As Election Day approaches, Trump intimidation efforts are increasing and Americans in both parties are worried that the other side could use unfair tactics to sway the election. Why does the public fail to serve as a check on anti-democratic practices? Matt Graham finds that only a small fraction of voters prioritize democratic principles over partisan and ideological interests. And by increasing ideological differences, polarization has hurt democracy’s valuation. Larry Bartels finds that large numbers of Republicans countenance anti-democratic moves. And it’s not based on their support for Trump, but their broader ethnic antagonism. They both say we should worry about American democratic backsliding among elites because public support for democracy won’t save us.

Guests: Larry Bartels, Vanderbilt University; Matt Graham, Yale University

Studies: “Democracy in America?” and “Ethnic Antagonism Erodes Republicans’ Commitment to Democracy.”


Matt Grossmann: When and why do Americans accept anti-democratic practices? This week on The Science of Politics. For the Niskanen Center, I’m Matt Grossmann.

As election day approaches, Trump supporters are egging on voter intimidation and Americans in both parties are worried that the other side could use unfair tactics to sway the election. How worried should we be about American democratic breakdown? And why does the public not serve as a check on anti-democratic practices?

This week, I talked to Matt Graham of Yale University about his new American Political Science Review article with Milan Svolik, “Democracy in America?” He finds that only a small fraction of voters prioritize democratic principles over partisan and ideological interests. And by increasing ideological differences, polarization has hurt democracy’s valuation.

I also talked to Larry Bartels of Vanderbilt University about his new PNAS article, “Ethnic Antagonism Erodes Republican’s Commitment to Democracy.” He finds that large numbers of Republicans countenance anti-democratic moves, and it’s not based on their support for Trump, but their broader ethnic antagonism.

Graham found that voters don’t punish politicians much for violations of democratic principles.

Matt Graham: Our article is about how strongly voters are willing to support democratic values and something that I think is important to understand about the approach we’re taking is that governments can be more or less democratic. So the traditional image of democracy dying as a military coup, where it happens in an instant and democracy just no longer exists. Today, when democracies fall, it tends to be elected leaders breaking them down from within, degrading democracy slowly. And that gives voters a chance to punish politicians who are working to make the system less democratic. And so our approach is designed to examine voter’s willingness to support democracy in that context.

Now, traditionally researchers measured support for democracy by directly asking people things, like, how important is it for you to live in a democracy? And even in this moment, people tend to overwhelmingly say that democracy is important to them, but in the real world, people don’t just get the opportunity to choose more democracy or less democracy. At the ballot box they’re making a bundled choice. And so our approach to learning about support for democracy was to put people in situations where they revealed their support for democracy through their choices, rather than just stating it directly.

And so we asked our survey subjects to make a series of choices between hypothetical candidates, they each had a political party, policy positions, and other attributes like age, race, gender, a career. And we found that people reveal a much lower level of support for democracy than they state. And in the scenarios we had that were most like the real world with polarized politics, you had a Republican taking two conservative positions, a Democrat taking too liberal positions, we found that only 3.5% Of our sample changed their minds when one of the candidates was randomly assigned to take an undemocratic position. And the message that I take from these findings is that people do care about democracy, but they also care about other things. And in the kind of polarized political environment we have today, where Democrats and Republicans are so clearly different from one another, they’re so far apart on the issues and in terms of what groups they stand with, the combined force of party and policy, and all these other factors tends to outweigh people’s otherwise desire to support democracy.

Matt Grossmann: Bartels found that many Republicans are willing to support anti-democratic statements.

Larry Bartels: There are two important findings here, I think. One is not unfamiliar to political scientists, but taps a dimension of a broader phenomenon that people have studied off and on for the past five or six decades, I guess, which is about the attachment of ordinary Americans to democratic values and norms of one kind or another. We’ve often been struck in our research by finding that people are strongly attached to democracy at a very broad abstract level, but that when it comes to actually putting democratic principles into practice, they tend to go astray much more often, either due to confusion or lack of understanding of the norms, or probably more often because they have other values and political priorities that conflict in particular circumstances with what we take to be democratic norms.

And so in the survey that I did in January, I tried out on a sample of Republicans and Republican leaning independents for questions about democratic norms and found a fair amount of agreement with statements that seemed to contradict important features of the way we typically think about the democratic process and the appropriate role of citizens in it. So that’s the first general finding, that there’s just a lot of anti-democratic sentiment out there.

