Adam Berinsky is a former RSF visiting scholar and the author of Political Rumors: Why We Accept Misinformation and How to Fight It. In Political Rumors Berinsky examines why political rumors exist and persist despite their unsubstantiated and refuted claims, who is most likely to believe them, and how to combat them. In a new interview, Berinsky discusses his findings. The interview has been edited for length and clarity.
Adam Berinsky is the Mitsui Professor of Political Science at the Massachusetts Institute of Technology.
Q. What motivated you to write Political Rumors? What makes misinformation a political rumor? Why is it important to study them?
I became interested in political rumors in the summer of 2009. I’ve studied American politics since the 1990s and I’ve seen a lot of crazy things. But even by those standards what was going on in 2009 was especially crazy. So, I became interested in rumors. I was looking at rumors about President Obama's birthplace – whether was he born in the U.S. I was also looking at rumors about “death panels.” Did the Affordable Care Act, which then was in front of Congress, have provisions that would create so-called death panels that would decide whether or not to provide care? I thought of them as a something of a lark. Like, oh, this should be kind of interesting to study. Back then people asked me, “Is this political science?” and I said, “I think it is.”
But over time, misinformation became a bigger issue. And it's something that really took off. In 2009. Political rumors and political misinformation are unfounded claims that are spread through society. It’s important to know that people aren’t necessarily doing this on purpose, but it can be very damaging for democracy.
Q. There has been a large focus on political misinformation in recent years. Are political rumors new? How has misinformation and rumor transmission changed over time? What has stayed the same?
There have been political rumors back to the beginning, even before the founding of the United States. And it’s not just in the U.S. I think that’s an important thing to note. This is really a global phenomenon. I study American politics, so that’s where my focus is, but I give talks to different groups around the world. I talk to folks from Finland and Spain, and they say, “Well, you think your stories are crazy, let me tell you what's spreading here.” So, this is something I think it's just part of the human condition. People like to gossip; people like to like to talk about crazy things. I think this has always been there.
I think what is different, though, is the means of transmission. I grew up in New York in the 1970s and ‘80s. There used to be pamphlets and leaflets posted up on the side of construction sites, saying things like Mondale was in cahoots with Reagan during the 1984 election. This is how things spread. I'm old enough to know what a mimeograph is. People would send mimeographs – newsletters – to each other to spread those rumors. Now with the internet, the transmission has become a lot easier. It's easier for people to spread information. But more importantly, it's easier for people to find like-minded folks on the internet. So, you don't need to go to a JFK assassination convention to find other people who are interested in the intricacies of the magic bullet. These are things that you can find online.
Another thing is that the internet has really increased the potential spread of rumors beyond just the people who are active believers, to infect the information ecosystem. What I mean by this is, say I don't necessarily believe a rumor, and I don't have interest in it, I can still watch other people talking about it online. And if I don't pay a lot of attention to politics, it might be that my exposure to political rhetoric is these kind of conversations between friends, relatives, or they could even be people I don't know. Watching these conversations on social media, I think, allows the spread of these rumors and misinformation beyond just the core believers.
Q. You identify four groups to describe how citizens engage with rumors. What are they? And what role do they play in rumor transmission?
When we think about rumors, we think there are people who believe them and people who don’t. But there’s a little bit more to it than that. At one level, there are creators – people who are motivated to create misinformation. Now, it’s very important to look at the political motivations and the economic motivations of these folks, the creators. However, I study the mass politics side of things, so creators are outside the system of what I look at in my book. I think it's really important when we focus on the mass public to think not just that there are the believers, people who say that they accept these rumors, and the disbelievers, the people who reject these rumors. There are also folks in between – the uncertain. These are the people when you ask, “Do you think that Obama was born in Kenya?” They won't say, “Yes, I think it's true.” They won't say, “No, that's not true.” They might just say, “Well, I'm not sure.”
I've done some surveys where I follow up with these folks. And I ask, “What is it that you heard that led you to the beliefs that you have?” And so, for the people who believe this rumor, they might say, “Well, you know, I saw on this website that the birth certificate in Hawaii is fake.” People who reject the rumor will say, “I've seen these charges, they're ludicrous.” But people in the middle say, “You know, I'm not really sure, I hear some things.” And essentially, “Where there's smoke, there's fire.” The people in the middle are people who are not quite ready to accept rumors, so that's a good thing. But they're not ready to reject them either.
