Do Americans Support Democracy As Much As They Say?

The storming of the Capitol and the votes by some Republican elected officials questioning the results of the 2020 election have many asking what force could act as a check on these increasing anti-democratic tendencies in American political life?  

Milan Svolik, Prof. of Political Science at Yale, may hold some answers. He investigates whether the American public would act as a check on anti-democratic politicians, and reveals how much we truly value democracy when we’re presented with tradeoffs.

Transcript

Anthony Fowler:

I'm Anthony Fowler.

Wioletta Dziuda:

I'm Wioletta Dziuda.

William Howell:

I'm Will Howell, and this is Not Another Politics Podcast. So what an extraordinary week. We've seen the president, in the aftermath of spreading all kinds of disinformation and lies about an election, helped organize and then gave a speech in which he directed a mob to descend upon Capitol Hill that then took over Congress. During those events, a number of people died. It had all the symbolism and feel of an insurrection, and it raised lots of big questions about the role of race in our politics or about parties and party factions and raised profound concerns about the health of our democracy, the stability of our democracy, the endurance of our democracy, things all that we've tried to grapple with on this show of ours.

Anthony Fowler:

Absolutely. If you've been watching political news for the last week or even the last four years or more than four years, you probably get the sense that there are a lot of Americans who don't care very much about the health of our democracy. I mean, of course, a violent mob intentionally trying to invade the capital building and disrupt the certification of an election is a horrible manifestation of that, but it seems like a good time to wonder about the state of democracy but also the extent to which Americans actually care about democracy. Do we have any compelling evidence or any reasons to think that things are as bad as they seem on television?

Wioletta Dziuda:

Yeah, so I think, for all of us, if I say that this is not how we planned to start our podcast in 2021, but by a strange coincidence, we were planning to start with a paper that actually asks the question, "To what extent are voters willing to trade off the qualities of their perfect candidate for some changes in how democracy works and for some antidemocratic behavior of the politicians?" This episode that you are going to hear is an episode that we've recorded before the events of the last week, but this is going to shed some light on what happened last week and maybe how we should learn from those events and how optimistic or pessimistic we should be about them. So, yeah, without further delay, Will, you've talked to someone who was trying to tackle those questions.

William Howell:

I did. I spoke with Milan Svolik, who is a graduate of the University of Chicago. He's now a professor at Yale in the political science department. He and Matthew Graham have a paper that came out just this past year in the APSR, the American Political Science Review, entitled Democracy in America: Partisanship, Polarization, and the Robustness of Support for Democracy in the United States, in which they try to grapple with precisely the issues that you two have raised, which is, namely, can we vest any hope that the public will step in and function as a kind of guardrail of last resort against the entreaties of a demagogue? So why don't we give the interview a listen. Milan, welcome. It's really a pleasure to have you on the show.

Milan Svolik:

Thanks for having me.

William Howell:

So can you just say a little bit about your work in this area and what brought you to this particular paper and the particular question that you're asking in this paper?

Milan Svolik:

Yes. So I study democracy authoritarianism around the world, and one of the most prominent trends that we observe is that when democracies break down these fates, it is not one single sharp event, but it usually a process that starts by a democratically elected incumbent gradually undermining democracy. That is, it is somebody who previously had to gain a majority or a plurality of votes in a perfectly democratic election who then uses the power of the office to try to undermine democracy and perpetuate his position in power. We try to examine in the United States can we count on the public to serve as a democratic check, and what do we mean by that? If an elected politician proposes to do something that violates our key democratic principles, are Americans willing to punish that politician that they would otherwise vote for by defecting from that politician?

Early on after political scientists started doing surveys, this is the 1940s, '50s, '60s, they also started asking Americans and later citizens around the world about their commitment to democratic principles. Ever since then, Americans score very high, around eight out of 10 on average recently, but, crucially, it's so also in most of the world. So one problem I see with asking about democracy in this way, especially in the United States, is Americans know that the politically, the socially acceptable way to answer that question is in only one way, and that is yes, I think it's very important to live in a country that's governed democratically. The problem is, first of all, it's at a highly abstract level, so it's a very general question, but, crucially, it doesn't capture what you're willing to sacrifice for democracy. What is the price that you're willing to democracy?

Crucially, when we observe democracy being undermined around the world, the choice is not between candidate A and candidate B and the only difference between them being that one is undemocratic, right? Usually, what you're posed with is a trade-off where you really like candidate A but candidate A is doing something undemocratic and then there's candidate B, who you don't like for many policy, partisan reasons and who is democratic. The question is in a choice like this, even though you say it's important for you to live in a country that's governed democratically, is it important enough so that when candidate A does something undemocratic you're willing to say, "Okay, and I'm willing to punish candidate A for that by voting in this particular election for candidate B."

William Howell:

Okay. Okay, so that's what you want to investigate, and you do it in two ways. One is through a survey experiment, and then the other has to do with something of a natural experiment. You take advantage of this curious event that happened in Montana. We'll get to that in a bit. But tell us first about this candidate choice experiment that you run. How did you set things up, and how, in setting it up the way that you did, were you able to evaluate these different kinds of claims?

