STEVE GOLDSTEIN: The power of social media on politics and elections isn't restricted to President [Donald] Trump's tweets and the retweets of his followers. And it isn't always more conventional celebrities like Oprah Winfrey and Tom Hanks or younger singers and musicians who are having an impact. Sometimes it's people who are mostly famous for being online influencers. Yotam Shmargad, [assistant] professor at the University of Arizona, studied — among other things — how closely voter registration tracked with influencers online. He defined "influencer" as anyone on Twitter who on average received at least 10 retweets for candidate-related comments. His focus was on congressional candidates, and he's with me to talk about the study and his conclusions. Yotam, let's talk specifically about the accounts you followed. How did you characterize them?
YOTAM SHMARGAD: What I characterize them as activist accounts. They readily post messages, tweets, about politics, and they have a clear partisan leaning. So they're either usually almost always posting pro-Democratic or pro-Republican messages, and they don't have any affiliation to kind of a formal political group. They're not working at a political organization. They're not working for a campaign, at least given their Twitter bio or Google searches that I performed. And so really, the only way to kind of call them is, is an activist. And well over 50% of the accounts that both retweet congressional or governor candidates and have that 10 retweet per tweet threshold are these activist accounts — they're not kind of traditional media folks.
GOLDSTEIN: So, Yotam, is there a way to measure what these activists in some way influenced people who are following them to act in a certain way? By that, I mean, are there examples where maybe more people registered to vote but they didn't actually vote or it sort of raised awareness in some sense, but didn't necessarily move a person to action? Was there a way to measure that? Did you try that?
SHMARGAD: Yeah. So these connections are obviously very difficult to make, because so many things are going on it's hard to know that, you know, one thing, one endorsement actually played a significant role in getting people either to vote or to vote for a particular party. In a study in 2016, I kind of — I did the same thing, looking only at congressional races, so only at Senate and House races. And I was able to establish a relationship between candidates' indirect influence. So on average, how much their retweeters were being retweeted themselves and the vote, the voting share of, of that candidate in, in their particular race. Not only did I find that relationship, but it actually also turns out that indirect influence is associated with a smaller vote gap in 2016 than in the previous election year in 2014. So this doesn't have to do with some particular district-level characteristics or state-level characteristics. It's really something about the candidate that's producing this indirect influence, allowing them to do better in, in, in the kind of, the voting booth than the same party's candidate in the previous election year.
GOLDSTEIN: Well, right, considering how lesser known candidates typically have a lot less money to spend on the traditional things we'll see — the advertisements, the mailers. It would seem that those who are able to get some influencers online to either support them, retweet, whatever it may be, it would seem that is a really dramatic shift in the sense that that's a way that the lesser-funded candidate actually can have an impact where he or she in the past would have really struggled even more.
SHMARGAD: That's right. And so we only — I only find this relationship for candidates who had less money than, than their competitor. There's, there's kind of no relationship between anything I see on Twitter and how well a well-to-do candidate does. So if a candidate is much well-funded, much better funded than their competitor, there's no relationship between what's happening on Twitter and what's happening at the voting booth. Both direct and indirect influence, there is just no relationship. The only relationship that uncovered is with indirect influence and for candidates that have less money than their competitor.
GOLDSTEIN: We've obviously seen this development over the last three-and-a-half to four years where we have a president who uses Twitter in a way that is incredibly influential. It's almost as if he speaks to his supporters via Twitter as opposed to the more traditional means. How impactful do you think that has been? And do you think we will see more people down a ticket as, as elections go by? Whether, let's see, in a, in a midyear — a midterm election in 2022 — do you expect to see more people who are running for Congress, whether they are well-funded or not well-funded, actually see this is even more of a platform?
SHMARGAD: Right. So speaking to the president specifically, it's really hard to know how much Twitter is having an influence compared to the host of other venues that the president is able to get his messages out. So you know, Donald Trump in many ways is more of a TV candidate than, than a social media candidate, although obviously he does, he does tweet quite frequently. Donald Trump created his entire career as a TV character. A lot of the reason that people know about Donald Trump is because of the attention that he received from the media on kind of traditional platforms, like TV and news and billboards and, to some extent. And so it's really hard to know how much Twitter specifically is moving the needle when all these other things are going on. But with respect to congressional candidates in lower level races, I think what one of the things — that it seems like one of the things that's really going on is that these influencers are able to align local candidates with kind of national priorities. So the influencers will take a message that a local candidate has produced and then align that message with kind of either Trump's policy or Hillary Clinton's policy and policies in 2016. And presumably, Biden's policies these days.
GOLDSTEIN: Thinking of the old traditional, whether it's the newspapers or the signs we'll see, and so-and-so endorses so-and-so. And we, we often debate in newsrooms whether endorsements were ever meaningful or if they mean anything now. Is in some ways the influence of social media or people who are retweeting others, is that in some way a new generation of the old, his newspaper endorses so-and-so?
SHMARGAD: Yeah. So that's the argument that I try to make. There's, there's been a concept, as you've described, of the elite endorsement, which is that somebody in the party endorses a candidate. Usually that's being discussed, at kind of a primary, really the presidential primary step, where in order to know who's going to win a party's primary, it's really helpful to know who else in the party, who else that's influential in the party is endorsing them. Now, that kind of formal party support is very useful in many cases. And so in most cases, it is the case that the, the candidate, the primary candidate who receives the most endorsements, is ultimately going to be the person who runs under under the party label. That was not the case in 2016. And so Donald Trump was not the the primary candidate who received the most endorsements and yet still was able to navigate himself and win, and win the primary. And so the reason for that is that there are these other forms of support, these informal measures of party support, that are not captured by, by what elites are suggesting voters should do. And, and what I'm arguing is part of this informal party support I'm able to capture by, by looking at Twitter and looking at how influential the endorsers of candidates' Twitter messages are.
GOLDSTEIN: Yotam, thanks very much again for the time and for the work you've done.
SHMARGAD: Thank you, Steve. Really appreciate talking to you.
GOLDSTEIN: Yotam Shmargad is an assistant professor in the UA's School of Government and Public Policy.
