In 2016, Donald Trump dominated media coverage in the race for the Republican nomination and he is on track to do so again this time. Does the media react to events and signals of public support, moving from one candidate to the next, or does it just focus on the frontrunner? And is media attention the main moving part in presidential primary campaigns? Zachary Scott finds that the media only sequentially highlights candidates in some nomination contests. But Trump dominated coverage more than others, in part due to his fearful and personalized rhetoric. Kevin Reuning finds that public interest follows rather than brings media coverage. Media attention led to increased poll support for Trump in 2016, but not for the other candidates. At least in 2016, the conventional story that Trump garnered outsize coverage and benefited seems correct.

Guests: Zachary ScottKevin Reuning

Studies: “Replicating the Discovery-Scrutiny-Decline Model of Quantity of Media Coverage in Presidential Primaries“; “Media Coverage, Public Interest, and Support in the 2016 Republican Invisible Primary.


Matt Grossmann: Does the media drive presidential primaries? This week on The Science of Politics. For The Niskanen Center, I’m Matt Grossmann.

In 2016, Donald Trump dominated media coverage in the race for the Republican nomination, and he’s on track to do so again this time. Does the media react to events and signals of public support, moving from one candidate to the next, or does it just focus on the front-runner? And is media attention the main moving part in presidential primary campaigns?

This week, I talked to Zachary Scott about his Journal of Elections, Public Opinion and Parties article, Replicating the Discovery, Scrutiny, and Decline Model of Quantity of Media Coverage in Presidential Primaries. He finds that the media only sequentially highlights candidates in some nomination contests. In others, they mostly stick with the front-runners, but Trump dominated coverage much more than others, in part due to his fearful and personal rhetoric.

I also talked to Kevin Reuning about his perspective on politics article with Nick Dietrich: Media Coverage, Public Interest, and Support in the 2016 Republican Invisible Primary. He finds that public interest follows, rather than brings, media coverage. Media attention led to increased poll support for Trump, but not for the other candidates. At least in 2016, the conventional story that Trump garnered outsized coverage and benefited seems correct. I talked to both about what’s in store for the 2024 cycle.

Let’s start with Scott, reviewing models of presidential campaign media coverage. All right. So what is the discovery, scrutiny, and decline model of the presidential primary process?

Zachary Scott: So the discovery, scrutiny, and decline model is a theoretical model of media coverage, specifically the quantity of media coverage in presidential primaries, that comes out of John Sides and Lynn Vavreck’s 2013 book, The Gamble, which is covering the 2012 presidential election both through the Republican primary, and also the general election between Mitt Romney and Barack Obama. And I believe they’ve also brought it up in their two subsequent books, Identity Crisis and Bitter End.

The basic idea of the discovery, scrutiny, and decline model is that often, there will be a perceived front-runner or favorite in a primary campaign, not necessarily always, but often. But then who is seen as kind of the main challenger to that candidate, that can wax and wane throughout the cycle, often as a function of media coverage, which tends to flow in this cyclical pattern.

So first, one of the candidates, among the many who are competing will be, quote-unquote, discovered. This will then precipitate a sharp increase in media coverage. So the media will land on, “Oh, this challenger. Wow. Look, maybe they could really contend for this nomination.” They will see a huge surge in coverage. As that goes on, the model suggests that this coverage will then start to turn more negative, that because the candidate is receiving so much more focus, more of that focus will be on the negative things about the candidate. Things will be discovered that are not good. This will then lead to a decline. I believe the model suggests it will probably come in support, followed by media coverage, and then the process will repeat.

In practice, often even within the 2012 campaign, there wasn’t necessarily scrutiny per se. It was more just the media discovered a shiny new thing, and so it just moved on to a different candidate, even though they hadn’t necessarily landed on something negative about the previous candidate. So it kind of fits. It can work in a few different ways.

Just to give practical examples, for people who remember the 2012 campaign, Mitt Romney is receiving a pretty steady amount of coverage throughout. Early on, Rick Perry is a potential contender, and then he gets a huge amount of coverage, and then he has some air-headed moments in some debates because he recently had surgery and was on painkillers, and also because I believe the media discovered that his family had hunted at a hunting lodge that was named about the racial epitaph. So that fits that model, right? A lot of coverage. Suddenly, that coverage becomes more negative, decline. Others, like Rick Santorum, never, at least to my mind, received all that much negative coverage. They just kind of ran out of time, and then their coverage then declined because they simply could no longer viably compete for the election.

So that is the basic pattern that you will see. This cyclical repetition of increases of coverage for some candidates that declines as they’re replaced with others. It’s not necessarily going to happen for every candidate in the race. Jon Huntsman never sees this happen, but you will see it for at least a few candidates every election cycle is the suggestion of the theoretical model.

Matt Grossmann: So you’ve studied the applicability of this model across Democrats and Republicans since 2000, finding it sometimes works and sometimes doesn’t. So what were the big findings and takeaways?

Zachary Scott: Yeah. This may be the simplest quantitative model, or paper, I’ve ever written. Probably will ever write. All I really did was plot some lines and look at them. The basic test I wanted to do, take a step back. When I read The Gamble, I was really smitten with this model. Because the 2012 Republican primary was the first one that I really followed that closely, because I was in college, I was taking a lot of political science and a lot of journalism classes. It was just something I was paying attention to. And I remember taking a class with the president of our college chapter of the College Republicans, who just every four weeks came in with a different t-shirt for a new candidate, whoever was surging The beginning of that fall semester, he was a Rick Perry guy, and it just kept repeating. He was a Herman Cain guy for three weeks.

And so when I read this model, I said, “Yeah. That makes a ton of sense,” that this was a person who was obviously very politically attuned, probably a little skeptical of Mitt Romney for whatever reason, could be ideology, could just think he wasn’t going to win a general election, whatever it was, wanted someone else beside him as an alternative, and kind of floated along with the media coverage as they repeated this process. I thought it was an excellent description of what happened in that election cycle.

As I went through grad school, I kind of can think of it as we would statistically refer to as overfitted, right? That’s a book that’s focused on the 2012 election campaign specifically, it’s a theoretical model to describe that election. And if you target a model to describe one thing, you tend to lose your ability to explain other things.

And as I saw the 2016 campaign unfolding, I didn’t think it really matched all that well with what was going on in either the Democratic or the Republican sides. The Democrat side, there just weren’t all that many candidates. There were only ever five really that ever made it to the debate stage. Two of them dropped out after a couple months. Martin O’Malley never got a surge. It was just two. There’s not a repetitive cycle. The Republican side, Donald Trump sucked all the oxygen out of the room. No one else really had all that much surge and decline relative to where he was at.

So I wanted to see, okay, well does this actually kind of broadly apply? So I just took the media coverage from all of these elections from 2000 to 2020, ultimately. That gives us nine presidential primary campaigns, although we kind of have to throw one of them out, because in the 2000 Democratic race, there were only two candidates. So again, you can’t even have a surge and decline when there’s only two.

