Do Democrats and Republicans now hate the other side with no way to breakthrough? Or can we tone down our social divides with shared values? Jon Kingzette finds that negative perceptions of the other party are driven mostly by ideological differences and are targeted at the politicians in the party rather than ordinary citizens. We may not be so tribal after all. Jan Voelkel finds that liberal candidates can earn support by framing their policies with conservative values. We can gain support by signaling that we have more in common than it appears.
Matt Grossman: Reducing polarization with shared values, this week on the Science of Politics. For the Niskanen Center, I’m Matt Grossman.
We hear the partisans now hate the other side. Is there any way to reach the opposition? If polarization is based on differences in ideologies and values, we may be able to tone down the social distance or demonstrate that we share some of the other side’s predispositions.
This week I talked to Jon Kingzette of Ohio State University about his journal of experimental political science article, Who Do You Loath? He finds that negative perceptions of the other party are driven mostly by ideological differences and are targeted at the politicians in the party rather than ordinary citizens. We may not be so tribal after all.
I also talked to Jan Voelkel of Stanford University about his working paper with Joseph Mernick and Rob Weiler, Navigating the Progressive Paradox. He finds that liberal candidates can earn support by framing their policies with conservative values, even though most stick to one side or the other. We may be able to signal that we have a lot in common.
Kingzette starts by breaking apart affective polarization into end party love and out party hate.
Jon Kingzette: Actually, since 1978, there’s been consistent question wording that asks, “How do you feel towards the Democratic Party and how do you feel towards the Republican Party?” Two separate questions, but they’re both on the same scale where, from zero to 100, where zero indicates very cold or negative attitudes, 50 is neutral or the midpoint, and then 100 is very warm or positive attitudes. Because it’s been on the ANES for years as consistently since the ’80s, this has been how we’ve measured affective polarization. Typically, when people measure affective polarization, they have taken partisans’ attitudes towards their own party and towards the opposing party. So if someone’s a Democrat, they take their attitudes towards the Democratic Party, how they feel towards the Democratic Party on that scale, and subtract out their attitudes toward the Republican Party. That’s how they create a measure of affective polarization is take this partisan difference in attitudes towards the two parties.
The way I like to break it apart, instead of combining those into a single measure, is to instead look at what predicts partisans’ attitudes towards their own party and toward the opposing party separately. The reason why I do this is, first, there’s simply not that high of a correlation between partisans’ attitudes towards their own and opposing party in all timeframes, so I think the correlation in 1984 is something like .39. I think the correlation now is something like .65, so it’s something where there’s a little more validity to combining those measures today than there was, wow, that’s 40 years ago now. But if you want to look at trends over time, I think there’s good reason to break them apart into looking at in-party attitudes and out-party attitudes.
Matt Grossman: Most of the change is in out-party attitudes.
Jon Kingzette: Since the 1980s, partisans’ attitudes towards the opposing party or the out-group party have essentially monotonically declined, and especially since 2000, they’ve declined pretty rapidly. But partisans’ attitudes towards their own party have remained basically constant over the same time period with very small shifts year by year but there’s no consistent trend, so we basically have in-party attitudes remaining roughly the same over the last 40 years and out-party attitudes drastically becoming more negative, and especially in the last 20 years those attitudes have become more negative.
There is a slight caveat. I actually looked at this yesterday. Online samples report more negative attitudes towards the opposing party than telephone or face-to-face interviews, so a small part of that decrease in out-party attitudes, or I guess increasing negativity of out-party attitudes could potentially be simply how we are fielding the ANES survey, but most of it, I think, is genuine attitude shifting to become more negative.
Matt Grossman: Asking these questions about the party brings politicians to mind, not public partisans.