The second finding is that I tried to figure out where these attitudes come from or what it is about these people and their political views that seems to promote anti-democratic sentiment. And looked at a broad range of possible explanations, but was surprised to see how clearly the one that stood out was, what I call, ethnic antagonism, concerns of various kinds about the role of African-Americans and immigrants and other minority groups in American society and politics these days.

Matt Grossmann: For Graham, the key factor was ideological alignment. If one candidate is much closer to you on the issues, you’re less influenced by their anti-democratic moves.

Matt Graham: The important aspect of a polarization is that it raises the stakes of elections, it makes voters feel like there’s a really clear choice between candidates and they might feel much closer to one side or the other. So in our research design, we took some steps to see how people’s willingness to support for democracy varies with the stakes of elections, if you will. So one to three weeks before we did the candidate choices, we measured people’s views on the policy positions that the candidates would later take. And then because the policy positions were randomly assigned, our subjects for sometimes in situations where one candidate was much closer to their views than the other candidate. And in other situations, there wasn’t much difference, the candidates would take two similar policies or one candidate would have an advantage in one area and a disadvantage in another area.

And so people supported democracy the most in situations where there weren’t big differences on the other dimensions. Democracy is most important to somebody’s choice when they just barely favor one side over the other. And then they see the undemocratic position, they’re not all that strongly committed and they change their mind at a pretty high rate. But when you’re in a situation like many voters are in, where the choice between Democrats and Republicans is just extremely stark, you might see something that your party is doing, and you might not like that, but the other reasons are just so stacked up in favor of your party that you still vote for them. We’re able to split out our results into situations where the voters are seeing relatively large and relatively small differences between the candidates and when people are relatively indifferent between two candidates, democracy matters a lot. But when people see a really strong reason for policy or partisan reasons to support one candidate over the other, it’s a lot harder for democracy to affect their choice, even if they do care about it. They also care about other things.

Matt Grossmann: Bartels found that the best predictors of Republican views, among many he tested, were related to ethnic antagonism.

Larry Bartels: There were a bunch of questions on the survey that tapped a variety of different specific attitudes and what I did was to try to group them into broad categories. I had a measure of people’s attitudes about President Trump specifically that were based on approval measures of various kind and perceptions of his character on a variety of specific dimensions. And then a variety of questions about partisanship and their attitudes about Democrats and Republicans and the Democratic and Republican parties.

Then there was a set of things about economic conservatism, their view about the role of the government and the economy and healthcare and the environment, and so on, attitudes about business and labor unions. And then a set of attitudes about cultural conservatism, that was a range of things touching on traditional moral values and religion, abortion, gender attitudes, but also attitudes about tradition, patriotism, the relative affection or closeness that people felt toward people in rural areas and people in cities, for example.

And then a set of questions about political cynicism, their views about politicians, generally, regardless of parties and how the political system works. And then finally this measure of ethnic antagonism, which was tapped by a variety of items, some of which are familiar kinds of questions that political scientists have asked about race, feelings of favorability or closeness to particular ethnic groups, for example, but variety of others as well. And one set of questions that seemed particularly powerful idea of others as well. And one set of questions that seemed particularly powerful in terms of capturing this in my analysis was questions about whether specific groups had unfair political influence or unfair access to government resources. And it turned out that people who said that African-Americans and immigrants and Latinos and people on welfare were too powerful and were getting more than their fair share of stuff from the government were classified as being ethnically antagonistic. And those attitudes tended to overlap quite a bit with more basic feelings about these groups, but also things like the perception that speaking English is essential to be a true American. And that African-Americans are using racism as an excuse for not working hard enough. And that discrimination against whites is a big problem and so on. So it’s a kind of a broad range of specific attitudes, but they seem to fit together pretty strongly and to be a strong predictor of these anti-democratic sentiments.

Matt Grossmann: Let’s look into the details starting with [inaudible 00:01:14]. They find that anti-democratic moves are just treated like other scandals, bad, but not necessarily influential.

Matt Graham: The structure of our approach doesn’t necessarily depend on undemocratic behavior being the bad thing. You can think of voters making trade-offs between party and policy and anything that a candidate might do that you don’t like. And so in our case, we had two benchmarks that aren’t strictly related to democratic values. We had some candidates be presented as having had multiple extramarital affairs. We had candidates who were convicted of underpaying their taxes. And our subjects tended to dislike those things as much, if not even more, than our democracy positions. And I think what this really sheds a light on is why other scandals don’t seem to always matter as much as you would like them to.