But this still gives us eight other races. And I just looked at the share of coverage that candidates tended to receive over, and I operationalized the surge as did you receive a plurality of coverage? So whoever arose to be the most-covered candidate at a certain period of time. Although ultimately, I think I was even more charitable than the model in just kind of looking, did it look like you got close? And when you do that, I think that the discovery, scrutiny, and decline model, it does a good job of explaining the media coverage, the patterns in quantity and media coverage, in about half of those eight primaries where it’s kind of eligible to happen.

2012 Republican primary is the best case. You see it really clearly with Santorum, you see it really clearly with Rick Perry, with Herman Cain. Newt Gingrich is kind of weird, because he surges three different times, but still, broadly works. Again, doesn’t describe everyone. There’s not really a moment for Michele Bachmann, except maybe something around the Ames Straw Poll, not Jon Huntsman, but broadly, yeah. Works. The 2008 Republican primary is another one where I think it broadly worked with Romney, with Giuliani, with Fred Thompson, and with John McCain, the eventual nominee.

And then the Democratic side, it works pretty well in 2004, Kerry, Dean, Wesley Clark, Lieberman, Edwards. And in the 2020 Democratic primary kind of. It’s getting a little hazy, but that may just be because there’s so many candidates. But I think it broadly describes what happened to [inaudible] Joe Biden, Bernie Sanders, Pete Buttigieg, Elizabeth Warren, and Michael Bloomberg. So four primaries in which I think the discovery, scrutiny, and decline model does a pretty good job, and I think anyone reasonably would look at it that way.

But in the other four, it doesn’t. It doesn’t really help you explain the 2000 Republican primary in which George Bush is just kind of perpetually getting a lot of coverage. Everyone else is fighting for scraps until you get this big John McCain surge, but that’s not a cyclical pattern. That’s only two candidates. It doesn’t help you with the 2008 Democratic primary, in which it’s Hillary Clinton and Barack Obama throughout. John Edwards a little bit at the beginning. But there are other candidates there, and they never see that pattern.

And it doesn’t help you with either of the primaries in the 2016, Democratic or Republican side, as already discussed. So I think it’s a useful model for about half the races, but it still leaves half, and the question remains as to why.

Matt Grossmann:So before this model, we had kind of our, I guess, historical model coming out of the party reforms that did focus on the role of the media, but also money, and especially momentum from the early states. And then a big intervention, and the party decides where the focus was on endorsements and really consensus of party elites around nomination. So how does discovery, scrutiny, and decline compare to those models? And is it fair to, I guess, say that those models work sometimes when the discovery, scrutiny, and decline model does not?

Zachary Scott: So you mentioned the post-reform era, and just for listeners, I think that’s probably worth some clarification. The means by which the parties have picked their nominees has varied historically. I won’t go through each phase, but through a lot of American history, the nominees were chosen at the conventions themselves. And the means by which the delegates to that convention were chosen, that varied, but delegates primarily represented the parties of those states. They ended up at these conventions, and nominees were chosen there.

After 1968, on the Democratic side, and the Republican side follows not long after, the parties reformed this process to tie delegate selection very closely for most delegates to the outcome of caucuses and primaries within those states. And now, there were caucuses and primaries before then, but they were largely responsible for not a majority of the total delegates, so a candidate could sweep all of these primaries and caucuses and still not have enough delegates to ensure themselves a nomination.

And in fact, part of the reason this happened is that in 1968, the Democratic nominee did not run in primaries or caucuses and still got the nomination right? So you could simply skirt this, what I believe Larry Bartels refers to as an outsider strategy, or I’m sorry, an insider strategy, simply just ignore going directly to party voters entirely and end up with a nomination. And in fact, outsider strategies that tried to specifically eschew party support and go directly to the voters never worked. After 1968, you have to go to the voters. You have to try and win over support in state primaries and caucuses to accrue delegates to try and secure the nomination.

So part of what I love about primaries is that they seem very complicated with all of the different candidates, and they take forever, but fundamentally are very, very simple. They are coordination problems. People within the party want someone who will win and will generally represent what the party wants. They do not want to lose the general election. It’s a matter of figuring out what candidate will do that thing.

Typically, you solve coordination problems in one of two ways. One is that you communicate. If a coordination problem is as simple as you and four friends are trying to figure out where to go to dinner, one possibly should text each other, and you figure out where can we all agree that we should go to dinner? When you have nominees chosen at conventions where everyone is meeting in one place, it’s pretty easy to coordinate, right? It’s still hundreds, thousands of people, but they’re all in one place, and they can discuss and horse trade and figure out who the nominee should be. When the nominee is decided by the voters, millions of people spread across the entire country, direct communication is impossible. You can’t coordinate via that means.

The second solution to a coordination problem is a focal point. Where can we all kind of fixate on and assume that everyone else is also going to want to go? If you’re trying to decide where to go to dinner with four friends, that may be the place that’s in the most middle or the most popular place. Something like that.

I think that you can really understand most theories of primaries, things that focus on momentum, things that focus on endorsements, on money, as means that candidates can use to become the focal point, right? That is, I think, the heart of the party-decides model, is that as these kind of coalitions no longer have direct control over the nomination at the convention, they can still use what tools they have available to them, namely endorsements, to signal to the voters that a certain candidate is the focal point, that this is where they should line up.

The thing is that that’s not the only means that a candidate can use to become the focal point. Maybe it’s money, right? Maybe they can just simply buy ads and put their face in front of enough eyeballs that people assume they’re the focal point. It could also be media coverage. These are all different means in which candidates can become focal points, and so the question really becomes in what situations are different strategies more or less successful? For example, it may be the case that being the focal point in the media is more successful when the party can’t signal that there is a focal point that people should line up behind instead.

In large part, I think these models have not always done a great job of talking to each other, and I will throw my own work under the bus. I don’t think I’ve done a great job at that either, and I think that’s something that we really need to consider. What are the situations in which certain focal point-oriented strategies are more or less successful relative to other ones?

Matt Grossmann: So how does the media decide what candidates to cover? You mentioned that they are sometimes responding to events, but they say that they’re responding to public interest or surges that occur before they start covering them.

And there are some examples of pretty direct decisions by the media, such as in this cycle, Rupert Murdoch reportedly sort of wanting someone other than Trump, and first settling on DeSantis, and then maybe more directly in recent weeks of Fox News, pumping Vivek Ramaswamy. Is the media independently making these decisions, or are they responding to kind of a consistent set of factors?

Zachary Scott: So newsmaking, I think, is best understood as the product of a confluence of factors. Journalists and editors and owners in some capacities, although I think Rupert Murdoch may be a bit more of an extraordinary case, they do have a lot of discretionary authority, but they don’t, at least in practice, tend to ignore reality.

