MODERATOR: All right. Good morning and welcome to the Washington Foreign Press Center. We’re happy to welcome all of our journalists in D.C. and those who will be participating from New York by digital video conference. Thank you for coming out to the latest in our mid-term election programming.

Before we get started, I wanted to remind everyone that our briefer is offering his views in a personal capacity and does not represent the official policy views of the U.S. Government.

Our briefer today is professor Ethan Porter, an assistant professor at the George Washington University School of Media and Public Affairs who received his PhD in political science from the University of Chicago in 2016. Today he’ll be talking about social media and the elections, and a lot of the information in this briefing he will also have in his upcoming Cambridge book False Alarm: The Truth About Political Mistruths.

And with that said, I’d like to pass it along to our briefer. Thank you for coming out.

MR PORTER: Thank you. Thanks for coming. Thanks, Bryce, for having me. So I’m going to have a few remarks. I’ll sort of give you a lay of the land and then feel free to shoot questions. So some of this you might already know. I’ll give you a 30,000-foot view and then I’ll walk you into some of the recent research on the subject.

So, as you might know, the majority of Americans are now consuming social media, though there is some evidence that perhaps demand has begun to plateau, which is to say that the number of Americans who actually are signing up for social media may have tapered off. When it comes to – when it comes to younger people in particular, there’s strong evidence that they are actually receiving their news primarily through social media. So rather than reading a newspaper or watching a news channel on television or even listening to a news program on the radio, people are receiving their news through Facebook. Facebook is a major source of news for young people. And again, for the young people, it might be the majority of them are now receiving their news over Facebook.

So the ubiquity of social media has made it a tempting target for political campaigns, which, again, probably isn’t news. In the past, in previous elections, there’s been a kind of Wild West quality to the role of social media in politics, which is to say it was largely unregulated. What was being shared on social media by political campaigns was done so perhaps outside the purview of federal regulators and certainly often unbeknownst to the media companies themselves. That’s changed. It’s now much more difficult for a political campaign to actually issue its advertisements and distribute its messages over social media. There are really aggressive regulations that, in Facebook’s case, Facebook has put on itself to control the flow of political information on Facebook. You need to go through a verification process to verify that you are in fact a political campaign, and you then need to show that your messages accord with federal election law and also accord with Facebook’s own terms of service.

So from a campaign’s perspective, the role of social media in the election has changed, which is to say what you’re doing is now – what they’re doing is now being scrutinized more seriously. It’s easy to overstate the role of social media in American elections. It’s easy to think that because of a few bad actors, a few bad apples, social media is simply causing people to vote a certain way or hold certain beliefs. That’s largely actually not the case. Media effects are small. And by media effects I mean the effect that any one media message or media story is going to have on people’s political beliefs or political behaviors. I think this is often disappointing for the media to hear, but media effects were small before social media and they have continued to be small in the age of social media.

Now, one thing I really want to emphasize: Just because media effects are small doesn’t mean that all sort of media – all media efforts by campaigns are within the law. Right, we still have a set of rules and laws in this country that govern the kind of messages people can put on social media and how they need to identify themselves. So whether or not media is effective on social media is a different question from whether or not it is legal. I think that’s really important to be clear about.

So why are media effects small even in social media? Social media gives users the opportunity to encounter countless messages, right? People will scroll through their Facebook feed, their Twitter feed, and they’ll see lots and lots of messages, meaning that the cumulative impact of one particular message is probably likely quite small, even messages that are designed to spread misinformation, to sow antipathy toward the other party. These messages are actually really unlikely to be effective at achieving the desired result. The best metaphor that I can think of, especially in a campaign cycle, is the game tug-of-war, right? Especially in a heat of a campaign season, both parties are flooding the airwaves and flooding social media with messages, and the net effect then of any one message is going to be pretty limited.

So social media is pretty small. That doesn’t mean it’s not concerning, right? Particularly when it comes to misinformation and disinformation. So I’ll give you some recent research into this. I’ll give you a little bit of the lay of the land.

So the empirical evidence on who consumes misinformation on social media suggests that the most – far and away the most common users tend to be older. Older users tend to consume a lot more fake news over social media than younger users, which suggests that there’s some element of social media literacy that probably matters, which is to say younger people might simply be better at realizing, okay, this is clearly bogus, and I can separate fact from fiction better than my elders. There’s evidence of that.