Milan Svolik:

We took a representative sample of Americans. Each of those respondents was faced with a choice between two politicians. Each politician was described by the attributes that we typically see in elections, so they had, as I mentioned, a party, an economic policy, a social policy. They also were described by their age, gender, race, and occupation and experience in that occupation. But then, crucially, each of them was also randomly said to have endorsed a violation of a democratic principle, so something that either violated electoral fairness or checks and balances or civil liberties.

As a result, we wanted to look at what percentage of our respondents are willing to, as I mentioned previously, defect from a candidate who proposes something undemocratic and at the level of the electorate that would give us the percentage overall of voters who are willing to punish undemocratic behavior and, as a result, the aggregate punishment that a candidate who would dare to behave undemocratically could expect to face from an electorate. Once we have this representative electorate making these choices between candidates, we can examine them then by the intensity of partisanship in a subset of the electorate so we are able to divide the electorate into partisan centrists and partisan extremists. We are able to look at the policy preferences of these voters, and we are able to order them from left to right on those who are extreme leftists versus centrists.

What we see throughout is very much consistent with our theoretical expectations. If you are a policy extremist or if you are a strong partisan, that subset of our sample is less willing to punish a candidate that they like but who proposes something undemocratic by saying they would vote for a candidate that they otherwise wouldn't vote for, whereas centrists ... And centrists here means either policy centrists or voters who in a survey say they're independents and either they are what we would call pure independents or partisan leaners, that is, they say they are independents but when asked one more time, well, they would say, "I lean Republican," or, "I lean Democrat." These are the voters that punish the most. In the aggregate, what we find is that roughly 11.7% of our representative electorate is willing to defect from a candidate who would propose something undemocratic.

William Howell:

Okay. And so the effects are most acute or largest in magnitude for moderates. But, of course, we know that extremists aren't in play, they're not subject to much mobilization efforts, except in terms of turning out the base, but not in terms of persuasion, that elections in the main are decided by moderates. Could I read these findings and take comfort in what you've found? That is, can I say, "Look, these moderates" ... Actually, an effect of 10 to 12%, just by virtue of one antidemocratic position that's being articulated, that's a big deal, no? That's the stuff of elections swinging from one candidate to another.

Milan Svolik:

So there are two answers. One is that what you just described, the effect that I described as 11.7%, is our most optimistic estimate. This is the estimate that we get when we completely randomly assign policy positions, partisanship, and undemocratic positions. The possible issue here is that, as a result, we get candidates that we don't always see in the real world, right? So we sometimes get a Democrat who is proposing to abolish taxes or we get a Republican who wants to legalize marijuana or something like that. An exercise we do in the paper is that once we have this total set of experimental candidates, we try to reduce the sample to those that look like what we observe in the real world. What this means in our context is that it's always a Democrat running against a Republican and the Republican, of the four possible positions from left to right that our candidates could have, only adopts the two on the right and the Democrat the two on the left.

Once we induce this party policy alignment, we see that the punishment actually drops to 3.5%. This is a big drop, right? This now means we're only three-point-something percent away from effectively not punishing a candidate for doing something undemocratic. Crucially, if you look at the margins by which congressional districts, for instance, are won, only a small fraction of congressional districts in the United States are won by a margin that is smaller than that or just as small as that. In other words, in the vast majority of congressional districts, a candidate could do something undemocratic or the larger party, the majority party, could do something undemocratic and get away with it.

William Howell:

And be just fine. And get away with it. Okay, so let's pack our bags and go to Montana then.

Milan Svolik:

Yes, asking in a survey respondents to choose between two candidates is one thing, and the advantage of that setting is that you are able to manipulate a lot of the features of the candidates, including their undemocratic positions. We wanted to be sure that also, in the real world, we would be able to arrive at a similar conclusion. The problem with the real world is that we almost never get to observe two candidates competing against each other and both being perfectly democratic and then the same exact election but one of them being not democratic. Almost that happened in Montana. In Montana, there was a special election for a House seat vacated by Ryan Zinke, who joined the Trump cabinet. This election happened in 2017. It was between the Republican, Greg Gianforte, and then a Democrat.

Crucially, what happened in that election is that Montanans got to observe two elections. Those who voted absentee before election day, on the night before the election, Gianforte assaulted a journalist. I believe the technical term was body-slammed the journalist. All the major newspapers, the three major newspapers in Montana, initially endorsed Gianforte. After this happened, they issued editorials on the morning of the election denouncing him and dis-endorsing him. So election day voters in Montana saw two candidates, a Republican and a Democrat, but now the Republican attacked a journalist, and so we are able to compare those who voted on election day to those who voted before election day and try to see whether that extra key information, that Gianforte actually is willing to attack a journalist, changed people's minds and changed how they voted.

William Howell:

Yeah. So, just in your description of the events, you point out that the journalists were quick to change their mind. There, we see immediate evidence of people making precisely the kinds of trade-offs that you were describing before. But we don't see, in the main, that kind of evidence in terms of voting behavior. It's a tricky kind of comparison, of course, because people who vote absentee are systematically different from people who vote on election day, so they can't just compare the mean levels support before to the mean levels of support on election day and expect to be able to make clear claims about what the impact of that body slam was on not just the poor journalist but on the thinking of voters. So how do you deal with this?