In the book, I really focus a lot on these groups. If you ask people the questions I asked in my book – Do you think Obama is a US citizen and therefore eligible to be president? Do you think that George Bush and Dick Cheney were somehow involved in 9/11? – depending on the question about 20 to 30%, sometimes even 35% of people say they don't know. And it’s really important to focus on these folks because we're aiming for not just minimizing rumor acceptance but really maximizing rumor rejection. And if we don't account for these people who say they don't know, these uncertain folks, we're basically creating uncertainty in the political system. To give an example, thinking about Donald Trump's claims about voter fraud in the 2020 election. It’s good that only a small portion of people accept these rumors. But there's also a small portion of people who reject the rumors. And there's a lot of people who said, “I'm not really sure.” And if it comes to faith in the political system, and you ask people, “Was the 2020 election stolen?” And they say, “Well, I'm not sure.” That's a really bad outcome for democracy.
Q. What makes someone susceptible to believing political rumors? Do “believers” really believe political rumors? Or are they just trying to express general displeasure with a political party?
When I started my work, I wanted to know what the different drivers of rumor belief were. And I found is that there are two dimensions. One dimension is general belief in fanciful stories. So, think about at Thanksgiving. You have a relative who always has a crazy story about something or the other. You try to avoid them. “Just don’t mention politics around Uncle Joe,” right? That’s one end. And then you have people on the other end who reject all sorts of rumors. So, there’s just a general tendency that people are more or less likely to accept rumors and misinformation.
But there’s also a political element. Democrats are more likely to reject rumors about Democrats and Republicans are more likely to reject rumors about Republicans. And what I think is really interesting, which I talk about in the book, is that there’s an interaction here. So, it’s not that all Republicans think that Obama wasn’t born in U.S., but Republicans who score high in these kinds of predispositions, or who are predisposed to believe rumors, are going to believe that. It’s the same thing with Democrats. So, there’s that interaction there. And I think it’s important when we think about solutions, to recognize that there’s some personality component, but there’s also a political component. And so, when we think about combating misinformation, we need to recognize that, in part, this is driven by politics. So, the solution, in part, needs to recognize and account for the political scene.
Then there’s also this question of whether or not people really believe this. When I started this project, people weren’t really sure it was political science. They were saying, “Oh this is all very interesting, but people aren’t telling you what they really believe. They’re telling you things that are going to make their political opponents look bad.” So, “Was Obama born in Kenya?” “Yes.” “Is Obama a Martian?” “Yes.” Right? And so, I spent some time early on trying to demonstrate to other political scientists, and social scientists more generally, that people really believe these things.
So, I wrote a paper where I wanted to demonstrate that, yes, we really should believe people. And when they say they believe rumors, they really do believe rumors. So, in a study I did things where I would basically try to give people incentives to not support rumors. So, if I’m looking at belief in the rumor about Obama not being born in the U.S., before I start, I can say, “Look, you know, people have lots of beliefs about different things. They also have feelings towards politicians. Tell me what you think about Obama.” And then they would say, “Oh, he’s terrible. I hate him.” And then I’d say, “Okay, now we’re going to move to a different kind of question on beliefs about whether he was born in the U.S. – factual beliefs about him.” And so, I would say, “Regardless of how you feel about him, where are your beliefs on where he was born?” So, essentially, asking people to set aside their negative feelings, to tell me what their beliefs on his citizenship were. In other domains, this has been shown to reduce having negative beliefs about the facts concerning whatever the target is. In this case, Barack Obama. And what I found is that it didn't move their beliefs at all.