So for instance, we saw Bob Menendez post to reelection and New Jersey despite some pretty significant scandal. And so I think an important lesson from all of that is that we shouldn’t assume that people who don’t change their vote choice over something don’t care about it. So when we see social conservatives voting for President Trump despite some of his obvious inconsistencies with their values, we shouldn’t infer that those people don’t really value those things. We should be cognizant that people will often hold their nose and choose something that they view as the lesser of two evils.

Matt Grossmann: There are several examples of anti-democratic moves that voters discount relative to agreement with the candidates.

Matt Graham: Breakdowns of checks and balances, breakdowns of rights, ultimately make it harder to sustain a democratic system. There are symptoms of democratic decay, and there are also things that make it easier for elected autocrats to consolidate their power. And so there are things that are popular will themselves, and there are also things that help sustain popular government even if they’re not the most pure conception of the popular will. And so I think that’s the sense in which checks and balances and civil liberties are really important. Now, as far as whether undemocratic behavior can be justified by the other side’s behavior, the dynamic that we’re in right now is very difficult. And we have two partisan teams that have been doing one another injury for a century and a half. I maintain a pretty balanced media diet. Both sides have endless wells of examples of how the other side’s behavior justifies their own. And the argument’s that we can’t just surrender while they cheat.

They do have some merit. This is a really pernicious dynamic we’re in. Some of the work by Jennifer McCoy and Murat Somer is really valuable here. I recently read a thought provoking piece by Adrienne LeBas argued that in Burkina Faso it was the opposition’s consistent willingness to counter the ruling party’s moves with mass protest was crucial to the elites ultimately reaching a settlement. And in the larger comparative literature on why countries democratize or why countries remain stable, elite commitment to democratic norms on both sides or however many sides there are, is really important. And so you can’t sustain democratic norms with only one party upholding them. And so sometimes these downward spirals get to a point where the elites decide we need to do something better, we need to improve the mechanisms through which we resolve our conflict. And that’s how you really build elite support for democratic norms. If both parties view the other party as not upholding them, and both parties are willing to cheat or maximally exploit their advantage, it’s a really difficult dynamic.

Matt Grossmann: They also used a Montana real-world situation to test their theory in the wild.

Matt Graham: The situation in Montana provided a convenient real-world opportunity to test our framework. And what happened there is that the night before the election, the Republican candidate, Greg Gianforte, assaulted a journalist who was repeatedly asking about Gianforte’s healthcare position. And absentee voting is very common in Montana. And the absentee voters didn’t know about the assault but the election day voters did. And so we collected vote totals from five counties in Montana, I think 87 precincts, that were separated by absentee versus election day. And that allowed us to use a difference in differences strategy to estimate how the assault affected the choices of voters who likely knew about the assault when they voted. And we found very similar results to what we saw in the candidate choice scenario, which was a lot more artificial.

Now, in terms of how well the Gianforte situation fits our framework, one way to frame it is that it’s an instance of political violence. It displays contempt for the free press, both political violence and contempt for the free press are inconsistent with democratic values. But it is admittedly not as clearly undemocratic as the positions that we chose ourselves. And so the model doesn’t say that the bad thing that the candidate does has to be undemocratic. We saw in the more artificial scenarios that voters do apply a similar logic when they’re evaluating a bad thing that a candidate does that’s not strictly undemocratic. And so the Montana experiment clearly shows that the basic logic of our framework plays out in the real world. And we’re ultimately limited to the universe of things that actually happened the night before the election and lend themselves to a good empirical strategy.

Matt Grossmann: Bartell’s study looked directly at agreement with several anti-democratic statements, with some differences in how many Republicans support them.

Larry Bartels: I’ll just read them to you. People were asked first to agree or disagree strongly or not strongly with each of these items. The first one says the traditional American way of life is disappearing so fast that we may have to use force to save it. The second one, a time will come when patriotic Americans have to take the law into their own hands. Third, strong leaders sometimes have to bend the rules in order to get things done. And fourth, it is hard to trust the results of elections when so many people will vote for anyone who offers a handout. So the responses to those items tended to cluster together. The last one about the legitimacy of elections generated stronger support among these people than the others. About 50% agreed that the traditional American way of life is disappearing so fast that we may have to use force to save it. And a similar number said that strong leaders sometimes have to bend the rules in order to get things done.