Milan Svolik:

That is right. That's an important objection. What do we do is that we develop the so-called differences-in-differences approach. That is, we, rather than taking just the precinct-level vote share among absentee and in-person voters in that 2017 election, we take the difference and compare it to the vote in the 2016 election. So now we have differenced out, here, any stable differences between in-person and absentee voters. In effect, we are comparing two differences, the shift between 2016 and 2017 among absentee voters and the same shift among in-person voters.

If the Republican receives fewer votes, that shift is less favorable for the Republican among in-person voters, then we can be more confident that that change is not because of the difference between absentee and in-person voters but rather because, relative to the same benchmark, now the in-person voters are actually voting less Republican. What we see in this case for the sample of precincts that we're able to examine is that, in fact, about 3.5% of the voters are now voting less in favor of Gianforte than they would be if they voted absentee. This shift is primarily concentrated in moderate counties and moderate precincts. In extreme precincts, there's either no different or, in fact, in a very few, it appears that they are actually rewarding Gianforte.

William Howell:

Oh, okay. So then we're in this moment, and we see all that's happening in the aftermath of this election, and we see both 74 million people voting for Trump on election day and all of the antidemocratic, I think what many people would claim would be antidemocratic, appeals by Trump in the aftermath of this election and the willingness of all kinds of people to abide those claims and affirm those kinds of claims. How does this paper of yours speak to this moment?

Milan Svolik:

One way that we can take our study is really to get a realistic estimate of what the price is that voters, and, in this case, Americans, are willing to pay for democracy. I think it's a sobering account compared to the benchmark of surveying people and asking them questions like, "How important is it for you to live in a country that's governed democratically?" Here, we have an answer where the end action is exactly what we care about, the choice between two candidates. One thing that I think is very sobering about this is that this is the United States. This is one of the oldest democracies in the world. So when only three-point-something percent of Americans are willing to change their voting behavior in response to undemocratic behavior by a candidate they like, I think we need to also reassess how we think about voters elsewhere in the world who are also often in a similar position but who, in these classic surveys, answer that they are highly committed to democracy. Our study suggests that Americans do value democracy, but that support is highly elastic, and it's highly susceptible to especially partisan polarization.

William Howell:

Okay. So if I could just take us from bad to worse and see if you'll come along with me, what you don't talk about in this paper is why an elected official, why a candidate would behave undemocratically. You simply identify the cost associated with that. But, of course, there are benefits from behaving that way. Some of those benefits have to do with having just an appetite for power or wanting to advance a policy agenda. Another thing, though, the instrumental value of behaving undemocratically for an incumbent is that it's a way to improve his or her electoral chances in the next election. It's a way to limit the franchise or to suppress the vote or to make the chances of their removal from office that much more distant.

What you have here is simply we're going to fix the electorate and treat it as constant and show that not that many people are going to step up and offer a corrective. That, though, in combination with the recognition that the antidemocratic behavior itself has the effect of marginalizing potentially those who might vote against you or who represent a threat to your chances of holding onto power, make the story that you're telling now all the more harrowing.

Milan Svolik:

I agree. So we, on purpose, in fact, focused on the kind of violations of democratic principles when we were assigning these statements to candidates that often occur in the United States, things like gerrymandering, voter disenfranchisement, the candidate suggesting that the governor ignore court rulings by judges appointed by the other side, and we document that this indeed happens, especially at the state level.

Consistent with what you said, I think the most realistic interpretation of why it happens is not because there are erratic politicians throughout the United States, but it's because they understand that there are political benefit to some of these measures that are undemocratic, and because they are in a democracy, however, those measures have to be very gradual, and they have to be such that voters don't react too strongly. I guess one way to interpret our results is that you can succeed in that and you can succeed in that especially in places where your party is at an advantage because then you know that even if there is some backlash, it might not be large enough to cost you the control of the legislature or governorship while simultaneously you might actually advance your position.

Anthony Fowler:

Okay. I mean, I think this is pretty clever. There are lots of experiments like this out there, these kinds of conjoined experiments. I'm not aware of anybody who has specifically tried to get as this question, and this seems like a nice way to get at this question. The finding is that on average these things do decrease the candidate's support. I think on average the candidate performs something like 11 percentage points worse, and then when you restrict attention to when it's just Democrats and Republicans with disparate policy positions, it's something like three or four percentage points.

So these are negative effects. They're not huge negative effects. Three or four percentage points is certainly enough to tip the results of lots of important elections, but it's not even as large as, say, what we typically think of as the incumbency advantage or the effect of a scandal or even, in their own experiment, they have the effect of an extramarital affair, for example. So their interpretation is that that number is relatively small and voters don't care that much about these democratic principles when there are other more important things at stake to the voters, like are they getting their preferred school funding package or something like that.

William Howell:

And I think it's worth underscoring that last point, which is the starting point of the paper for them is to say, look, when you survey the public and you say, "Do you care about democracy," everybody says, "Yes, hooray, we love democracy, I care about democracy," but when they're presented with trade-offs in the final analysis, they want to come out and say, "Yeah, but not so much." I mean, yes, at the margin, you might care about democracy, but you care about a whole bunch of other stuff a whole lot more, so we can't count upon the American public to step up and force politicians' hands and ensure that they behave within democratic guardrails.