I can’t recall many instances of party favorites, front-runner candidates, candidates who had obviously a lot of support, either from party elites or in polls, who were ignored by media, or at least not for long stretches of time. What’s more likely is that, yeah, they’ll be getting coverage, and then where the media is exercising discretionary authority is in who is portrayed as the primary alternative.

Now, that is still significant, because that suggests certain fault lines within the parties, right? If you just think of it this way, if the media are suggesting in 2020 that the primary candidates are Joe Biden and Bernie Sanders, that is suggesting that the primary fault line within the Democratic Party is not based on, say, age or generation, right? Those are two very old candidates, and so it’s highlighting a certain ideological or more establishment fault line as the most significant one within the party and ignoring others that may also exist and felt quite saliently by party voters.

So that’s a very significant power, but they are typically, with the possible exception of Donald Trump in 2016, which is not an insignificant exception to make, usually not ignoring what other factors are suggesting are powerful, important candidates within the race.

Matt Grossmann: So the other article for this episode looks specifically at the 2016 Republican primary and tries to assess what drives media coverage, and whether media coverage drives polling bounces for the candidates, and finds that the media are not really following online searches, at least as one measure of public interest, but that public interest is following media coverage, and that media coverage does not tend to increase support for other candidates, non-Trump candidates, but it does matter for Trump. That is, it is true that media coverage of Trump preceded his polling. So how would you compare that set of findings to yours?

Zachary Scott: I think that’s broadly correct. The question I have is how well that fits only the 2016 election, how that is also an overfitted result. I think that voters primarily find primaries to be overwhelming. It is my job to study them, and I find them overwhelming. They’re tough to make sense of, because there can be so many candidates, they take so long. There tend to be much smaller differences between the candidates in the race than there are in, say, the general election, so you’re parsing maybe minute differences. The differences are frequently stylistic rather than policy-based, which it creates a whole separate dimension of preferences.

They’re tough, and how voters respond to that is they want the focal point. They want to know, within some confines, where can they line up that they think other people are also going to be? And media coverage can be a powerful way of suggesting that. Not necessarily because the media coverage is significant per se, but for what people perceive it to suggest about other indicators; that if the media is covering this candidate, then maybe that means that other people are likely to already like them.

It could also be a third party effect, which is people tend to estimate that they are not affected by media coverage, but that other people are very strongly affected by media coverage, so it could be a result of that. I don’t know the exact mechanism, but I do think that people find media coverage, the quantity of coverage specifically, to be a useful queue to where they should line up that other people are likely to also line up.

Now, getting to the second point about the 2016 election, no primary, at least until 2024, we’ll see how this one plays out, sees a discrepancy in media coverage like the 2016 Republican race. In a field of 17 candidates, if everyone was getting equal amount of coverage, people are getting 6 or 7% of the coverage. Donald Trump, I don’t think is ever below 40%, and he’s frequently above 50. In a field of 17 candidates, a single one is getting sometimes a pure majority of media coverage.

If media coverage is not strongly correlated with support for the other candidates, well, relative to where Trump is at in any given point in time, they’re not getting all that much media coverage. They’re getting more, but it’s not necessarily all that much, especially if you think about it from the perspective of the voter. If they’re just paying some attention and they’re seeing four stories about Donald Trump and one about Ted Cruz, and yeah, maybe last week they wouldn’t have seen any about Ted Cruz, but that’s still not all that much relative to what they’re viewing. And that, again, is an exceptional case. Most primaries don’t look like that.

So I think the broad result that the media coverage can drive interest and the relationship the other way isn’t clear, I think that’s probably right. The degree to which that is idiosyncratically related to support for only certain candidates and not others, I’m not sure if that is a function of just that particular race or if that applies more broadly.

Matt Grossmann: You have also studied the appeals that candidates directly make in presidential primaries, including studies of populism and emotional appeals. What were the biggest findings and takeaways from those studies? How widely are those appeals and does that help us understand some of the unique dynamics of these primaries?

Zachary Scott: I recently published a paper in PRQ on populist appeals and presidential primaries, the first one you referred to. That was another project where the data collection that went into forming the corpus I used for a lot of these papers led me to something I didn’t otherwise know, that informed that project. Kirk Hawkins, who does a lot of research on comparative populism, he gave a talk at the University of Maryland when I was there in grad school about the serge of populism in presidential primaries, had some measures of the populism used by Bernie Sanders, and by Donald Trump, and a little bit by Ted Cruz, and was making the case that American scholars need to do a better job of reckoning with growing populism within American politics and specifically in engaging with the comparatives’ literature.

At the time I was collecting all of these speeches, and I just happened on my computer with me, I pulled them up and I started reading excerpts to him and he said, “Yeah, those are great examples of populism.” I said, “Okay, well these are by Jerry Brown in 1992.” Clearly populism is not a new thing within American presidential primaries, which led me to think if it’s not new, then how do parties react to it? Because maybe if we have populism continuously fomenting, but not leading to takeovers of the parties as the suggestion was that was imminent, then maybe there’s something that the parties are doing that is somehow engaging with or minimizing the threat of populism.

The approach I took to that is I looked at how the parties react when they face a populace insurgency, and specifically more electorally viable populace insurgency. When a populace candidate emerges in these presidential primaries who actually does poll pretty well or wins over some delegates, what do the parties then do?

And what I find is that I took this pretty large collection of speeches by presidential primary candidates throughout the primary season, about 4,000 speeches in total, and I find that once the candidates who win the nomination… Joe Biden after March 2020, but from that point in time when Joe Biden is the presumptive nominee up to the convention, you find that when nominees, presumptive nominees have faced a electorally viable populist challenger, they start to adopt some of the core issues of that populist challenger. They essentially assimilate some of the message as a concession to that populist insurgency to say, “Look, I hear you. You are really concerned about healthcare. And so, I’m going to start to assimilate that message.”

And they also do so with non-populist challengers, but the effect is actually quite stronger and more consistently found for populist challengers. And as a further extension of that, I took the actual party platforms from each of these election cycles and I essentially took a plagiarism detection software that is used to design if students are cheating, and I simply compared the words in platforms to the speeches by populist challengers. And I find that a populist challenger who does better in the polls… Or actually not the polls, I should say. A populist challenger who accrues more delegates, platforms tend to plagiarize from no speeches at a higher rate than they do from speeches by populists who don’t get that chair.

Again, suggesting that parties react, that when they see that they have faced a electorally viable populist insurgency, they then try and make some sort of policy concession to signal that they are willing to play ball with this populist insurgency. Interestingly, and I argue consistent with the asymmetric party’s thesis, that says that the Democratic Party is primarily a coalition of interest groups, whereas the Republican Party is primarily an ideological movement that is less resistant to offering policy concessions to keep themselves together. It’s the Democratic Party that does that in response to populist insurgencies, that Bernie Sanders runs in 2016 and Hillary Clinton says, “Okay, let me see what I can do to move my positions around to accommodate this.” Whereas insurgencies within the Republican Party don’t yield that same type of concessions.