The other thing that I’ve shown extensively in my own research is that misinformation, misleading statements can quite easily be corrected, which is to say that if a campaign or candidate makes false statements or misleading statements, if that misleading statement is then paired with a correction, there is good reason to think that the people who received the initial misinformation are going to respond to the correction, right?

So the journalistic practice of fact-checking I think doesn’t get enough credit. My research has shown repeatedly and exhaustively that fact-checking does actually move people in the direction of factual accuracy.

The challenge, though, for social media is that, again, according to the best available evidence, there’s a mismatch between the people who consume misinformation and the people who consume fact checks, right? So the people who consume misinformation are quite literally almost never the consumers of fact checks, right? So when I run a study in a stylized lab, I can force people to see corrective information, but it seems that without a researcher pushing them to receive corrective information, people who consume false information are really unlikely to ever go seek out corrective information.

Finally, the other sort of constant in a lot of my research is that factual corrections have little bearing on people’s political attitudes or views, right? So when a supporter of a particular political party receives corrective information about a false statement made by that political party, there’s good reason that that person will actually accept the facts but then will not change his or her attitudes toward the political party that has received the correction, which I think is really important to understand because it suggests that people are receptive to factual information across party lines in this country, but the relationship between their party preferences and their – the facts that they have at their disposal is a pretty weak relationship.

So that’s really about it. I’m happy to take questions and answer what I can.

MODERATOR: All right. For questions, please state your name and the organization you work with and try to limit yourself to one question, if possible. If there’s anything that he’s not able to answer, you can reach out to us at, and we’ll try and answer any questions that need follow-up.

QUESTION: Hi, thanks for doing this. Nirmal Ghosh from The Straits Times. I wonder if you could speak to – you’ve quite well described the social media environment as very democratic, egalitarian, everybody has access to everything. But there are people who move off the more public platforms into more confined spaces, and these are essentially even more hermetically sealed echo chambers. For example, I think there’s an app called or a company called uCampaign which has developed these platforms. And you also have the phenomenon – I don’t know about the U.S. but in some other countries – of closed WhatsApp groups. And misinformation as we know travels faster than the facts, right? So I wonder if you could speak to these sort of more in the shadows in social media. What’s going on with it?

MR PORTER: Yeah, so there haven’t been in the United States – as far as I’m aware, there haven’t been cases of misinformation over WhatsApp leading to terrible, unspeakable crimes in the way that there have been in other countries. It doesn’t mean that we’re somehow invulnerable from that problem; it just hasn’t happened yet, so far as I know.

In general, there have been different companies that have tried to create social networks just for conservatives or just for liberals, just for Democrats, just for Republicans. Those have mostly not done well because it seems that people enjoy – and again, this is speculative, but if you just sort of look at the evidence, it does seem like Americans take pleasure in experiencing their friends’, their families’ lives more so than politics. We’re not a very politically engaged country, so social networks that are just about political identity are probably going to attract actually a tiny, small fraction of the population, especially in comparison to the number of people that Facebook attracts. In other words, people like seeing photos of their friends and relatives more than they like politics.

I also think with echo chambers, it’s easy to overstate the harm that echo chambers can bring about. I think some of the evidence for that is really mixed in the United State. Elsewhere, though, I really can’t speak to.

QUESTION: Sriram Lakshman. I work for the Hindu newspaper in India. You said media effects are small typically, right? But we never consume social media in drops, one message at a time, so I’m just curious if you could comment on that. After a certain point, I am constantly getting a barrage of information over months.


QUESTION: And that brainwashes me.


QUESTION: Does it or doesn’t it?

MR PORTER: Well, it depends on what we’re interested in, right? If we’re talking about does media have the ability to affect how people vote, in total maybe it does, right? But for any one message, like you said, it is likely to be sort of part of a much larger set of messages that we’re receiving at any one time, right? So it’s possible the media is having long-term effects in structuring our views, but a political campaign – I mean, I see this all the time and I’m sure you do too. It says if we release this ad, we’re going to win by this much. And that’s really actually not the case often. Effects of one particular media message are pretty likely to be small.