Jon Kingzette: Most of the time, around the ANES, these questions, I think importantly, are asked, “How do you feel towards the Democratic Party? How do you feel towards the Republican Party?” What I find in my research is that, when you ask those questions about the party, what everyone immediately thinks about is that you’re asking about politicians in the party. That’s what’s going to come to mind to survey respondents as they’re being asked to report their attitudes towards the two major parties in the United States.
I feel that an experiment in which I ask people to report their attitudes towards the two parties, towards politicians in the Republican and Democratic Party and towards what I called ordinary people who belong to the Republican and Democratic Party. What I find is that partisans’ attitudes towards the opposing party and opposing party politicians are essentially the same, but they have much more positive attitudes towards ordinary members of the opposing party. Basically, people get along or at least have more positive feelings towards the out-group partisans that they might actually come across during the course of their everyday life compared to politicians and the party that we typically ask about.
Matt Grossman: Voelkel finds that politicians usually run on liberal policies and rhetoric or conservative policies and rhetoric. But the best approach is to run on liberal policies and conservative values.
Jan Voelkel: We find that candidates receive the broadest support from voters if they link policies that are intended to reduce economic inequality with conservative values over talking about these policies and linking them to conservative values. In addition to that, we also find that currently candidates typically either run on economically progressive policies, so these policies that are intended to reduce economic inequality and value rhetoric among liberal values or on more economically conservative policies while talking about them in terms of conservative values, you find that people are not really on [inaudible 00:06:59] in terms of what would maximize voter support.
To put that into a bit more context, the idea from this paper was that we started with the observation that while Americans would like to live in a society with less economic inequality and typically support economic policies that would reduce economic inequality, they typically do not elect the politicians that run on these platforms of policies intended to reduce economic inequality. A big takeaway from that is that we often typically assume that certain types of policies and certain types of values go together, but actually the link between policies and value is it’s much more malleable than typically assumed. You can talk about policies intended to reduce eco…
So you can talk about policies intended to reduce economic inequality by linking them to moral principles or values such as reducing harm or economic fairness or equality. But you can also link such policies to values such as loyalty, respect for families or respect for traditions.
Matt Grossman: Moral framing is both possible and voters deem it reasonable.
Jan Voelkel: The first work that has been done on this was on the environment. Matt and Rob have shown that you can frame support for pro-environmental policies by linking them to the dimension of priorities so that you want to protect mother earth, that you want to keep the environment pure and [inaudible 00:09:01]. Yeah, and so this has been then extended more and more. And we were the first to do it in a context of political campaigns. And I mean, when we started this, this was a question for us, if we would be able to do this. And so we thought, well, you can argue for these policies from very different angles. You can say, and that might be the most straightforward way for people, because that’s the way that it is typically done in politics, is that we should increase taxes or we should increase access to healthcare because it will minimize harm that people are experiencing. These are people who are in need of more support.
But you can also talk about these policies such as more access to healthcare by saying that you want decent lives for Americans who provide for their families and who have served the US in the past. Yeah. So you could make a case for these progressive policies from a lot of different moral angles. And in the paper, we find that people perceive such arguments from different moral angles as similarly consistent. So they don’t see a candidate as more consistent if they argue for these policies in terms of liberal values compared to when they argue in favor of these policies in terms of conservative values.
Matt Grossman: Some candidates do differ in what values they discuss. So the possibilities are out there.
Jan Voelkel: By seeing that there was variation, for example, within the democratic party, but also within the Republican party, in terms of how politicians talk about the policy platforms. There was some evidence that it is possible to do so in a way that is understood as coherent and, to some extent, persuasive. So it doesn’t seem impossible to do it in the current political environment, to talk about progressive economic policies with more references to values such as harm and care versus more reference to loyalty and respect for traditions.
Matt Grossman: The sweet spot is in liberal policies and conservative values, but reframing could be a wider strategy.