Wioletta Dziuda:

I think my reading of the paper was less pessimistic than the reading of the authors themselves. Let's talk about how large we think those numbers really are. Are they really small, or are they actually pretty large? One thing that came to my mind when I was reading is that those politicians were running for a state legislature, and then if you think the undemocratic positions that they were taking were, for example, regarding the behavior of the governor. So, as a voter, when I'm deciding whether to elect a state legislator, I think, "How likely is it that the views that she expresses are going to somehow affect the outcome?" So if I think about her views on what governors should do, it's very unlikely that her views are going to affect what the governor will actually do. But her views on immigration or her views on funding state schools, those are actually likely to somehow affect legislation. So it's not so surprising that I'm going to cringe when I hear all these undemocratic views but that I'm going to ignore them when voting.

Anthony Fowler:

Yeah, I agree with that 100%. I think that was one of my first thoughts reading about the experiment as well, is, well, for how many of these things can a state legislator even make a difference? Whether or not the governor complies with unfavorable rulings by judges they don't like, it's not really something that the state ... It'd be nice if everyone in the state legislature agreed that there are certain democratic norms that the government should abide by, but there's not a whole lot a single state legislator can do to change, for example, whether the governor uses executive orders. Certainly, a state legislator cannot change the number of polling stations without the cooperation of lots of other officials.

Wioletta Dziuda:

Yeah, I think I completely I agree. I think there are additional complicating factors in this survey. So, for example, if you think about gerrymandering, I am opposed. Gerrymandering is legal in the United States. It's a crazy idea to me. All else equal, I would rather have politicians engaging in that. But I know that they all do, and knowing that the opposite party, the party that I don't support, actually actively engages in gerrymandering makes me, to be honest, want my party to engage in that, too, just to level the playing field, just to make everything fair. So things that I consider very undemocratic I might still support or not oppose, not actively oppose, because of the equilibrium we are in, because of the situation that we are in.

William Howell:

I think this is important. I quite agree that part of what I suspect voters are hearing when they're presented with these kinds of what are being called undemocratic positions is a sense of both leveling the playing field in the way that you've described, Wiola, but also a willingness on the part of elected officials to fight hard and to do everything possible in order to advance good policy in an equilibrium in which affecting change is nearly impossible. So supporting a government issuing executive orders when legislators don't cooperate can be seen as, look, we've got to do something. There are problems here. This is another avenue by which we can affect change. I'm not just going to say, "Ho-hum, what are we to do. Change is hard." I'm going to fight hard on behalf of the interests of my constituents, and to that extent—

I mean, it's tricky because that, too, can become irrational, a rationale for what is decidedly antidemocratic behavior. So there is that piece. But there also is what may be communicated here is just a willingness, again, to fight hard and do everything possible in order to best represent the interests of your constituents. This is not a story about trying to corrupt a democracy through and through and co-op the state for one's private interests necessarily.

Anthony Fowler:

So there's even more that complicates this, I think, which is let's not forget that the estimated effects here are negative, right? The voters on average are penalizing politicians who take these positions, and they need not be negative. If you think through some of these things, if you had just paid attention to political discourse on cable news, you might've thought these estimates would be positive because for many people, if you're a partisan extremist, this might be something you really like. Say I'm an extreme conservative Republican and I should actively want the legislatures to say, "Yes, the Republican governor should do whatever he wants and should ignore journalists and ignore judges and so on." Yet, nevertheless, we're still getting these negative effects.

One thing that I find interesting about a lot of these undemocratic positions is that they already are bundling together some trade-off. There's a trade-off already in here, right? I might, of course, all else equal, not want a governor to just ignore unfavorable court rulings. That seems bad. I care about democracy and so on. But, at the same time, I might want my own party's governor to implement whatever policies they want because I tend to agree with that party on policy. So there's a trade-off built into that treatment already, and if I cared even more about achieving a policy goal than I cared about democratic norms and institutions, then a lot of these things I might expect a positive effect.

The fact that I get not a positive effect but, in fact, a negative effect on average suggests that, in many cases, these respondents care even more about democratic norms than they care about achieving some short-term policy goal. So I think there's a way to read that result as saying, actually, these respondents care a lot about democratic norms. Not only do they care about democratic norms a little bit, they actually care about them even more than they care about achieving some immediate policy goal.

William Howell:

Yeah. A core part of the paper is to suggest that the effects aren't uniform, that the effects are largest among centrists, that's where the willingness to punish antidemocratic behavior appears to be most pronounced, but that, as you say, even the extremists look dimly upon this kind of behavior when there's a long tradition. So I study the presidency. Henry Jones Ford, who was a progressive writing a century ago, talked about how there's a kind of political valor for presidents who break through the constitutional form.

It's this notion that there are all these impediments, procedural and small democratic impediments, that stand before you but that what you're going to do is rise above them in order to best serve the interest of the people, come what may, and that what political failure looks like is somebody who sits on their hands and says, "Yes, I would like to do something on behalf of you dear people, but what am I to do because there are procedures that we have to respect and norms we have to abide," and that that kind of posture in our politics is to be not just frowned upon but to be repudiated. That's the sign of political failure, and yet here's a paper that says, no, even in this moment of crisis, even in this moment where we're tied up in gridlock, where we've got widespread polarization between the parties, we nonetheless see a public willing to step forward and to say, "Nope, when you engage in these kinds of behaviors at the margin, I will be less likely to vote in favor of you."

