Plain English with Derek Thompson - The Four Biggest Myths About Political Persuasion

Episode Date: August 16, 2024

Today's episode is about how we change our minds—and what political science tells us about the best ways to change the minds of voters. Our guest is David Broockman, a political scientist at the University of California Berkeley, and the coauthor, with Josh Kalla, of a new essay in Slow Boring on Kamala Harris, Donald Trump, and the most persuasive arguments and messages to decide this election. Today, David and I talk about the four biggest myths of political persuasion—and in the process, David will attempt to do something that I’m not entirely sure is possible: He’ll try to change my mind about how persuasion works. If you have questions, observations, or ideas for future episodes, email us at [e-mail]. Host: Derek Thompson Guest: David Broockman Producer: Devon Baroldi Links: "What's Better Than Calling Donald Trump 'Weird'?" https://www.slowboring.com/p/whats-better-than-calling-trump-weird "Consuming cross-cutting media causes learning and moderates attitudes: A field experiment with Fox News viewers" https://osf.io/preprints/osf/jrw26 Learn more about your ad choices. Visit podcastchoices.com/adchoices

In the fall of 2014, a group of hackers pulled off the biggest Hollywood heist of all time. They broke into computer servers belonging to Sony Pictures and released hundreds of thousands of top-secret documents. The attack would cause an international incident, upbent thousands of lives, and changed the movie industry forever. From Spotify and the Ringer podcast network, I'm Brian Raftery, and this is the Hollywood Hack. Listen on the big picture feed, starting August 19th. Today's episode is about how we change our minds and the political science of voters changing their minds during an election. I don't know that I've talked about this before, but I have a very specific theory of persuasion. My theory is that nobody ever changes their mind.

Not ever. That's a strong version of the theory, at least. So maybe I ought to back up and explain where it comes from. In 2019, Matthew Feinberg, a psychologist at the University of Toronto, and Rob Willer, a sociologist in Stanford University, published a paper on the moral foundations theory of persuasion. They said political differences are often fueled by different moral stories, different moral concerns. Conservatives tend to care more about order, safety, patriotism. Progressives tend to care more about equality. justice. And one reason why these groups are so profoundly unpersuasive to each other in debate

is that both liberals and conservatives typically craft arguments based on their own moral convictions rather than the convictions of the people they're trying to persuade. We tend to assume that our way of looking at the world is correct and therefore universal. But values are more like a state currency. They work within certain networks, but they don't always translate across borders. The Canadian dollar works very well in Ontario, but it doesn't facilitate any transaction in Japan. And similarly, progressive values are persuasive among progressives, but they lose their purchase in, say, an evangelical church in rural Oklahoma. It's considered an example here, and I'm stealing from my colleague Olga Kazan, staff writer at The Atlantic, who has written about this several

times, think about the topic of immigration, which clearly divides left and right. If you're a progressive trying to persuade a conservative to join your side, you might be tempted to base the argument on your own moral code. You might say, refugees are a human rights crisis. You might say, a child born one mile south of the Texas-Mexico border has the same rights in eyes of God as a child born one mile north of that border. Justice, equality. These are fine values, but they are disproportionately progressive values, which means they're unlikely to persuade anybody who isn't progressive. To persuade conservatives, you should make arguments on conservative grounds. You could say, our country was founded on the American dream, and that's all these

folks want to do. Work hard for their families. Pay taxes. and build a strong community. You could appeal to family needs. You could say there's a strong work shortage in construction and home health care. If you want more houses for your kids or more home nurses for your parents, we need to expand legal immigration

so that your family can feel taken care of. When I read Moral Foundation's theory, I have, I think, a weird take on it. I don't think this is persuasion at all. I think the subtext of these persuasive efforts is essentially the recognition that people's moral foundations, their deepest opinions, are essentially fixed. If you want to persuade somebody of anything, don't tell them why they're wrong. Tell them why they're right. You're right to want cheaper housing.

You're right to want home health aides to take care of your aging mother. and achieving both might require more immigrants. If you support liberalized immigration law, like I do, you don't have to change your mind. You just have to see the ways in which we agreed all along. In the last few years, I've quietly followed the research on political persuasion, and I've started to wonder whether my pet theory, nobody is ever persuaded of anything, is correct.

Several recent papers have found that political persuasion is actually more common than I assumed, and it doesn't always work at the level of moral foundations theory. Sometimes people change their mind, not only because they interpret familiar facts through new frameworks, but because they really do learn new facts. If this is true, then my cynicism is misplaced. People learn. They change. Their minds can be changed. And this has important implications for thinking about our own minds and the political campaigns trying to change them. Today's guest is David Brockman, a political scientist at the University of California, Berkeley,

and the co-author with Josh Cal of a new essay in Slow Boring on Kamala Harris, Donald Trump, and what political science predicts will be the best arguments and messages to decide this election. Today, David and I talk about the myths of political persuasion, and in the process, David will attempt to do something that I'm not entirely sure is even possible. He's going to try to change my mind. I'm Derek Thompson. This is plain English. David Brockman, welcome to the podcast.

Thanks so much. It's great to be here. Very quickly, and before we dive into the meat of this show, who are you and what do you study? Yeah, so I'm a professor of political science at UC Berkeley, as well as the director of our Center for American Democracy at Berkeley. And I study public opinion and political persuasion in the United States. I want to go over some common assumptions about politics and political persuasion. And I wanted you here on the show because I want you to tell me if these conventional wisdoms are right or wrong. Does that sound okay?