QUESTION: Thank you, sir. This is Jahanzaib Ali from ARY News TV Pakistan. I’m a bit late in the press conference, as usual. I do that. (Laughter.) So sir, what I understand, what I’m listening to is you say that the social media cannot do any change for the results of the elections and especially there are – social media is a huge platform for the fake news. So I mean, it doesn’t matter, right? You say that it’s nothing to do with the elections, that it can’t influence on the results, right? That’s what you say?

MR PORTER: I think it’s easy to overstate the probability that social media will have a decisive impact on the election, right? It doesn’t mean that there’s not a possible world where social media is not affecting the election decisively, but the idea that social media is going to determine who wins, I think that’s unlikely. Does that make sense?

I would also say too, the other thing to keep in mind, the other emphasis that I would have, is that when it comes to misinformation, people can be corrected of that misinformation, which I think we often forget, and we do so at our own peril. The problem, again, as I mentioned in my remarks, is the mismatch between people who consume misinformation and the people who consume fact checks. Those two are not meeting enough.

QUESTION: Hi, my name is Philippe Gelie. I’m with Le Figaro of France. Some candidates have used social medias to produce videos, some of them becoming viral. Have you any idea how they do compare to traditional TV ads in terms of efficiency, considering that what you implied earlier is that on the social networks, people see what they want to see, right?


QUESTION: Which may not be the case on TV. I’d like – can you compare the two?

MR PORTER: Yes, yes.

QUESTION: Thank you.

MR PORTER: I think your instinct is right here, which is to say if I’m a political campaign and I’m buying advertising on television, I’m going to reach people who I might not necessarily be able to reach otherwise. This is – now TV ads now have what’s called addressable TV even, where I can say I need to reach 6,000 people in a particular congressional district, I know exactly what these people are like and I want to only go on those people’s televisions, right? That allows a degree of targeting that is not possible and is certainly not possible with viral video, right? Viral video, if it goes viral, it’s likely to go viral because it’s – likeminded people enjoy the video, right? And campaigns are often interested in persuading other people, right? They want to sort of reach other people and say no, no, vote for my person, not your person, that kind of thing. And that’s – it’s really hard to do and sort of – although it might be the case, it’s hard to figure out theoretically how that would work.

MODERATOR: All right. First we’ll go to New York, and then if the gentleman in front of Philippe still has a question we’ll move.

QUESTION: Yeah. Nikkei Business, Japanese magazine. So my question is especially the last election we see many, like, Facebook impact, and also Cambridge Analytica, that kind of issue. So that kind – educate people we have to be careful for the information comes from the social media. How do you think this going to impact to the midterm election or any other election after this?

MR PORTER: Sure. I think a few different things. When it comes to Cambridge Analytica, they sold – during the campaign they told people we’re doing all these really fancy things and we’re doing – using these psychographic methods, but in reality they were a bunch of crooks, which is to say that they were doing things that were against Facebook rules and maybe against the law. People I think have become more cautious about sharing their own information online, right – there’s a premium for privacy now that Tim Cook, the CEO of Apple, has talked about, right? People understand at the highest levels of tech that people are going to be more protective of their own personal information than they were, and that’s because of things like Cambridge Analytica.

What this means is that, as I mentioned, media effects are probably small in general. They’re probably only going to get smaller, because people are going to become more hesitant to share their personal information, which is the bread and butter of social media advertising. If people aren’t sharing their personal information, then social media advertising will become even less effective.

QUESTION: Hello. Yoshinari Kurose from Japanese newspaper Sankei Shimbun. So does that mean that the candidates are not putting so much stress or importance on social media compared to the TV ads or newspaper ads?

MR PORTER: That’s a great question. People – there’s a – campaigns are still putting lots and lots of money toward television advertising, especially – and I – this is true the last time I checked. I – maybe we should check this again. On the Democratic side, it’s tilted really strongly toward TV, and that will likely continue to be the case. However, there’s obviously a platform – there’s obviously a space for digital media as well. The rise of YouTube, Hulu, other sort of social media sites that allow – and Hulu’s not really a social media site, but other video sites that allow users to directly – campaign advertising purchases to directly target users, those are likely to increase in the share of purchasing that they represent, Hulu especially. And any time you can force people to watch an ad, which is not the case with television, especially with fast-forwarding and rewinding, that’s a good thing for your campaign.