And I think that that’s helpful for scholars of American politics in conceptualizing that populism happens and parties aren’t static in response, that they can be dynamic and may inform how we think about the threat of populism moving forward. The second line of research you discussed is some of my research that’s co-authored with Jared McDonald, the University of Mary Washington, that is looking at broadly symbolic appeals is how we characterize it in elections, mostly presidential primaries, although we’ve started to branch out to other levels of campaigns as well.

We’ve been doing a couple of different things with that, including we have some working papers in the pipeline. I think just as a representative case, we have an article in APR that looks at differences in candidate identities and how they use emotional appeals. We find that Republican candidates tend to use more fear appeals than Democratic candidates on average, and that women candidates tend to use more joy appeals than men candidates on average, even controlling for factors like their pole standing or the phase in the race. And the other finding from that paper is a null finding, which is that Davin Phoenix’s fantastic book on how anger appeals affect voters of different races implies that Black candidates will avoid using anger as a rhetorical appeal, because it just doesn’t activate racial minority voters in the same way, and also that it can come with a penalty.

We don’t find a difference in use of anger by candidate race, either looking at Black versus other groups or looking at racial minority candidates versus white candidates, which was surprising to us. I love that book. I fully expected that result to replicate and it simply didn’t, and there is a lingering question as to why. We have since, in a working paper, extended that to general elections and we find presidential general elections, that we find the basic results holding. And we have a couple other projects in that pipeline that go beyond just emotional appeals. Look at things like trait appeals. There’s again, part of this broader research agenda on symbolic appeals and campaign rhetoric.

Matt Grossmann: So you have another line of research that looks at when the media and candidates are talking about the same things, and I think related to that is some research on what happens when we have crowded fields and new candidates enter. What are the big findings there?

Zachary Scott: What led me to be interested in this question of how media cover the agendas of presidential primary candidates was really the 2016 election or the 2016 Republican primary, specifically. I like to play a game with people when they ask me about this research, which is, “Tell me what issues Donald Trump talked about in his campaign.” And people say immigration, they say trade, they say himself and all of those things are correct. And then I say, “Okay, tell me what issues Jeb Bush talked about,” and no one knows.

And he had an agenda. He talked a lot about healthcare and he talked a lot about education, actually, and those did not get through. He received some coverage. The tone of that coverage may be positive, negative, I haven’t actually looked at that. But regardless, the coverage didn’t talk about the issues he wanted to talk about. He had a case that he wanted to make to the voters and the voters didn’t get it. Now, if they had gotten it, I don’t know that they would’ve liked him more, I have no idea. But the simple message, there were 17 candidates in that race and they all came up with a message they thought was their best case to make to the voters. And for some of them, the voters got it, and for others, they didn’t. And I was trying to, in my dissertation, figure out why.

So what I use as a theoretical argument is my undergrad degrees were partially… [inaudible] degree in political science, I also got a degree in journalism and I hated being journalist, but I found the socialization into how to actually do journalism and learning how to be a professional very interesting, and we would get whizzed on these newsworthiness values. In my introductory journalism class in 2010, we had a quiz every two weeks in which we were asked, “What are the six newsworthiness values?” And let’s see, conflict, human interest, proximity, importance, timeliness and unusualness. To this day, I can recite them. And so, I wanted to see, is it the case that those newsworthiness values that drive how journalists process what is and is not news is affecting how they portray the candidate’s agendas when they actually go to generate coverage?

You referenced two different papers that are both in that line and taking different hacks on that question. One focuses on if candidates can portray their agendas in ways that are more enticing to journalists. What I ultimately argue in that paper, and this is an article that was published in journalism, Mass Communication Quarterly, is that candidates can use rhetorical newsworthiness cues to affect how enticing their message appears to journalists by leveraging the newsworthiness values that I just recited 13 years later.

Journalists think that conflict is an important feature of what makes something news. Can candidates affect, can they change how they present their message to seem like it involves more conflict, to essentially attract journalists to it? And in that paper, I argue that the way they can change how conflict appealing their messages is by invoking more anger. And I show that there is a pretty strong correlation between the amount of anger that a candidate uses in their announcement speech, and how well their agenda was presented in mass media coverage. Now, that I said correlate, I was very specific in that word choice because it is a correlation, and I had some reviewers along the way who suggested other plausible intervening variables. Maybe it’s not anger, maybe it’s negativity. If you control for candidates negativity, negativity is not correlated, anger is, maybe it’s not anger, it’s just general emotionality. If you control for how overall emotional the announcement speech was, it’s anger, it’s not that overall emotionality.

Points in favor that it is anger specifically that makes a announcement speech seem more newsworthy to the media. Another thing I suggested is that human interest is this important newsworthiness value that candidates who talk more about themselves and define their appeal more around who they are as people. For example, maybe if you’re going to talk about, let’s say, economic hard times, you talk about how your family specifically dealt with economic hard times. Obama would talk about how he personally understands the student loan crisis because he had recently paid off his student loans. Finding a personal way to portray it that will appeal to journalists sense of human interest as a newsworthiness value and will be more enticing to them. And again, I find that there is that correlation. The candidates who talk more about themselves, they tend to have more success at getting their agendas across in the media. And then, so candidates who are angry and talk a lot about themselves tend to have more success. Well, if we look at the external world, we can sometimes see that that plays out.

The second paper, which was published in APR is focusing on how the introduction of new candidates can affect what is not always explicitly described as a newsworthiness value for journalists, but I think still is active, which is simplicity. Journalists, I don’t think like to define what they do as finding a simple way to portray the external world, but you have a word count or you have a 30-second segment you’re going to air, you can’t dive into all the details, so they tend to like stories that are really simple, they’ll be able to portray very clearly to their audience.

And the basic argument is that when a new candidate enters the race, that brings a bunch of new fault lines with them. Okay, it’s a new way of conceiving what the different divides are between candidates, and that will essentially make the race less simple. And using a broader collection of speeches, not just announcement speeches, and applying some topic modeling to them to figure out what are the messages of the speeches versus the media coverage of the candidates, because the number of texts is simply too large to analyze via content analysis, I find that when you have new candidates entering the race, overall candidates tend to have less success at getting their agenda across. And in reverse, when the field winnows and people drop out and it concentrates in a fewer and fewer candidates, the argument being that the race becomes simpler, the fault lines become clear, journalists and the candidates who are still there see an increase in how much their agenda is permeating intermediate coverage.