Anthony Fowler:

So if we wanted to square the results of this paper with ... It certainly seems like there's a mass of people out there who celebrate these undemocratic developments in some cases. How would we do that? I could think of a few different possibilities. One is that the people that we see on cable news are in fact a tiny, tiny, tiny sliver of the American populate. They're negligible, and none of them were in this study, and, essentially, none of them are in America. They just happen to get a lot of news attention. Another possibility is that there very much are these heterogeneous effects. If you are extreme enough and partisan enough and so on, then, yes, there are some people for whom these effects are positive. This experiment just happens to not have maybe enough data or the right kind of data to be able to detect that heterogeneity.

William Howell:

But they do find this in the Montana experiment. He does find some evidence that people from the most extreme districts appear to have rewarded, well, the candidate who body-slammed the legislator the night before election day and that could be seen as, "You see, here's somebody" ... It could be an affirmation of how much they hate the press, right, that they are truly antidemocratic. Another way of understanding that, though, is that, no, this is somebody who's really going to fight, who really is going to stand up and do the right thing, come what may, and is not going to put up with what's perceived as being anti-American or purely obstructionist behavior and that that's what we want in the people we elect to office, not people who body-slam other people, per se, but somebody who's really going to fight hard.

Wioletta Dziuda:

Yeah. I think this brings us to the distinction between what's truly undemocratic and people believe is undemocratic. So if I think about what's happening right now in politics, a lot of people on the right say, "We think that actually the election was stolen, and we cannot fathom why Democrats are not trying to defend democracy," or they say, "Look, our voices have not been heard. How can this be democracy? For so many years, we've been abandoned, and now, for the first time, we have a president that actually seems to be channeling the will of the people." So I think maybe that's one way to try to square the results of the paper with what we see outside, that, in the paper, people clearly think about those issues in isolation, not in a particular context, and they see that those positions are undemocratic and then they'd express their position, "We do not like undemocratic things." But in real life it's a little bit of a grayer area of what I think is democratic, and what I think is democratic for me is not democratic for you.

William Howell:

I quite agree. This is, in some ways, the thing that I have the hardest time with the paper, in that the paper's suggesting, and usefully so, that what we ought to be doing is paying attention to trade-offs. On the one hand, I like the policy. On the other hand, I don't like the antidemocratic behavior. Which effect dominates? But I think much of what we observe is our understanding of what constitutes democratic behavior is itself a function of where we sit in our politics. So when we look out right now and we see Trump railing against this election, you see people on the left saying, "This is a coup," and, meanwhile, on the right, they're not saying, "Well, it is a coup, you're right. This is antidemocratic. But Trump's the right guy because has the right policies, and so I'm willing to stand with him." That's not their understanding. Their understanding is that, "No, there really was widespread fraud. There really is a sense that our voices have been systematically ignored and there's an election that's been stolen, and I can't believe that you people on the left aren't willing to own up to that and to face it."

Now, in between all this is the status of evidence and the fact that you can't change certain people's mind by presenting evidence that suggests there was in fact no fraud. Well, then, how seriously are we going to take you when you say, "I, too, stand for democracy?" Yet, at least as they express their views and their willingness to support Trump, it's not that, well, he has the right policies but I'm troubled by his antidemocratic behavior, it's that no, no, no, he's the one who's the true voice of the people all the way through and our system really is rigged and we need to provide a corrective to our broken politics.

Anthony Fowler:

But that could be precisely where what you see on cable news is not representative of what's really going on in the electorate, right? I'm sure there are, in fact, many Trump supporters who hold exactly the position you said that they don't hold, which is, "I agree with Trump on immigration and taxes, et cetera. Of course, I don't love all this other stuff, but I still voted for him nevertheless." That must be a common view. But that's not the person you see screaming on cable news. That's the person who you see quietly voting for Donald Trump and not talking to their friends about it.

William Howell:

True, but I think there's also evidence, though, that many people who fit the characteristics that you've just offered believe that, in fact, the election was stolen, believe that, in fact, it was widespread fraud. It isn't that they say, "I'm really troubled by Trump's misbehavior here, but he will deliver us continued tax cuts." It's that they think he's fighting the right fight. I think there's lots of people who believe that, and they're not just the crazy people who are screaming on cable news, but a lot of people believe that.

Wioletta Dziuda:

Yeah. So I think it's actually very important, it will be very important to know the difference, how many people belong to the first camp and how many people belong to the second camp. Part of me agrees with Anthony that probably there are lots of people who just say those things because they want to feel better about themselves, but, truly, they know that the election was not stolen and they support Trump only because of the policies. But I also agree with Will because, look, people, during the primaries in 2015, they had the choice between Trump, who already was showing some undemocratic tendencies, and other conservatives with very similar policy positions who were more democratic, who were specifically saying, "What Trump is saying is wrong and we have to preserve democracy."