But over 70%, almost three quarters, agreed that it’s hard to trust the results of elections when so many people will vote for anyone who offers a handout and less than 10% disagreed with that item. Generally, the disagreement rates range from 10 to about 20 or 25% with the rest saying that they were unsure or didn’t either agree or disagree. So there was a difference just in terms of the overall level of support for that one by comparison with the others. In terms of the way they were related to these various political norms and values, in every case, ethnic antagonism was the strongest single predictor of agreement with all four of these statements but there were some differences in detail with respect to the other domains. For example, people who are most enthusiastic about President Trump were the most likely to say that it’s okay for strong leaders to bend the rules. That’s probably not surprising. People who scored high on political cynicism were more likely to distrust the result of elections. But as I say, in each of these four cases, ethnic antagonism was the best predictor of people’s responses.

Matt Grossmann: He says, Republicans likely have a different route to different anti-democratic attitudes than Democrats.

Larry Bartels: I think it would be interesting to look at the nature of anti-democratic sentiments among Democrats, but I think it would require rather different questions. The kinds of values and frustrations that are tapped by these questions, I think, are really much more likely to be salient on the Republican side than for Democrats. And I suspect the explanations would be quite different as well. If you look at the distributions of ethnic antagonism, they’re vastly different for Democrats and Republicans. 98% of the Democrats scored below the Republican average and more than 98-

Scored below the Republican average and more than 98% of the Republicans scored above the Democratic average on ethnic antagonism. So these are very different distributions, it’s not even clear to me that it makes sense to call it ethnic antagonism at the opposite extreme where many of the Democrats fall. But it does seem possible that their attitudes about social groups and about the role of minorities in the country and about cultural diversity would factor into their views about democracy. It might be related to their responses to questions, the kinds of frustrations that they’re more likely to have about the current democratic process.

Matt Grossmann: Even if behavior doesn’t match attitudes for all of these people, it could still have a big impact.

Larry Bartels: Most of these people are not about to go out and engage in lawlessness or violence, but even if it’s one 10th of 1% of these people that amounts to about 30,000 people, that’s not enough to bring down American democracy, but it’s enough to create a hell of a lot of political turmoil. On the other hand, I think the sort of distinct reason to be concerned about this is that it seems plausible, at least, that this kind of sentiment in the general population, or specifically among the supporters of a particular political party or political leader might encourage elites to take anti-democratic actions.

I think in concrete terms, the most immediate threat is probably not from ordinary citizens themselves, but from political leaders who are in a position to take actions that will erode democratic processes and institutions in some way. And although we can’t be sure, it seems plausible that having a basis of support for those kinds of actions, even in the kind of casual way that they might be expressed in surveys would increase the probability that political elites would be tempted to engage in that kind of action.

Matt Grossmann: Graham also sees real lessons for today.

Matt Graham: Our framework applies to what we’re seeing now, because voters are making a choice between two starkly, different versions or visions for the country. A lot of people know what side they’re on, whether that’s because of their policy positions, whether it’s because of their social identities, they know which party supports their religion or their race, maybe it’s the combined weight of all of those things. And those attachments that people have in the real world are often maybe even usually much deeper than any attachment we could generate to a candidate through some randomly generated text on a screen.

And that gives voters a really tough choice. And so to me, it seems that there are plenty of voters out there who strongly disapprove of many of the things that president Trump says and does, including his clear lack of commitment to democratic principles, and they’re going to vote for him anyway. And I think our framework sheds light on exactly why people are willing to hold their nose and do that because they value more than one thing and sometimes one thing outweighs another.

Matt Grossmann: Both analyses, according to Bartells, give a permission structure for expressing these views.

Larry Bartels: It will be interesting to know to what extent people would be willing to agree flat out that it’s okay to violate democratic norms in order to achieve their own political ends. I think one of the reasons that these questions that I asked resonated with people is that they didn’t frame the choice in quite that way. They alluded to values that are hugely important for a lot of Republicans and pit them against democratic norms, but in a way that may not have forced them consciously to realize that they were making a choice between their values and democracy as a system.

Similarly, the Graham and Slavic piece presents people with these hypothetical choices. And I think to the extent that that works, it probably works because it doesn’t hit people over their head with the idea that they’re making a choice specifically between their partisan political preferences and their respect for democratic norms. These factors are kind of mixed in with a variety of other things that people are told about these hypothetical candidates. And that might provide a more realistic sense of things than to ask them straight out, is it okay to violate democratic norms if our party is going to benefit from it?