Jan Voelkel: Linking conservative symbolic politics and liberal or progressive operational politics together may achieve the broadest base of support, because the majority of Americans support progressive policies and identify as ideologically conservative. So that seems to have the biggest potential for forming a broad coalition. But the replication of our study would certainly be interesting also because we find that using conservative value framing as a candidate who runs on economically progressive policies did not decrease support among liberal voters. So it might be that using liberal value framing as an economically-conservative candidate could read you as support among conservative voters because conservative voters may care mostly about this symbolic dimension, but that is something that we don’t know so far, and it’s something that future research needs to explore.
Matt Grossman: But Kingzette says that won’t work for everyone. Affective polarization is still real for some.
Jon Kingzette: It’s still a majority of partisans have negative attitudes towards the opposing partisans, so even towards ordinary people they might come across in the mass public. And there’s even, I think, still about 10-15% who report attitudes at zero. So they feel very cold, very unfavorably towards ordinary people on the other side. I am not all that concerned about that. So I think this is where I think thinking theoretically and really carefully about what these results mean compared what we might find harmful for democratic functioning is really important.
So it’s really cheap to just say, “Yeah, I dislike Republicans,” or, “I dislike Democrats,” on a survey. And I think what’s interesting, and one of the puzzles that I think I maybe haven’t solved quite yet but I think this research begins to sort of solve, is we have these measures of dislike towards opposing partisans. And yet we don’t see widespread discrimination in the mass public, or I guess I should say, I don’t think there’s widespread discrimination, right? People work alongside people in the opposing party. People have family in the opposing party that they might visit a little less or set boundaries that we’re not going to talk about politics or something like that, but they still visit those family members.
I’m not that concerned about people reporting their dislike on a survey. I think if those reports of dislike for opposing partisans really bled into behavior in a powerful way, then I would be concerned. So for example, let’s say that even 20% of people who have family members who are in the opposing party stop having relationships with each other because of their opposing partisanship. I think that would be a real problem. But I think the evidence is that those sorts of riffs in real life relationships are much, much smaller than the percentage of people who are willing to report negative attitudes on a survey.
Matt Grossman: And partisans don’t necessarily even like all the ordinary members of their own party.
Jon Kingzette: So I expected people to dislike the parties, disliked politicians who belong to the parties, but have more positive feelings towards ordinary people in the parties for both the out party and their own party. But what I find instead is that partisans actually have-
What I find instead is that partisans actually have much more positive attitudes towards their own party. So when they’re asked to report their feelings towards the Democratic party or the Republican party, they’re more positive then towards politicians in their own party and towards ordinary people who belong to it. And I think what’s going on here is that I think it’s actually a signal that the parties are good at what they do, at bringing coalitions of people together who might not share sort of policy positions or not share all policy positions, who might have sort of broadly different sort of perspectives about how politics ought to operate. And that those parties are actually really good at doing that and sort of garnering support for the party brand, even though it’s a big camp. And so, yeah, I think that’s part of it, is that on that end, it’s a signal that the parties are good at sort of building that brand up and making a variety of people excited about it.
But I think when people think about the ordinary people who… So it’s not that surprising that people report more negative attitudes towards politicians than towards the sort of broad party. That’s not too surprising because people despise politicians. But the surprising finding is that people report more negative attitudes towards their own partisans compared to the party. And I do think it’s because everyone can think about sort of ordinary people that they know in everyday life who they share partisan affiliation with, or who they might suspect they share partisan affiliation with, but they don’t like, or they don’t like their views, or maybe they’ve just angered them in some other way. And so, I think that’s part of it, is that they probably don’t share sort of policy positions or sort of broad perspectives with some of the people that they know also belong to their own party. So they report more negative attitudes towards those people.
Matt Grossman: Bipartisan identity drives in-party attitudes, but ideology explains out-party views.