Fast forward to 2021. You have Republicans who hold similar positions who are running for Congress, who are running for Congress, who are in Congress or planning to run for Congress, and some of them actively decide to support what we would view non-democratic positions because they think they would be rewarded for those positions. So it seems that at least in ... Maybe they also respond to what's on cable news, but at least in the mind of Ted Cruz, taking an antidemocratic position, what we call an antidemocratic position, is going to get him rewarded, even though he doesn't change his position on any substantive issues.

Anthony Fowler:

But we don't have much evidence that that's true, right? Even Ted Cruz doesn't have much evidence that that's true. If you just looked observationally, you would see that, as we discussed in a previous episode, Trump notably outperformed other Republican candidates running in the same places in 2020, which seemed to suggest that voters are actually penalizing Trump for ... Maybe he matches their policy views, but he's got other shortcomings, including his undemocratic positions, that might have caught the votes.

Wioletta Dziuda:

So I agree. This remains to be seen, and I'm actually sitting at the edge of my seat. I hope you're right. I hope that Ted Cruz and all these Congress people are mistaken, that, indeed, them taking those undemocratic positions are going to backfire. But they seem to believe that that's something that voters actually want, which would suggest that voters actually perceive those actions to be fine, to be perhaps non-undemocratic.

William Howell:

Yeah, this is something that ... Milan doesn't offer an account of this in the paper, which is to the extent that their effects are negative in the paper, and so why would anybody then behave antidemocratically? Why would you do this if it could on average only have the effect of decreasing people's willingness to support you, and particularly among marginal voters, particularly among centrists? What are you doing?

Wioletta Dziuda:

Especially, why would you take those positions when you know that they are not going to change any outcomes? This is a very good example of what's happening right now in Congress. They all know that Biden will be elected president come January 20th, and they still say, "I'm going to take this undemocratic position." So, yes, that's even more puzzling in light of the results.

Anthony Fowler:

I have three explanations in mind. This parallels the debate we have sometimes about why elected officials are so extreme, even though all the evidence suggests that they would be electorally better off if they could moderate their positions. One potential explanation is they just can't help themselves. They really are extremists who don't care about democratic principles. They just can't help it. This is their true position, and it's going to come out one way or another over the course of the campaign.

Another possibility is they have the wrong perception about the electoral returns to these behaviors. Maybe all of the people they talk to are their own staffers, who are extremists, and the people that come to their rallies, who are extremists, and they tune into cable news and so on, and they think that there are lots of people who are going to reward them for this when, in fact, there aren't.

Then maybe a third possibility that also is somewhat consistent with the data is that this is kind of a thing you do when you're behind. You throw a Hail Mary. You gamble for resurrection. You don't do this normally when you're ahead, but when you're Donald Trump and you see that you're about to lose the election, that's when you start to try to undermine the health of our elections. You don't do that when you expect to win.

William Howell:

Can I offer a fourth, which is a slight variant on that third one, which is that there are rewards to be had from behaving antidemocratically, both in terms of you and members of your party maintaining power going forward and you and your members of your party's ability to advance an agenda, for which there are decided electoral rewards to be had. So, look, if, again, what I do is wrap myself in constitutional niceties and say, "Yes, I would like to advance this policy, but I can't because the courts have said I can't," a set of people in the electorate are going to say, "What are you doing," whereas if I overcome those obstacles and I actually deliver for them, I actually pass the tax cut, there are electoral rewards to be had there. In that sense, there's a political rationale for behaving antidemocratically. It's about your ability to deliver for your base and secure the rewards associated with doing so and shifting the terms of subsequent political contests to your advantage, which also is beneficial.

Wioletta Dziuda:

Yes, but, still, we see those undemocratic statements and movements, even accompanied by a statement, "I know this is not going to have any effect." So what you are talking about, what both of you are talking about, is a situation in which I have a chance to actually, yes, do something antidemocratic but this is actually going to lead to my policy position being implemented. But, here, people, the very people who engage in those antidemocratic positions, they acknowledge, "There's nothing we can do. It's all going to end up the same. Biden is going to be the president, and we're going to have all these liberal policies implemented. But I still am going to do those things that can damage our image abroad, that can damage our ability to impose democracies on other countries and talk about imposing democracies in other countries." So, again, I hope Anthony might be right, that they just are wrong that these positions are popular, but they seem to behave as if they believe that these antidemocratic positions were popular with certain constituency.

Anthony Fowler:

Yeah, and it could be that there's all kinds of internal party politics that explains this kind of stuff as well, that they're not trying to please the voters, they're trying to please Donald Trump and they're trying to please their state party captains and so on, who might be 100% onboard with these positions. But I don't think we have any compelling evidence that this is actually what the voters want, and I think this paper's showing at least somewhat compelling evidence that they don't. I take that as reassuring for the future of our democracy. So should we talk briefly about Montana? The last little analysis in this paper focuses on this case, a 2017 special election in Montana's US House seat. Greg Gianforte is a Republican. The day before the election or very near, right before the election—

William Howell:

The day before.