Matt Grossmann: Graham agrees that both studies are important, despite different designs.

Matt Graham: The two approaches are different and I think there’s value to both of them. The approach that Milan and I took tried to look at the structural obstacles that polarization presents to electoral accountability for violations of democratic norms. And we tried to choose positions that weren’t clearly associated with one party or another, because we wanted to know about the underlying dynamics. We wanted to get closer to the question of, are Democrats and Republicans inherently more or less willing to countenance this sort of thing. And approach like Larry’s approach is really important because it looks at the specific dynamics of what’s happening now.

And it also sounds quite important just because there’s such a long history of race playing in to undemocratic things that Republicans now do and Democrats did in the past when that dimension was flipped. One thing that is important to recognize about the situation that we’re in right now is, especially at the state level, in terms of ballot access, in terms of opportunities to gerrymander, Republicans are in situations where disenfranchising people serves their electoral interest. And Democrats are in situations where giving people the franchise, it serves their electoral interests. And so the way the self-interest lines up, I think really affects how these dynamics play out and I’m glad other people are studying that.

Matt Grossmann: Martell says there’s lots of threats to democratic systems, some general and some specific to US Republicans.

Larry Bartels: I think there are three levels here. I think it’s probably a pretty consistent feature of democracies that although people express verbal attachment to democratic norms in the abstract, they turn out in practice to be pretty willing to prioritize their own substantive political values and preferences when push comes to shove. I think there’s a general pattern in democratic systems that ethnic conflict is dangerous and makes it hard for democracies to work successfully, there isn’t a kind of distinguished record of successful multi-ethnic democracies.

But I think the circumstances of the contemporary Republican party are distinctive in that there’s been so much attention over the past decade, I suppose, to demographic change. And in particular to this shadow that many people find so threatening of a Democrat demographic transition in which we end up no longer being an essentially white America, but rather a majority-minority country. For people who are alarmed about that prospect, the political stakes are very high and so I think that’s one of the most important inflammatory factors in the current political saturate.

Matt Grossmann: But either way, these aren’t attitudes we all agree to in advance of political competition.

Matt Graham: A lot of the early work on support for democratic norms interpreted them as consensual features of American political culture and found that there was generally higher agreement on people who were better educated and more engaged with the political process. And the people who did this work often interpreted a lack of agreement with norms as being kind of indication of misunderstanding or being disconnected from American political culture. I think more recently, we’ve come to recognize that people’s attitudes about these broad process, questions and norms are strongly related to their substantive political values.

And here it turns out that well, these political attitude dimensions that I measured are, in some cases strongly predicted to their responses, the kind of standard demographic factors like education and urban versus rural locale turn out not to be very strongly related to their responses. Given the distribution of these characteristics in the sample that I had, the modal person who was agreeing with these anti-democratic statements was somebody who had some college education and lived in the suburbs and paid attention to politics, not somebody who was uneducated and isolated from political discourse.

Matt Grossmann: They both agree that the current anti-democratic attitudes are not all specific to Trump, though Trump has enhanced it versus Bartell’s.

Larry Bartels: Yes. I tend to think of the president more as a symptom that has a cause here, I think this feeling has been growing among Republicans for some time. I think Trump successfully appealed to it in the course of the 2016 campaign and since, but I don’t think he’s the cause of it.

But because of it, it may be that people are more willing to endorse these statements now than they would have been without him on the scene. It’s kind of legitimized, anti-immigrant, anti-minority rhetoric in a way that certainly previous Republican leaders mostly haven’t.

Matt Grossmann: Graham says Trump supporters are like all of us, willing to put other interests ahead of the anti-democratic moves.

Matt Graham: There’s a narrative that is driven by people who pay attention to politics all the time. They’re on Twitter all day every day. Their eyes are glued to cable news. They can’t avoid hearing about these things all day and all night. And then they look at the stability of President Trump’s poll numbers or something like that. There’s this narrative that President Trump’s supporters must not care about any of this stuff. They must be okay with it. They saw this thing happen or they saw that thing happened and they didn’t immediately become Democrats. And so they must condone it, or I guess maybe they do condone it, but they must like it, they must think it’s good. And I think what our framework really does a nice job of showing is the fact that a political development doesn’t change somebody’s mind doesn’t mean that that person doesn’t care about it.