Jon Kingzette: When it comes to partisans’ attitudes toward their own party, easily the most important predictor is the strength of their partisan identification. So do people call themselves strong Democrats, Republicans? Do they call themselves Democrats, Republicans, but won’t say they’re strong Democrats or Republicans, versus someone not even saying they’re Democrat or Republican in the first place, but then saying, “Yeah, I lean towards one of those parties at the onset.” The strength with which they identify with their own party is easily the driver of in-party attitudes, which makes sense. So there’s research in social psychology which shows that essentially in order to identify with a group, or if you identify with a group, you’re going to have positive attitudes toward that group, especially in the case of something like partisanship, where you essentially choose to enter that group. So pretty clear that partisan identification or the strength of partisan identification is driving in-party attitudes.
When it comes to out-party attitudes, the most important predictor is symbolic ideology, which I measure was sort of a seven point measure of ideological identification. So does someone call themselves a strong liberal to moderate, to strong conservative with sort of some midpoints in there? The most important predictor of out-party attitudes and the relationships are sort of in the direction you expect. So Democrats who are more liberal have more negative attitudes toward the Republican party, Republicans who are more conservative have more negative attitudes towards the Democratic party.
But I also find that policy positions, and I sort of created an aggregate measure of someone’s liberalism or conservatism across a range of different policies, also predicts partisans’ attitudes towards the opposing party, and partisan identification also predicts partisans’ attitudes towards the opposing party. So essentially, in-party attitudes are driven solely by partisan identification, but out-party attitudes are driven by both party ID, but also by ideology. And that ideology, both symbolic and sort of policy positions or operational ideology, both impact out-party attitudes.
Matt Grossman: Voelkel sees some options for reducing perceived ideological distance. Even a technocratic framing for liberal policies might be unifying.
Jan Voelkel: When we started out this project, we had another technical frame and it’s like what we called it, which was very much about we need smart, efficient policies that should ensure economic growth. We model this based on an op ed from Tim Kaine that we had read. And we thought this is a sort of amoral, alternative control condition. And then what we found in our research is that actually conservative participants also perceive that as being consistent with conservative values. And then it also increased support compared to this other more neutral control condition that we had. So we’re now interested in seeing if that is another set of values that could be efficient in the political domain.
Matt Grossman: And they’re working on research, linking values to a effective polarization.
Jan Voelkel: And now all they’re also doing a lot of research on the affective polarization part. One thing that we’re interested in is how effective the prioritization and also attitudinal polarization fits in with cohesion for democratic societies overall. If we want to explore that link more, as I’ve said before, we want to see if there are other values that are previously unexplored. For example, if technical principles or neoliberal principles have become moralized over time and is now another way, that you could link policies to such values that have been mobilized over time in order to shift typical positions on these issues.
Matt Grossman: Kingzette’s other work finds that affective polarization might be easily softened with some basic familiarity.
Jon Kingzette: These are sort of social distance items that are sometimes used to measure affective polarization. Basically, how comfortable are you or how happy or unhappy are you? Or how do you feel about having people in the opposing party that are sort of close to you in life in some way? So do you want to distance from them in your social relationships? Are you sort of comfortable with having members of the opposing party around?
And so, what I did is I asked one of sort of the hallmark questions here, which is how happy or unhappy would you be if you had a child and they were marrying someone from the opposing party? And I also asked sort of a new one that I came up with, which was how happy or unhappy would you be if you had someone who was in the opposing party being a neighbor? And a majority of people are actually unhappy with having an opposing partisan being a neighbor. 56% reported that they would be unhappy with that. But then I asked a followup question that says, now imagine these neighbors are kind of people who always say hello to you. How would you feel towards those neighbors given this information? And that number goes from 56% reporting that there’d be unhappy to 8%. And I don’t think those results are necessarily surprising, but I think what they reflect is that it’s just how unimportant political considerations are for most people during the course of their everyday life.
So you have neighbors who belong to the opposing party, and I actually said they also put up yard signs to support the opposing party presidential candidate. But if you just tell survey respondents that those neighbors are also nice, their level of negativity towards them just drops so far down.