I also did studies where I essentially paid people to not believe a rumor about Obama. The way I would do this is I would do surveys through market research companies. And the magic word for those surveys is “full participations points.” You’re going to get fully compensated for your time after you’ve fully participated. So, I would say to people who were inclined to believe rumors about Obama, so in this case Republicans, “I’m going to ask you some follow-up questions about that. That’ll take about five minutes or so. Otherwise, we can go right to the end of the survey.” Essentially, if you want to tell me why this is the case, you’re going to have to pay a cost; you’re not getting your participation points until you answer all these questions for me. So, I’m getting at whether these are people’s true beliefs by triangulating through these different methods. And what I found is that there are some people who probably are cheap talk, right? I don’t want to say everyone who says that Obama was born in Kenya genuinely believes it in the depths of their soul, but most people really do say what they believe.
Q. What role do political elites play in the dissemination of misinformation?
I think it’s really important to think about the role of elites. A lot of the work in political science focuses on mass behavior, like the kind of work that I do. And the conclusion that could be drawn is that other people are just idiots. “Why do they believe this? Because this is clearly ludicrous.” I talk about this more explicitly in the book, but we know a couple of things about public opinion. First is that most people, most of the time, don’t pay attention to politics. However, they’re happy to answer survey questions. And they’re also happy, when it comes time to vote, to vote on particular issues. So, what most people do, they don’t totally cede their authority, but they take heed of what political leaders they like say. So, if I’m someone who pays casual attention to politics, and I know I like Donald Trump, when it comes to questions about voter fraud, am I going to do the deep research myself? Probably not. But I’m going to listen to what Trump is saying about that. And not just Trump, but other Republican politicians. And, I should say, this is true not just of Republican leaders and Republican elites, but Democratic leaders and Democratic elites as well. No party has the market on following their leaders.
This is something that is rational for folks to do. But what that means is that we’re putting a lot of faith in our leaders to guide us the right way. It makes sense, as someone who is busy with their life, their family, their hobbies, their interests. It wouldn’t make sense to spend a lot of time thinking about politics. And it makes more to sense to figure out which kind of team I’m on, which party best represents my interests, which leaders articulate the kinds of things I find compelling and listening to them. And this works fine as long as the leaders are responsible. Thinking about the current day and age, if a big source of misinformation – not the generation of misinformation, but the spread of misinformation – in this case, I’ve been talking about voter fraud, is coming from elites, then we can’t really blame the citizens for listening to their leaders. We should be blaming the leaders for providing messages that are unsupported by evidence, because that’s what the mass public is going to take up.
So, in one regard, my book could be read as saying, “Look at all this crazy stuff that people believe.” But on the other hand, we need to ask, “But why do they believe it?” It’s because they’re trusting the wrong people. They’re following political leaders who should be leading and should be making calculations that are based on what is good for society. But we’re in a catch-22 because their political interests, their electoral interests, are served by creating confusion. So, Trump loses the 2020 election, and by all available evidence, it was a free and fair election that he did lose. But it’s in his interest to create confusion and sow discord. It’s also in the Republican Party’s interest to sow discord and have misinformation. It’s not that this is malicious, but it’s motivated by political interest. And so that’s the tricky conundrum there. The solution is better elite discourse; it’s having responsible leaders. But if leaders benefit from discourse based on misinformation, then that can be bad. Liz Cheney is an example of a case where you have leaders that stand up to it. She is a Republican who essentially said, “This is crazy. There was no voter fraud.” And she was run out of her leadership position and then eventually run out of Congress because of the positions she took.
Q. It has been found that individuals who are more politically engaged are more likely to reject rumors. Is the solution to political misinformation more information?
I’m not saying information is a bad thing, but, that said, I think this gets back to this question about leaders. So, who knows more about politics? People who are more engaged with the political system. So, I can ask you questions like, “Who’s the President?” “What’s the name of the Chief Justice of the Supreme Court?” I don’t care that you can answer these particular trivia questions, but that you can shows use that you’re reading newspapers and you’re engaged with politics. We find that, in general, the more attention you pay to politics, the more you get these questions right. And, in the case of political rumors, the more likely you are to reject these rumors.
Now, this works well if leaders are giving messages that are in line with the truth. But what I find is that, observationally, the more attention Democrats pay to politics the more likely they are to reject rumors about both Republican politicians and Democratic politicians. The more attention that Republicans pay to politics – again, these are Republicans in the mass public – the more likely they are to reject rumors about Republicans. It has no impact on their rejection of rumors about Democrats. So, why is this? It’s because if you look at the political rhetoric, you see that Republican politicians are saying, “Here’s some crazy stories that may or may not be true!” They don’t even need to endorse it. They don’t need to say that the election was stolen. They say, “Was the election stolen? I don’t know.” Just asking questions creates the uncertainty.