Now, as part of that, again, observational research, I had some reviewers say, “I think your result is really just an artifact of horse race coverage,” that the media tend to focus more on the horse race early in the election and then later on they switch to covering substance when people are actually going to be voting and it’s more important, and that’s when you see winnowing, so you’re seeing candidates drop out at the same time that the media is increasing their substantive coverage for reasons that have nothing to do with the race becoming clearer.

So I replicate the results while taking, essentially, horse race out of the equation, and it does diminish the effect. It does become substantively weaker, but it’s still statistically significant in the correct direction suggesting that, yeah, that there is this phenomenon where as the race becomes more complicated with new candidates entering media struggle to… I shouldn’t say struggle. They tend to diverge more from the case that the candidates themselves are trying to make about why voters should consider them.

Matt Grossmann: Your Discovery, Scrutiny and Decline model found a few cases of fit with the Republicans and a few cases of fit with the Democrats, but some of your other papers found some distinctiveness across the two parties. So having looked at all of these different contests, how different or similar are the Democratic and Republican presidential races? Are there models that would only apply to one side or is it just depend on the specific cycle for each side?

Zachary Scott: Yeah, that’s a great question. It’s not something I think we as a field can confidently answer in any direction. My sense, partially informed by my own research and partially by the broader literature, is that there’s fairly minimal differences in how the media cover the races. I find some small differences here or there. I think in the paper on how the changing size of the field affects coverage, that Republicans tend to have less success at getting their agendas across the media, but it’s a pretty small difference. Overall, I have not found all that much evidence that the way the media covers Democratic and Republican primary campaigns is all that different. But as you suggested, other research I’ve done and that others have done, there does appear to be differences in the campaigns themselves between the parties and how the candidates make their appeals, not just issue ownership, but stylistic differences, their response to populism within the campaigns, that those do seem to be pretty different.

And that is pretty consistent, I think broadly with this argument that the Democratic Party is fundamentally a coalition of interest groups, whereas the Republican Party tends to be more of an ideological movement. I think the results do line up pretty well behind that thesis, which I think in turn begs the question of, “Should the media be covering the races so similarly, if they’re fundamentally different? Does the appearance of similarity, is that in itself actually a bias?” Which is an interesting question and it’s one that’s difficult to unpack, but it is something that I think normatively, we as a field should probably be doing a better job engaging with.

Matt Grossmann: Presidential primaries have been seen as a key lens for understanding how parties make decisions, what parties are, and yet some of your findings may suggest that that’s something of a mistake in that we only have a certain number of cases, they have pretty idiosyncratic dynamics. And if you take, say, like the Party Decides a Model or a simpler version of it and apply it to congressional races, it turns out the establishment favorite is identifiable and wins a lot. And so, maybe the kinds of things we should know about parties are more actively studied, are better studied somewhere else other than presidential primaries. What is your current take on that? How unique are presidential primaries as a lens on the parties and is it a sign, given that we seem to find some idiosyncrasies each time, that maybe this isn’t the appropriate lens for understanding the strength of parties or the dynamics of parties or what parties are?

Zachary Scott: Empirically, I think it is really difficult to study presidential primaries, because there are so few cases, the same lens that actually applies to presidential general elections. Sometimes someone will raise that point like, “Hey, you are basic, all of these on a set number of races, that’s really post the New Deal because that’s when we have reliable economic data, and you don’t have all that much data to actually work with. Are you actually sure you can generalize anything?” And the same basic logic applies to president primaries. They’re really, really tough to study. And because they’re essentially each individual case is so temporarily dependent on the situation in which it arises, it’s hard to generalize all that much across them. There’s always this counterfactual question of like, “Well, how would things have been differently in this particular time if this had happened, if you had had Twitter in 2000?” We had a bunch of isolationist candidates in the nineties and early 2000s and Robertson and Buchanan. Would those campaigns have been more successful, had social media been different or had the media ecosystem been different?

And those are unanswerable. Empirically, it’s tough to put too much stock in analyzing presidential primaries. That said, I think that what makes presidential primaries so important to study more than anything else is that they are one arena in which pretty much everyone within the party is forced to compete simultaneously. Parties are very complicated things. They can comprise a lot of different actors with different interests, and they’re not always active in the same fights, at the same time, and with the same level of resources, but everyone has to participate in the present or primary in some capacity.

And so, they’re just the one place where we can see it all, imperfectly, unfold. That said, that is really more an argument for studying presidential primaries out of convenience more than out of anything specific about them. And I have increasingly been thinking about how we can study the extended members of the party networks and what they do to lay the groundwork before the presidential primaries. I have this one working paper with Andrew Lug of University of Nevada, Las Vegas UNLV, that is very weird for me because it’s really about international political economy, a thing I never thought I would study. But the reason we have that paper is because my co-author was very interested in how isolationist sentiment spreads, and specifically in the insults about globalists, and how that took off. And he and I were talking about this and I started telling him a lot about Robert Welch and the John Birch Society and how this guy, who was never an elected politician, was a candy manufacturer, spent a lot of his own money circulating a newsletter and forming thinking groups, and how that is political power. And it’s political power that goes to lay the foundations that affect the way the presidential primaries unfold. And the presidential primaries are useful because they allow a situation to see that faction against all the other factions, and to see how they can pit their resources against each other, and how they form many coalitions.

And that’s convenient, but it’s ultimately just a reflection of what goes on between the cycles that is harder to measure and harder to study than even presidential primaries, which we already discussed have these empirical issues but are really important, and we need to think more about how we can do so. And increasing data availability formerly on social media, I think, was a very convenient lens and something that was handy, but it’s something we need to really consider because the primaries are long, but they’re still just a snapshot, and what goes on before the primaries has a really important effect, and it’s less visible, but just as important.

Matt Grossmann: Scott is skeptical that we can build a general model of how the media covers presidential nominations, as it seems to shift from cycle to cycle. Kevin Reuning studied the 2016 Republican nomination closely, trying to figure out whether media coverage followed or led public interest and support. I asked him about his main findings. So you studied the sources and effects of media coverage in the 2016 Republican pre-primary period. What were the big findings and takeaways?

Kevin Reuning: Yeah, there are, I think, two big takeaways from it. One, and this one is perhaps not super surprising, that media coverage drives a lot of interest in candidates. So in particular, we’re not just looking at, a lot of the research has been about media coverage and support. We were also adding in the interaction with just public interest in candidates as well. And so we find that media tension leads to public interest across sort of the whole field. We did not find a very strong relationship with media coverage and support, so increased media coverage leading to increased support, except for Donald Trump. He stood out as the sort of big exception where there was a clearer relationship between media coverage, increased media coverage leading to increased support for him.

Matt Grossmann: So you began with a discussion of how the media decide who to cover overall and in presidential primaries. So what are these basic media incentives and how much have they changed?