And so, for me, this answers the puzzle of how most partisans can get along during the course of everyday life, even though survey responses of partisans are just so negative towards the other side. It’s simply that most people aren’t thinking about politics most of the time, they don’t really care about politics most of the time, and that positive nonpolitical information just takes precedence over negative information or the information that someone belongs to the other side.
Matt Grossman: And Voelkel acknowledges that in the real world, there might not be a big enough framing effect to counter our polarized politics.
Jan Voelkel: It’s only true that it is a big limitation, that we did it on a hypothetical candidate. We also used a candidate that was given a name that was a male candidate. So that’s [inaudible 00:25:32], another potential moderator. We don’t know if it would work for a woman candidate because we haven’t tested that yet, if there was any interaction in fact.
I will say that there is work in political science from David Brockman, Josh Kalla, who found that, in general, the effects of campaign advertisement and framing on voting decisions is very, very small and probably nonexistent or equal to zero due to … Especially in general elections when pocket queues are very salient. It’s like we did that and did an extra conservative test off our hypothesis by making it salient that the candidate we presented belonged to the Democratic Party. So participants knew that and our framing effects still showed up.
I would also say that the effects were, on a scale from zero to 100 among conservatives, the effect was roughly 10 points. I would say even though it may not be large enough to bring participants on a dichotomous scale, from do not support the candidate to support the candidate, a 10-point increase is quite big. I would say we don’t know if it would survive counter-framing. If there was a big campaign trying to argue against this sort of a value framing, we don’t know if the results will be robust to that.
As I said earlier, we know that participants did not perceive the link between policies and values as inconsistent, and we obtained the effect for a variety of measures. So also increase of principles and how well-liked and how competent a candidate was perceived.
So there needs to be more work done in order to see if such an effect could be obtained in the context of real campaigns with real behavioral indicators. But I do think that our results are enough to warrant more research in this domain.
Matt Grossman: Kingzette says some views are irreconcilable, but it’s useful to find and acknowledge them.
Jon Kingzette: On some policies, there’s a little bit more room for compromise and room for people with different values to meet halfway. And so, I think there we can have debate and negotiations about what ought to happen. But I do think that, especially on certain policies and especially people who really don’t share values, we are at a certain point irreconcilable. So we just have different views. We’re not going to come to an agreement.
But my perspective on it is that that is okay and that that is part and parcel of democratic politics, so long as people agree to the rules of the game. So I think a lot of what the Republican Party is doing now is trying to undermine democracy, basic institutions and processes, which I think those irreconcilable values with the small D democratic values that most of the Democratic Party has, or at least has heaps and bounds more than the Republican Party now, there should be a common set of rules and norms that I think is not okay if there’s differences.
But then on policy, I just view it as part of democratic politics, is that at certain points, people are going to have different values and I think we should try to convince others that our values are right. I think that’s part of the game.
But sometimes we’re just going to stand at odds, and that’s okay. I think that’s the number one thing that I would like to bring across here is that I think there’s so much punditry that decries that we can’t come to agreements when a lot of people think very different things and have a different set of values.
I think it’s uncomfortable to recognize that there’s irreconcilable differences and that there’s not always going to be room for good compromises sometimes. I think oftentimes there is, but I think people get uncomfortable with that idea and then say that it makes them fight even harder to try to find compromise and to find policies that everyone can agree to. I think it’s important that at a certain point we accept difference rather than try to fight against it.
Matt Grossman: There’s a lot more to learn. The Science of Politics is available biweekly from the Niskanen Center and part of the Democracy Group Network. I’m your host, Matt Grossmann.
If you like this discussion, you should check out our prior episodes on Polarized Opinion on Climate Change and Messages that Move Conservatives and Are Americans Becoming Tribal with Identity Politics Trumping All. For now, thanks to Jon Kingzette and Jan Voelkel for joining me. Please check out Who Do You Love and Navigating the Progressive Paradox, and then listen in next time.