Anthony Fowler:

... he body-slams, he assaults, physically assaults, a journalist, who by all accounts is just trying to ask him some questions about his positions on healthcare. They've got a clever idea where they're going to compare his support in the early votes where those people didn't know about the body slam to the people who voted in person on election day, and then they're going to do a difference-in-differences where they essentially compare early to in-person votes in previous elections to try to account for the possibility that the early voters are just different. We could probably critique the design a little bit if we wanted to, but it's an interesting idea, and they find that, indeed, if you didn't know this already, physically assaulting a reporter right before the election is bad for your electoral chances. So that's good. Again, it goes in the-

William Howell:

Again, except among the most extreme, right?

Anthony Fowler:

It goes in the expected direction. That's good. Yes. What do we make of this? What do we make of this result, other than the fact that doing crazy things is bad for you electorally? The authors want to say this is yet another example of ... This is an antidemocratic thing because he's assaulting a journalist for doing his job, and he's now the governor of Montana—

Wioletta Dziuda:

He still got elected, yes.

Anthony Fowler:

... so this seemed to propel him to the upper echelons of politics in the state.

William Howell:

Yeah, it does two things for them. One is it underscores their basic theme that the public can't be counted on to rise up en masse and vote people who assume antidemocratic positions out of office. It does that, and the other thing that it does is to say, look, the survey experiment is interesting and clever, but it's still a survey experiment, and they want to look at actual behavior, actual votes cast out in the real world, and there aren't very many instances where you have something that might be construed as a violation of democratic norms occurring randomly that would allow us to back out its effect on votes that are ultimately cast.

Now, we could not just quibble but raise real concerns about the extent to which this difference-in-difference design works, right, that the difference that we observe in a presidential election between early voters versus same day voters need be the same as the kinds of differences we observe on off-year elections. But then there is this larger issue about the extent to which we want to say is the learning having to do with a violation of democratic norms, or is the learning, "My God, this guy's crazy, he's unhinged," or, "Wow, he's a tough dude and I really like that." Does it have anything to do with the broad themes of the paper? People can come to their own views.

Anthony Fowler:

If you were going to write a model, if you were going to write a formal model of when to body-slam your political opponents, you would say, do it, one, right before the election, when most people have already voted.

Wioletta Dziuda:

Yes, if you truly body-slammed—

Anthony Fowler:

If you really want to body-slam.

Wioletta Dziuda:

... like you were provoked.

Anthony Fowler:

And two, if you really want to body-slam a journalist, you should do it in Montana where the electorate is already pretty skewed in a partisan way so that it'll cost you votes but not enough for you to lose the election.

Wioletta Dziuda:

True. That's good to be.

Anthony Fowler:

So maybe we should take some solace in the fact that the reason we don't see regular body-slamming in Ohio and New Hampshire and North Carolina is because that would be electorally consequential for those candidates, so they're not doing it.

Wioletta Dziuda:

Yet.

Anthony Fowler:

Yet.

Wioletta Dziuda:

Welcome, 2021.

Anthony Fowler:

I'm sure immediately afterward—

Wioletta Dziuda:

You remember the things we were telling ourselves a year ago that will never happen, and then they all happened, so who knows what will happen this year.

Anthony Fowler:

In light of recent events, in light of this insurrection in Washington and the new Georgia Senate results coming in, we got together again to just talk briefly about these same topics and think about the extent to which we might've changed any of our views in light of this new information.

Wioletta Dziuda:

I think the question still remains, to what extent they're most representative of a broader audience. Is this just a small fraction of the most extreme and vocal Trump supporters, or is it actually an opinion of an average Republican voter? During our podcast, we were talking about what's actually true. It's worth remembering that those events that happened in DC last week came just right after two Democrats won seats in the Senate. The two Democrats from Georgia won seats in the Senate despite the fact that Trump candidates were running on the other side, and they were aligning themselves very specifically with undemocratic actions. I think if we learn something from Georgia, it's that people are actually attaching quite a bit of weight to undemocratic actions and they are willing to switch their votes and vote for a less favored candidate otherwise policy-wise just to express their disappointment with undemocratic actions. I think that explains also why a lot of Republicans decided to abandon Trump just right after the riots in DC, because they realized that there is a lot of opposition among voters against those undemocratic attempts.

Anthony Fowler:

Yeah. I agree with everything that's been said. Obviously, the events this past week are extremely upsetting, yet I come away thinking ... One, I don't think this mob speaks for a large group of Americans. I think there probably have always been crazy people who would be happy to march on Washington and try to storm the Capitol if given the opportunity, and maybe Donald Trump was the coordinating device for them, like, "The president told us to do it. Okay, great, now we can really fulfill our dreams." That's obviously disappointing and frustrating that that's the case. We unfortunately learned about the perhaps incompetence of our security and our Capitol police and our organization in order to prevent something like that from happening, so that was very troubling.

But on the other hand, I think we learned something about the resilience of our democratic institutions. Even Mike Pence and even Mitch McConnell, who would probably love to overturn the results of the 2020 election results, they didn't, and they weren't willing to do that, and they very strongly renounced the protestors, as they should have. But maybe it wasn't obvious to all of us that they would, and they did. The vast majority of the American public seems to realize that this was a terrible thing that happened, and they reaffirmed their support for a democratic institution. So, in some sense, I think even in this perfect storm of things going as badly as they could for the health of democracy we still weathered that storm well. That gives me some confidence in the health of our democracy going forward.