Kevin Reuning: Yeah, there are two, not necessarily competing incentives, but they end up being often competing. One is around just sort of journalistic norms. So these are the things that are taught in journalism school for those who go to journalism school or top through socialization in your profession. And a lot of those revolve around a sense of wanting to be unbiased or providing just sort of the facts on the ground. Though, there has been a big conversation in journalism about what unbiasedness means and what being a sort of neutral arbiter means, especially since 2016. And for those, the incentive is often then to be the journalist that knows what’s going on in the campaign, that knows which candidates really have support, and only focusing on the candidates that really are going to be good, not good, going to have a high chance of actually competing.

In contrast, there is a media incentive around just the fact that it’s a profit-making institution, or at least in most cases, the media is a profit-making institution. And for those, you have to worry much more about viewership, right? What stories are actually getting read, shared, things like that. And that has changed a lot in the lead of 2016 and then continued to change since then in part because there is a lot more data for those in the newsroom about what it is actually people are reading, right? 1970s you sent out a newspaper, you did a news show, you didn’t really know what people are paying attention to. Now, since there’s all these analytics on websites and online video, you know exactly what people are looking at and when, and their advertisers know what people are looking at and when. So that incentive can lead you to want to cover more of what people are interested in reading, which can often lead to coverage of things that are maybe not politically that important, but are sort of eye-grabbing, or attention-grabbing, I should say.

And so those are sort of the two different incentives, and that’s what we saw with 2016 in particular was that Donald Trump was seen in two ways. One is sort of sometimes explicitly talked about as a sideshow, something that wasn’t very serious, and so a serious journalist wouldn’t want to cover that. On the other hand, it became very clear that there was interest in Donald Trump as a candidate from viewers, too. And so media would want to cover that because media would want more people coming to their websites. And so those are sort of the two competing incentives.

Matt Grossmann: So you found that public interest measured by these online searches followed generally rather than bringing media coverage. How should we interpret that? And if the media isn’t following this particular indicator of public interest, is it possible they’re looking at other things like crowds or early state performance to guide their coverage?

Kevin Reuning: So one thing I think is important to note in this, and I feel like I can say it’s been several years, I’m no longer a grad student, and trying to worry about getting a job, is that I wish we had more fine-grained data. We’re looking at day-level interest and news coverage. Some of these dynamics might just be washed out in the fact that we don’t have information about what in the 10 minutes before something is happening, something blowing up on Twitter at the time, and that’s leading to media coverage, or is it the other way around? So I think there’s some parts where there is room for more research at trying to find more fine-grained information or getting more fine-grained data on those sorts of very particular dynamics that we just couldn’t get at the time. The other aspect of this is the actual events themselves and how people find out about them.

So Donald Trump famously wrote a golden escalator down to start off his campaign. People have to find out about that somehow, and they find out about that through the media. And so what is sort of missing in here is the fact that candidates are doing things, and those things that candidates do get filtered through the media often before they get filtered into the public at large. Though, that also opens up a broader question about what the media is, and especially given splintering of the media landscape plus the rise of just people who are social media influencers slash journalists slash just pundits, like what their role is since they’re disconnected from all this.

It’s a hard question, I think, to parse. I think what we can say from our finding is that, at least for these sort of big events, that if the media isn’t there to cover them or the campaign trail events, I should say, not necessarily big events, the media isn’t there to cover them. I don’t think anyone’s hearing about them. If John Kasich didn’t actually even make it into our data because there are such few search trends for him. And I think that’s in part because the media wasn’t there covering his, there weren’t many stories out there. So how would you even know to search about John Kasich unless you are a dialed-in politico of sorts.

Matt Grossmann: So, as you mentioned, you found that media coverage only helped Trump directly in terms of the subsequent polling. Why do you think that was, and maybe give us a flavor of the scale of the Trump campaign’s media coverage because it certainly did seem to be unending, and is that backed up in your data?

Kevin Reuning: Yeah, so I mean, Donald Trump dominated media coverage, not surprising. I think throughout the entire time, he had the majority of media coverage of all the candidates. So not just a plurality of it, but of all the media coverage of all the candidates, the majority of it went to Donald Trump, even though the field had lots of candidates in it at the time. And so there’s that potential aspect of it. There might be an effect where if you just can’t get a foothold in the media, people aren’t going to find out enough about you to really decide to support you or not. I think the other part which is harder for us to do in this context since we only really have a case of one of this, is the fact that Donald Trump was to a degree distinct from the rest of the Republican campaign or Republican candidates.

He was distinct both in his temperament as well as some of the policy positions he put forth, from having policy positions that might lean slightly towards the left to having policy positions that were racially conservative, if you want to use the euphemism there, right? To the comments that he made about immigrants. The fact that he was saying things that were distinct, I think, not only helped him make him stand out so the media was going to cover him more, but also if there is a subset of people out there that want those issues, that want those positions of a sort of nationalist conservative with more economic populism at times, if there are people that want that, then they can find that in Donald Trump. And that wasn’t in the rest of the Republican field. There wasn’t anything else like him there. And so I think it’s a hard research question to do to figure out why.

I mean, this is in some ways the question that a lot of political science, at least American politics, has been cycling around the last, what? Six years now, about what makes Donald Trump unique. It’s hard to answer. My gut says right now its some degree about the uniqueness of what he was and the fact that he just dominated the media to a certain degree. Of course, that then goes into, and I know your next question maybe is about this, is what does this mean for 2024 when he is not as unique of a character? A lot of his platform has been taken up by much more the Republican Party, though, it was interesting, there was only, I think, a few mentions of building the wall in the Republican debate, and that was a centerpiece of Donald Trump’s campaign was this Build the Wall slogan. And so he’s definitely shaped the party, but he still has some unique attributes. And so I think it becomes maybe less clear, maybe gives a chance for other candidates to potentially break through that.

Matt Grossmann: So we might be quick to point to policy positions, but obviously Donald Trump’s been getting media coverage since the 1980s outsized for his, I guess, direct influence in the world. And he also drew from a lot of, I guess, which you might call lowbrow consumer marketing, things like infomercials and multilevel marketing schemes and the hats and all of that. So, I mean, is it possible that what the media system got hit with is the same as the political science system? This was just a very, not a new set of tactics to get media attention, but a successful one in another arena that was newly applied to politics.

Kevin Reuning: Yeah, I mean, that is a good point. I mean, he campaigned in a way that was very different as well. I mean, I live in southwest Ohio. If I drive out of my college town, I still go past lots of Trump flags. I see Trump stickers all over the place. He was always very good at marketing himself, even when there wasn’t necessarily a whole lot of substance behind it. And that might just have played out here. I mean, you see CNN has made themselves way too much of the story here, which is not great on its own, but a lot of their post-2016 and now their post-2022 or whatever, with this executive being pushed out, there’s been a lot of like, well, what did we do wrong here? What did we cover wrong? And it feels, I don’t know, too easy to say this.