So, this makes a trickier proposition. Because if the solution was people who pay more attention politics reject rumors of all sorts, then we say pay more attention to politics. But if there’s a split in how politicians are talking about these political rumors, that’s going to be reflected in the answers that people give. So, more information, in and of itself, is not necessarily a bad thing. But the quality of that information, I think, plays a really important role.
Q. What do you believe is the best strategy to combat misinformation? Are there any limitations to that strategy?
I started writing this book in 2009. That was two sabbaticals ago. Most of my projects to that point had been five-year projects. I’m a political scientist but trained like a social psychologist, so if I want to answer questions of interest to me, I run surveys to figure out what people think. I run experiments to see if I can change the way they think. So, with this project, I thought, I’ll run some surveys, I’ll come up with some solutions, and then I’ll be done. Sort of a four-year project. So, I signed a book contract in 2012 saying I would finish this book in 2014. 2014 came and went. 2020 came and went. I eventually finished in 2023.
The survey part was straightforward, figuring out who believes rumors and why. But coming up with solutions was a lot harder. Now, I could feel bad about that, but I realize it’s not just me. After 2016, there was an explosion of work about political misinformation, and misinformation more generally, in the social sciences. And, I will say, that no one’s come up with the perfection solution. There’s been a lot of searches for the silver bullet that’s going to slay misinformation.
I have my own solution. In the book I talk about how it's as important who delivers the message as what that message is. For example, in the book, I talk about if I want to debunk rumors about death panels, I want to have good messages. But it's important to have those messages delivered by people with credibility. I find that Republican politicians – so, people speaking against their own apparent interests – are the best medium to deliver that message.
So, is that the magic solution? No, because it doesn’t always work. Other people have extended my work into the areas of voter fraud and environmental policy and have found it works well there. I've done other studies on healthcare and presidential elections where things are more mixed. So, I have a solution that works sometimes, and doesn't have huge effects. But that's my preferred solution. But there's lots of other people who are working in the space looking at other solutions. What's the best way to fact check? Is it better to directly contradict misinformation? Or to do it indirectly? Should we give people training before they encounter misinformation, so they know what to watch out for? Or do we want to do it while they encounter misinformation? There are lots of different strategies.
What I would say is, we shouldn't be looking for the ideal solution. In fact, I think that that's created problems and not just with my research timeline and publication date. I think, in general, coming up with a solution has been very tricky because we're looking for the solution. And I think instead we want to recognize that there are lots of different solutions and thinking about how these might complement each other. What I've been thinking about, in my time here at Russell Sage, is coming up with a framework where we can combine different interventions to think about things that might work together. So, it could be that there are two separate interventions that each reduce belief in misinformation by five or six percent. That’s a believable effect, that could be good. What happens if I combine them? It could be that they’re substitutes for each other; it could be that they work in somewhat the same way. So, that says it doesn’t really matter what I do.
But what I’m really hoping to do is find interventions that can work together. The analogy I use in the book is the Swiss cheese metaphor. So, if you have a block of Swiss cheese, and you slice it, there are holes there. Let’s say you turn slices so that the holes don’t necessarily line up. The idea is that each of these solutions is imperfect, but perhaps the shortcomings of one will be covered by another. People talk about a Swiss cheese model of pandemic response to COVID. We can think about a Swiss cheese model in response to misinformation. We can think about different ways that interventions could combine and overcome the shortcomings of each other. In an ideal world, maybe they could even build on each other. So, it could be that one type of intervention gives you the skills you need to identify particular pieces of misinformation, and another gives you the skills that you need to integrate the knowledge you’ve gained from that into your beliefs. It could be that, individually, these interventions work okay, but together it could be a really good thing. So, that’s what I’m looking for. There’s not a simple straightforward solution, again, that’s why it took me, like, 14 years to finally finish the book. But I think that this is something that is potentially promising.