In elections, people are making a binary choice. When people are asked in a public opinion poll whether they approve or disapprove of the President’s performance in office, they’re making a binary choice. And so I think it’s wrong and I think it’s irresponsible for people to infer from the stability of President Trump’s support that everybody who remains stable in that support must approve of all of these things that they see the President doing. Because one, those people who remain steady in their support probably aren’t following politics as closely. And two, the fact that they remained steady in their support absolutely does not mean that they like everything that the guy does.

Matt Grossmann: They both study the public, but also both agree that elites are more important. For Bartels it’s still useful to know that the public just won’t be a check on their moves.

Larry Bartels: I’m generally a believer in top-down models of politics. I actually just wrote a paper focusing on Europe, the title of which is Democracy Erodes from the Top. So I think the action of political leads is certainly most important in the immediate sense. I guess I would say what’s struck me in some ways most about the Trump administration is the starkness of the distinction we’ve seen between the behavior of on one hand people whose political positions and careers don’t depend directly on electoral politics. So judges, bureaucrats, journalists, all of whom have, I think sometimes, clumsily and ineffectively, but pretty actively and genuinely tried to stand up for democratic norms and democratic procedures when they saw them under threat. And on the other hand, elected politicians who seem generally very unwilling to pay any real partisan political price for standing up for democratic norms.

Matt Grossmann: Graham says the dynamic they found applies globally, but elite incentives may still be most important.

Matt Graham: The framework is drawing on some extremely basic facts about human decision-making. In any trade off that people make, if you have a lot stacked up on one end of the trade-off it’s going to be harder for any single factor to move you in the other direction. And Milan has shown that this same result holds in Venezuela, he’s doing some work in Turkey, he’s doing some work in Tunisia. And so to whatever extent electoral accountability does matter, I think we’ve shown that polarization is a serious obstacle to it. But you’re absolutely right that much of the comparative literature focuses on elite incentives to support democracy, institute it, uphold it. And the public’s views of whether something is or is not undemocratic and whether they should get out in the streets are heavily conditioned by opinion leadership, by somebody or another that as political scientists we usually refer to as the elites.

And so I certainly wouldn’t want anyone to look at our work and think that we think that public support is the only factor, but if we’re going to figure out how much public support matters, I think it’s important to be able to measure it well and understand the dynamics that affect it. And popular movements are really crucial in these things. Even in the elite-focused models in the comparative literature, what you’ll often see the elites balancing the cost of is, “Am I going to have an autocratic government in which I have to bear the costs of putting down mass protests every once in a while?” Or, “Am I going to get less for myself in exchange for the cost of not being the ruler?”

Matt Grossmann: Next, Bartels wants to look at Democrats.

Larry Bartels: Yeah. I’m interested generally in the question of how people think about these potential norm violations and I do think it would be interesting to look at Democrats and try to figure out both what particular kinds of norm violations they would be likely to accept and what the motivating factors for those turn out to be.

Matt Grossmann: And Graham wants to move towards solutions and avoiding overstating the problem.

Matt Graham: One of the next steps in my research agenda with Milan is to examine the role that social norms play in support for democracy. And I think that sensationalist coverage of work like ours that uses the finding that only three and a half percent of people change their mind in the most polarized settings, to say that only three and a half percent of people care is destructive. I think it might be doing some real harm. And so for this work specifically, I want to move it in a more solution-oriented direction. I want to find ways to get people to place a higher value on democracy. I want to figure out what makes voters want to hold their own side to a high standard. I want to be trumpeting that because if my role as a social scientist is just going to be to throw more fire on these problems then I would want to do something else with my life.

And in the bigger picture, I think the coverage of this work I’ve done with Milan is not the only instance in which sensationalist coverage of survey data might be undermining faith in democracy. When I turned in my dissertation yesterday, the argument there is that misinterpretations of survey data are dramatically overstating the size of partisan divisions and the prevalence of false beliefs and misperceptions, belief in politicized rumors. And I think too often, the instinct to run to the most pessimistic interpretation of the data and the most pessimistic possible portrayal of people is actually a source of misinformation. And I want to quantify the social harm that I see emerging from the systematic abuse of survey data in the public sphere.

Matt Grossmann: There’s a lot more to learn. The Science of Politics is available biweekly from the Niskanen center. I’m your host, Matt Grossmann. Please check out our recent episodes at or anywhere you find your podcasts. Thanks to Matthew Graham and Larry Bartels for joining me. Please check out “Democracy in America?” and “Ethnic Antagonism Erodes Republican’s Commitment to Democracy,” and then listen in next time.

Photo Credit: Donald Trump / Public domain