MODERATOR: Gentleman in the second – second.

QUESTION: I guess this question may have been asked in a different form before. You’re suggesting there’s —

MODERATOR: Could you give your name and outlet?


MODERATOR: Could you give your name and outlet?

QUESTION: Yeah, Sriram Lakshman from The Hindu. You’re suggesting that there’s some sort of self-selection going on in terms of the people who consume updates on facts and those who will believe anything, right, and those – there’s not much of an intersection between those two groups. That’s kind – that kind of makes you feel hopeless. How do you solve this issue?

And – but you’ve also said that media effects are small, and I get the general sense that social media just doesn’t matter that much. So in a sense that’s good news, since this is going on. Why does the President of the United States then spend so much time on social media? And that’s maybe not a fair question to you, but because you keep hearing he’s trying to fire up his base, he’s trying to fire up his base. But if media effects are minimal, why – I mean, he’s obviously getting inputs from experts, and maybe they’re just incorrect, but what’s going on here? Can you shed some light on that? Thanks.

MR PORTER: Sure, I’ll try to answer. I would say first I would never presume that anyone is getting advice from experts. Maybe they are, maybe they aren’t. Maybe he’s making arguments and making claims on social media for no reason at all and without any good evidence. That seems reasonable to – that might be true.

I would also say that traditionally in the study of public opinion, we accord the president a great deal of – we view the president as having an especial degree of influence, right, which is different from all other actors in American public life. So are people – are President Trump’s followers reading all of his tweets? No, probably not directly, but the President, by virtue of being the President, his messages are reflected and refracted throughout the media universe. They’re amplified, right? When he tweets something, it’s not only those on Twitter, it’s – everyone sees it. And so the effects of the President on – the ability of the President to have effects there are probably certainly larger than most other people’s and most other campaigns.

MODERATOR: The gentleman in the first row.

QUESTION: Nirmal Ghosh again from The Straits Times. I’m kind of curious, there’s this – you’ve said very explicitly that it’s difficult to measure or quantify and it’s probably improbable that social media advertisements and campaigns actually determine the result of an election. Now, what if the objective is broader and deeper than just determining which party wins what contest? For example, if you look at the federal charges against – on Operation Lakhta, for example, the Russian woman, Elena Khusyaynova, right? The objective was to sow discord, to basically erode faith in the system. Not necessarily one political party or one candidate, but to exacerbate social discord so the whole system starts coming apart at the seams, right?

So is social media effective? It seems to have been. Anecdotally, for us who consume social media and follow the news all the time, it seems to be pretty effective.

MR PORTER: Right. I suppose – right, as I think I mentioned earlier, it depends on what your objective is, right? If you’re a political campaign running a congressional campaign in Iowa, it’s really easy to be wrong in assuming how influential social media will be. If you’re a adversarial power, it might be the case that social media is, if en masse, more likely – capable of being effective at driving the public conversation, driving the conversation that elites have, right? Which is also an objective that especially can be achieved on Twitter.

So again, I think it just depends on what you’re talking about, right? When we – when I say the media effects are small, oftentimes in that – and the research, that’s because we’ve looked at media effects in terms of their ability to sway people’s beliefs about a candidate, about an issue, about a political party. And in that case, media effects are small. If you’re talking about these larger macro objectives, it may be the case that effects are larger. That’s, again, just sort of hard to say at this point.

MODERATOR: The journalist in the third row.

QUESTION: Thank you. I’m Zhaoyin Feng with BBC World Service. I’m wondering whether you could provide some more insights on live streaming on social media platforms. How is it different from other ways for the campaign to raise publicity for the candidates? We’ve seen candidates like Beto who was pretty much unknown a year ago and now has live streamed his whole campaign —


QUESTION: — and has gained a lot of popularity. Could you provide some comments on that?