William Howell:

Can't we at once be reassured and shaken? I think this is one of those cases where it is true that both chambers of Congress went back into session in the immediate aftermath of the insurrection and completed their work and thereby sent a powerful signal about the willingness and the capacity of the first branch of government to carry forward and in fact certify that the Biden-Harris ticket is going to take over the second branch of government come January 20th. That's a very big deal. It also is unnerving, to say the least, that this band was able to take over Congress and they came within a hair's breadth of actually entering the Senate chambers when the senators were actually in there and some of them had zip-ties, which suggested that they had plans to do more than just smash windows.

When we think about the preparations for this attack and the president's involvement in it and the different responses of at least some Republicans who saw perfectly fit to carry forward and reaffirm the lies that the election was somehow rigged or stolen, even in the aftermath of this insurrection, that's ... My God, right? You think, "Where are we?" Yet we were able to withstand that, which is why I say, I think, I can't help but feel at once shaken and reassured all at once.

Wioletta Dziuda:

Yeah. I think if I think about what happened, I'm much more pessimistic and much more disappointed with the behavior of the political class than with the American voter. I think the political class thought, inferred from the election of Donald Trump, that most of the Americans are waiting to compromise the democracy. They didn't think twice, and they went along with that. They themselves are willing to compromise democracy just to be reelected, just to be in power. But I think the lesson that hopefully they have learned from the Georgia election is that they were wrong about the American voter, that, at least, the people, the voter, is the one who cares about democracy to quite an important extent. So I'm more hopeful about the American public but extremely disappointed in the American politicians, and that definitely raises a question moving forward to if we get another demagogue elected, will they, again, assume that voters just like demagogues and they don't care about democracy and will they play along? This time, we were safe, but next time we might not be safe. They might push us too hard before we have the next election where the voter can express their opinion.

Anthony Fowler:

And to be clear, in Georgia, it wasn't a wholesale repudiation of the Republican candidates that we saw there. They, in the general election, barely won, but they didn't get the majorities needed in order to avoid a runoff. But they barely won, I guess, both their candidates, and then it was a swing, you can count it in the tens of thousands, slightly more than 100,000 votes, out of over five million votes cast, in the Democrats' favor. So it isn't that the Republicans all packed up their bags and said, "We won't put up with these lies en masse." But there were enough, given just how close the election was, to make a difference there.

Wioletta Dziuda:

Yeah, there were not, and as we talked about in the podcast, this was just an election of two senators, and as a voter, you might not like their antidemocratic attitude but you might say, "Well, they are not going to actually be pivotal in dismantling democracy," so you might be willing to play along with that, and the voters, at least on the margin, they weren't. But having said that, again, next time, it might be that this marginal voter won't be even able to express their opinion because by the time the election should roll around, politicians might have already changed the rules to the point at which we can't express our votes properly.

William Howell:

It is remarkable. It's remarkable that both Senate seats from Georgia are going to be blue. We're talking about a very solidly red state that has historically in recent history not even been competitive, not even been the kind of state we talk about politically. We're also talking about a special election, where typically turnout is low in special elections, and especially young turnout, minority turnout is low in special elections, so the runoff election, right? It's especially shocking that in a solidly red state in a special runoff election, you would still get a blue candidate emerging as victorious. We've also talked about the fact that voters typically prefer a divided government, or at least a lot of voters do, so, if anything, you might've thought now that we know that Biden is going to be the president, maybe that frees up a lot of conservative-leaning or moderate voters in Georgia to more confidently vote for a Republican for Senate.

All of those things are pushing against this result, and nevertheless you get two Democrats elected, albeit in a very close election. So that suggests that there's something else going on that the voters are upset about. It's not that Georgia all of a sudden became hyper progressive in terms of their preferences on taxes or immigration or something like that. They probably are genuinely upset about some of these behaviors of the Republican Party and these Republican candidates.

Anthony Fowler:

Yes. I think that people will be pointing ... Those who would and will criticize the president for many things will, on the right, point to his losing the Senate by virtue of these lies that he was trafficking in the aftermath of the election as being really counterproductive towards Republicans' ability to hold onto the Senate. We talked in our ... When we were discussing the paper, we talked at some length about whether or not these effects are big or small, and this is a case where the effect on the Georgia electorate was small, in the sense that the number of people who switched votes was a small number of people.

William Howell:

We don't really know. We don't know. I mean, in the runoff, as the effect of this antidemocratic stuff.

Anthony Fowler:

Yeah, I don't want to suggest that, but it surely is the case that those who voted in ... The vast, vast majority of people who voted in both the general and then the runoff voted consistently Democratic or consistently Republican, right? To the extent that there are differences, some people who voted are who changed their mind and then voted D, it's a reasonably small number. But to the extent that we want to attribute that in turn to Trump and his antidemocratic, small democratic entreaties, that the small number of shifts induced a massive political shift, right? It led to both races switching and the Democrats getting control of the Senate. So little effects can have huge political repercussions.

William Howell:

Thanks for listening to Not Another Politics podcast.

Wioletta Dziuda:

Our show is a podcast from the Harris School of Public Policy and is produced by Matt Hodapp. Thanks for listening.