He’s not a traditional candidate, and it seems he is willing to do, I often joke that he’s sort of demonstrated that the goals, that it’s having no shame is basically a tool or a strength. That not being willing to apologize or not wanting to apologize, not being sort of things that I would be embarrassed about, having people call out things I’ve said. The ability to just sort of do it no matter what seems to be a strength that, I guess, from my perspective, thankfully not every candidate has though maybe that’s what the Democrats need now, is someone who is unwilling to ever admit that they’re wrong and to just say things to get attention. It makes me think of the Florida member of Congress, Grayson, who got in trouble or not in trouble, but kept talking about Republicans having blood on their hands. Maybe Democrats need more of that to be able to push back against the trumping of the Republican Party.

Matt Grossmann: So we had, coming out of the 2012 election, this discovery scrutiny decline model, which did incorporate the media very directly into the rise and fall of political candidates. And you did test one part of that which just is their connection between media coverage and then polling gains. But the other side of it was supposed to be that scrutiny that once media gets a hold of a candidate, they tend to scrutinize the candidate, and then the candidate goes down in the poll subsequently after that. So is there any validity to that? And in 2016, why didn’t the scrutiny of Trump lead to a decline, and where do you think that model is today?

Kevin Reuning: Well, it’s again hard because I think there was a degree, just not for Trump. I mean, I know our empirics don’t necessarily fully support this as clearly for the other candidates. But we did see anecdotally the, I don’t know, second tier, the sort of not Trump candidates appear, get increased coverage, scrutinized, and then collapse, right? I think, in part, some of those might’ve been over the whole span of the campaign. They might’ve been washed down the data just because it was such a short part of it. And so I guess the response to this is in some way, going back, we’ve talked about before, that Trump might be unique. I suspect that we’re about to see the same thing happen to Vivek Ramaswamy, that he is going to get a boost. He already was sort of in the uphill turn going into the debate.

I think that will continue, and then I would guess he’s going to drop back down in a month, maybe two, right? I think we’re seeing DeSantis following that. It’s a little bit different because he sort of started at the top, or started, I shouldn’t say at the top because again, Donald Trump is the top, but he started as the second-tier candidate or second-rank candidate and is on the decline. Vivek will follow that. And then the question will be, who will be next? And it might be Nikki Haley might follow that model, and it’s possible that Trump could see this sudden decline if he is convicted. But I think we do still, to some extent, see those dynamics. They just, for whatever reason, don’t seem to apply as clearly to Donald Trump that they do apply to the people who are challenging Donald Trump in the campaign.

Matt Grossmann: So I know it wasn’t the focus of the paper, but you did test for debate effects, and we just had our first debate and can look forward to more. So what is the state of that research? Do debates matter in primary elections, either in driving coverage or candidate standing?

Kevin Reuning: So I’m not super familiar with what research has come out since, but the general take is that debates don’t do a whole lot, at least for direct support, because most voters don’t watch the debates. The people who watch the debates are either deeply involved in politics in some way already and have their mind shaped or their positions already. That’s really it actually. I was going to say there’s another group, but that’s really the group that’s watching the debates right now.

And I think the biggest chance that they have to have an impact is actually through a longer process in the media, that this is going to shape some degree of media narratives that will play out over the next few weeks, that there’s going to probably be more focus on Nikki Haley in the next few weeks because of her performance, more focus on Vivek, again, because of his performance. While there might be less of a focus on some of the other candidates that could have had a breakout moment, I’m thinking of, I guess, Pence. Though, Pence is already a strange character to begin with as the former vice president who was deeply involved in the events of January 6th, 2020, and potentially hated by a set of diehard Trump supporters. But I think to the extent that debates could have any sort of effect, it’s through how the media decides that it’s changed things in their head, that journalists decide. Honestly, one of the things that’s also changed about 2020 or changed since 2016, 2012, 2020, is that we used to see, I think, a lot more of the…

We had a 15, 10-year window maybe, where we got to actually see a lot more of this journalist discussion because it just happened on Twitter. Right? At least for people who are on Twitter, you got to see journalists much more explicitly talk about some of these things.

Matt Grossmann: So we talked a little bit about whether a similar kind of character would be covered. Similarly, on the Democratic side, we have had a few additional examples. You talk a little bit about 2016 coverage on the Democratic side. In 2020, we did see pretty dramatic media influence, but it was more consistent with elites and former candidates gathering around Joe Biden when he was threatened by Bernie Sanders. And he won a primary in South Carolina.

This time, we saw perhaps an interesting example where RFK Jr. was being hyped more by the conservative media even in a Democratic primary. So that might suggest that something might be different on the conservative media side more than the Democratic voters and candidates. I guess, what is your assessment now of how different we should expect the dynamics to be on the Democratic side versus the Republican side?

Kevin Reuning: Yeah. As you’ve literally written about here, there’s clear asymmetries in the parties. And these asymmetries, I think, play out in a variety of places, including in media.

I think both media… How media covers Democrats and how Democrats respond to media coverage, there’s much more of a willingness for Democrats to backtrack on things. I think the media, for as much as a lot of journalists are left of center, they are willing to be critical of Democrats much more than I think some of the conservative media is being willing to be critical of conservatives.

So there’s all these asymmetries at play. And I also get the feeling… There’s a great paper or there’s been a set of papers that have come out in science on Facebook and polarization. And buried in one of them is… Actually, the appendix of one on polarization is this great graph showing that, basically, media… Website sort of bias, they don’t call it bias, but it’s like website lean from left to right.

And you can see that there’s this whole spread of some very left-leaning and some very right-leaning websites. But what I found very interesting is that most of the left-leaning one, they have a range of people that are still reading them. I can’t remember any of them off the top of my head for the left side, but like Jacobin, to the left as they are, they’re getting some central-left people still reading it.

On the right side, the conservative side, there’s these sets of media sources that if you are not very conservative, you’ve just never heard of, like, again, the Western Journal News, Gateway Pundit. These are not things that are read by people that are not very far to the right.

And I just think that, I don’t know, it was eye-opening to me in that the degree of asymmetry in this… This is kind of going far afield from your question, which is about the dynamics of the Democratic Party. I do think that it would follow our more traditional expectations of party primaries that we’ve seen in the last 30 years or so would be my guess.

I think 2016 was unique in part because there was only one real clear challenger to Hillary Clinton, Bernie Sanders. There’s a few very long, long shots like Martin O’Malley and one other I can’t remember right now. But that made for a different set of dynamics. You have a two-person race, opposed to this multi-person race.

In 2020, we saw much clearer… much more similar dynamics. And I guess the question is, what will happen in 2028? It’s hard to predict the future, obviously. And I think a big part of it is about where the media ecosystem is by then. But I would suppose it’s going to follow the older model more than something newer. I just don’t think there’s anyone who has the hold of Democratic Party voters as Trump does the hold of Republican Party voters.