MR PORTER: Yeah. I think with any new technological development, it’s tempting to view it as revolutionary and different from anything before, but especially with Beto, this is – Beto is a lot like the profile that Wendy Davis had. If you remember, Wendy Davis ran for the Senate as a Democrat from Texas and she also achieved a good deal of fame in certain media circles but lost pretty badly. So my view of Beto is – again, I’m not in the electoral prognostication business, but I would be surprised if he were able to win because of live streaming, right? Again, to consume a live stream, you need to be really devoted. You need to say, okay, I want to watch what this person is doing right now. I think Beto has done live streams – he’s in his car driving around, right? If you want to watch someone drive around in their car, you are probably already going to vote for them. (Laughter.)

MODERATOR: All right. New York and then the journalist in the fourth row.

QUESTION: Hi, this is Shilong Yang from Xinhua News Agency. There have been so many talk about the social media, digital media. But for us, many of us working in the print media, there is an interest in and critical question: Is the print media a dying industry from your observation of the U.S. print media? Thank you.

MR PORTER: Well, again, I’m not in the prognostication business, but my own view, of course, is that since the election of Trump, it’s actually been pretty good for print media. People in the United States trust print media. They – and certainly they trust it more than they do some website they might randomly come across. So the Times, the Atlantic, these print legacy publications have done quite well. As the media environment has changed so rapidly, there’s real value add to publish a print newspaper or magazine that has a reputation and a long legacy of being trustworthy and sort of having some degree of integrity, and people respond to that in the marketplace.

MODERATOR: All right, in the middle.

QUESTION: Hi. This is Benazir from Pakistan. I’m the digital media editor for ARY News. My question is more about the influence of social media in business. Like, last time in the election the results were first tweeted and they were all over the Twitter. And same like in the elections in July in Pakistan. Advertisers have more interest in advertising during elections. Our – like our livestream, it goes like way – like 20 times higher than the normal. So I see more opportunities in terms of advertisers to advertise for the – like elections. So – and you think that advertisers do not have much opportunities. So don’t you think social media is influencing how advertisers are thinking in terms of spending for websites and livestreamings?

MR PORTER: Do you mean political advertisers or commercial advertisers?

QUESTION: Like, they were – anyone. Like, our revenues get higher, way higher during elections than normal days.

MR PORTER: Oh, sure. Yeah. So the people who are really making a lot of money right now are the cable company owners in the – sorry, the channel owners in Florida, in swing states, right? Swing states right now to advertise on traditional TV right now, the advertising rates are very, very high. I think that’s one way of looking at it. I also think it’s a really good time to be a television company owner right now, because you can charge a lot and both sides are willing to pay.

Remember I said earlier, I used the tug-of-war metaphor to describe, especially at this time of year, how campaigns view the advertising landscape with both sides spending lots and lots of money. The net result of that is probably that very few ads break through, but for the person who’s selling the ads, they’re doing pretty well after that.

MODERATOR: Any more questions? All right. If we have no more questions, I would like to thank our briefer for coming out and thank you all for participating. To find more about our midterm elections programming, visit our website, (Applause.)

MR PORTER: Thank you.

MODERATOR: Actually, it looks like we have a New York.


QUESTION: Sorry, thank you. Grace Hwang from Fuji TV. I wanted to ask you what you thought about the effect of bots on social media. Like, I was reading something where a bunch of bots just kind of magnified a certain issue. They made it seem like there was way more attention on it or way more views from a certain side, way more voices than there actually were, and it led to a prominent figure in media getting fired. So I know you’ve been saying that there’s not really a huge effect that social media has and there won’t be on the elections, perhaps, but do you have any comments on this?


QUESTION: Like, do you think these are just swells that will happen about certain topics but overall not have any impact at all or —

MR PORTER: Right. Great question. I think when it comes to bots, again, this is an example of how the electoral media landscape has changed from this election to the previous election. It’s now clear that bots can shape perceptions on social media. But as a result, companies have really cracked down on bots and now face – bots are taken off platforms. They are eliminated, and the people who proliferate them are penalized. That being said, bots are still concerning insofar as they’re out there and they can shape what people think matters. But again, like, the fact that we all know this indicates that it’s probably less likely to be impactful, the bot phenomenon, than it was in the previous election.

MODERATOR: All right. Thank you, everyone, for coming out. I’d like to remind you that this – our briefer was speaking in a personal capacity and not representative of the American government. Please join us again in the future. Have a great day.

MR PORTER: Thank you.

U.S. Department of State

The Lessons of 1989: Freedom and Our Future