Matt Grossmann: So political science got a bit of a Black guy after the 2016 primary that you studied because a particular over-interpretation of the book the party decides was seen as making it very unlikely that Trump would be the Republican nominee because he lacks support from party insiders. But that doesn’t mean that we should all throw out that model or the role of endorsements in this process.

So where does that stand at the moment? And if it was dislodged, is it a reason to return to the more old models, which might be a little bit more consistent with your findings that were more about media and momentum?

Kevin Reuning: Yeah. As I think, actually… I’m going to go into, I think, a bit of your next question here, if I remember correctly, about presidential primaries. Part of the answer to this is that we only have so many presidential primaries. And parties actually do a whole lot more than just run presidential primaries, or don’t even run, participate in presidential primaries. Right?

I think there is a world where The Party Decides position is still pretty true, and I see it… If I ever have a chance to write this paper, I’m going to, if I have the time to, is about more of a punctuated equilibrium world where the party elites are able to constrain things to a certain extent. But when you have these candidates that are outlier to the norm for what the party thinks, they have the chance of breaking through and pulling things in a new direction.

The Party Decides literature has a view of party change that is very top-down, of party elites and interest groups navigating differences of opinion to modify their positions. And what’s potentially, I think, missing a lot of that is the fact that you could have these political entrepreneurs that demonstrate that there is support for something that is outside of what has been seen as the party’s wheelhouse, so to speak.

And so I’m not ready to throw away that model entirely. I think we just need to think through more closely the dynamics of these exceptions to the rule which are important to understanding it by itself. Donald Trump has also… he hasn’t broken the entire system, right? There was just now party elites have taken on a new set of positions, or not even necessarily positions all the time, just ways that they approach politics.

And there’s been a new set of party elites brought that have risen as well. We see that very clearly with the chair of the RNC, Ronna McDaniel. [inaudible] Ronna or Ronda. With McDaniels, who had, let’s say, a change of heart over all of this, and is very much more willing to follow where both a lot of the elites are and the voters are.

The model is perhaps a little too simple, which is what we do often in social science. We simplify things that are useful at the time and then make them more complicated as we realize that not all of our findings fit with that simple model.

Matt Grossmann: So yeah, as you mentioned, we have other ways of looking at the party. And if we think the presidential primary process is too idiosyncratic or we don’t have enough cases, the endorsement or party elite-based models have held up better at the congressional level, for example, not that outsider candidates never win, but that endorsements seem to matter there.

Obviously, we could look at the parties through their interest group networks or their party organization as I know you have done. I guess how does our view of parties change if we downgrade the importance of this presidential primary process? Do they appear stronger? Are they evolving differently?

Kevin Reuning: The question if they appear stronger, I think, is interesting because it’s not necessarily that they are… Parties in the United States are famously weak, I guess, or party organizations in the United States are famously weak.

And it’s even hard to talk about parties in some ways because we have the classic tripartite party model; party in the government, party in the electorate, and party as organizations. And the degree of strength of those varies. And also, even when we talk about party as organizations, we have a much more complicated set of things from the organizations that exist around it to the actual official party organization.

I think the thing that we need to bring back in right now is a clearer look of where pundits and more media actors fit into this network, into this, I guess, set of party individuals or party things. The media clearly plays a role. And what we see more and more is parties, because of social media, being able to act as media in some ways too, being able to reshare news, shape how their voters or their members see things.

Not everyone is following the RNC on Facebook or Twitter, or their state party, or whatever. But for the party activists who I think are increasingly important, they are following these social media accounts. I think more needs to be also said if I’m just laying out research I wish would get done, looking at how people have opted out of politics and what that means for the power of party activists.

I think more and more people just see politics as something that other people do. I know, in some ways, we’ve had increased voter turnout and things like that. But there is, I think, a hardcore group of people who just don’t care about politics whatsoever because there’s so much anger and fighting.

And in some ways, that empowers activists even more. And those activists are the ones that are running party organizations, running these interest groups. And so maybe the organizations themselves aren’t that powerful, but I think the activists have become more influential.

Matt Grossmann: Trump is sometimes grouped with things happening in other parts of the world. And there is certainly some evidence that Democratic backsliding is occurring more widely across rich democracies.

So it is, I guess, a little bit weird for Americanists to explain Trump by saying we have these convoluted dynamics for our presidential primary process and specific media infrastructure that enabled Trump to rise. How do we place that series of explanations in this global pattern where people would point to wider trends or even, in some cases, the impossibility or difficulty of having a multiracial democracy succeed? How can we get our more particularistic explanations in that context?

Kevin Reuning: I’m going to end on some high notes here. The joke that I heard in grad school is that Americanists are just comparativists with an N of 1. And so I think there’s a lot that we can learn from the comparative political literature.

I do think there’s two tensions, and these are tensions that exist in a lot of political science, or there’s two points of tension. One is that, to a degree, the culture, the institutions of any single country are really important for understanding those country’s dynamics, right?

On the other hand, we do think there are some sort of general phenomena that are generalizable from one country to another country. And it’s possible that, as Americanists, we have done too much of the former, and we need to start doing a bit more of the latter as well.

The other question there, and I think it’s the big question that a lot of political scientists are wrestling with now, is the role of the internet and social media, which is often seen as a through line in a lot of these different cases. And it’s hard because in the general… What is it? From the 30,000-feet view, it looks like social media matters a whole lot.

But as you start looking more closely at specific dynamics, it’s hard to figure out how exactly social media matters. And so I think we can maybe get some better traction on these questions by looking at how the internet and social media intersects with different types of electoral systems.

And that could be helpful of us trying to understand American politics to draw on that literature. But I’m also not going to throw out the American politics subfield or field as a whole because I do think it’s useful for us to be very thoughtful about what the background of our case, so to speak, is, and understanding those in more specific ways. Right?

Yeah. As I talk to my students, we need to figure out what is unique about this case and what sort of things are unique in important and useful ways. And so some of it is that America has this history of race and racism that isn’t necessarily entirely unique, but the ways it’s played out in American politics are maybe more unique. And this shapes a lot of our ongoing politics, from how it shapes institutions to how it shapes people’s opinions and perceptions of the world.

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. If you like this discussion, here are the episodes I recommend checking out next, all linked on our website.

“Do Early Primary States Still Pick Presidents?” “Did Americans’ Racial Attitudes Elect Trump?” “Congressional Primaries: How the Parties Fight Insurgents.” “Racial Stereotypes in Voting for Obama and Trump.” And, “Do the Parties Prefer White Male Candidates?”

Thanks to Zachary Scott and Kevin [inaudible] for joining me. Please check out “Replicating the discovery, scrutiny, and decline model of the quantity of media coverage in presidential primaries” and “Media Coverage, Public Interest, and Support in the 2016 Republican Invisible Primary,” and then listen in next time.