Sounds great. Number one, I want to begin with this very central dynamic or dilemma in the 2024 election, and that is the popular assumption that Democrats should focus on attacking Trump. And I think the steelman of this conventional wisdom is that elections are a cacophony of noise. And if you want to be heard, you should say the same thing over and over and over again. And if you want to defeat your opponent, given the fact that most people are negativity bias, you should remind people over and over again why your opponent is unacceptable. We know from psychology that people are biased toward negativity.

They remember and retain negative messages more than the small and subtle, positive things that happen to them in their life. So if you put this all together, it leads, I think, to a very fair and rational assumption that the most important thing you can do in an election is to stay on message and attack your opponent. from your research, is this true or false? Yeah, well, I don't know that I can tell you definitively, here's what's true and false,

but I can give you, you know, I think our best guess based on the base available evidence and social science theories we have. And based on that evidence and theories, I would say that is likely to be false. And I think the basic underlying social science reason for this is that it's easier to persuade people about things they know less about. Donald Trump has been in the news for 10 years now. It's just hard to imagine someone who has, you know, been hearing about this guy, you know, probably, you know, tens or hundreds of hours of information over the course of the last 10 years, you know, even for people that don't pay much attention to politics.

And all of a sudden, you know, one adjective like, well, have you ever thought about the fact that he's weird, you know, really changing their minds about him? Versus in the context of this election with Kamala Harris, people know much less about her. In some recent surveys we've looked at, you know, a very large. share of voters don't even know what her position is on Social Security. And that's something that those of us who are really interested in politics would take for granted. But the average person who's just busy with their day-to-day life, like, why would they know that? And so it's partly, you know, campaigns responsibility to let them know. But together with my colleague, Josh Calla,

we've kind of applied this in the context of political campaigns in a few ways. There's a few patterns we found that are consistent with this. One is when we look at these experiments campaigns have done over the decades of what seems to be persuading voters in election. One telltale sign of this is that as you get closer to an election, and so people have heard more from the media, heard more from campaigns, the effects of campaigns outreach goes down. And so you can have these big effects on what people think back in August, but by the time you get to November, the effects of each of those pieces of outreach is smaller. That's one pattern consistent with that. Another is some work we did, again with my colleague Josh Call at Yale back in 2020.

when we found in the 2020 election, which has some eerie similarities to this one, we tested actually hundreds of messages about Donald Trump, and none of them worked. But then messages about Joe Biden, who at the time was similarly, like not nearly as well known as Trump, Biden had not been president for the last four years at that point, were actually much more effective, both pro and anti-Biden messages. And recently we had the opportunity to replicate some of that work in 2024. And sure enough, in the context of this election, we find the same thing. that all these attacks on Trump, none of them do anything to make people vote for Kamala Harris.

But telling people more about Kamala Harris, that really changes their view of her and makes them more willing to vote for her. This leads, I think, to a second conventional wisdom, which is that when we're thinking about Kamala Harris, the way to convince voters to vote for Kamala Harris if they're currently undecided is to persuade them that Kamala is, We see this right now, I think, all over the media, lots of people in politics and policy, telling Kamala Harris to pivot to the center, let go of your left-wing positions on Medicare for all or open borders, pivot to the center, embrace moderation, and clearly flag for undecided or persuadable voters that you are a moderate. How do you feel about this second, conventional wisdom, that the idea to win in a general election like the 2024 election

is to convince the modal voter that you're a moderate? I think on the one hand, we have, I think, very clear evidence that when politicians kind of become more extreme on either side, it loses them votes, right? So, you know, you can look at this in just really simple ways like, you know, politicians, you know, Democrats who supported Medicare for All did worse in the 2018 House elections than the those who didn't. That I think is different, though, than the question of, okay, you're in an election, your positions are set. What do you make sure that swing voters know about? So, yes, it's probably good to avoid taking positions on individual issues that people can hit you on. But I think the way

that kind of political elites, political commentators, think about politics is very much in terms of this like left-right ideology, right? Well, I think that part of the thought of like Kamala, quote-unquote, moderating, that I think is good is like, well, she should probably try to avoid taking positions that are unpopular. I think the other version of it, though, is she should take positions like, say, if she stakes out a really centrist position on the border. Maybe that'll be good in terms of insulating her from attacks from Republicans about that she wants open borders or whatever else. But the idea that that's going to send voters a signal that then they're going to infer, oh, well, if she's taking that centrist position on the border, then she must also have centrist positions on this other. issue like Social Security or whatever else. And generally voters and especially swing voters

and swing voters are the people who care less about and know less about politics than just about anybody. They don't really think in those ideological terms. And so I think there's a truth to this in that I do think issues matter a lot. And but I think the way voters think about issues because most voters have this constellation of maybe some centrist views, maybe some conservative views, maybe some liberal views. I think voters, voters are going to be thinking more about, well, how many issues does a candidate agree with me on? Or do I hear about them disagree with me about something specific that I know that doesn't sound right? Versus, well, you know, maybe I don't agree with them on that, but it tells me something that they're broadly moderate.

And I like that. The two points that I want to emphasize from your comment. The first is that swing voters are less ideological than decided voters. And that seems to suggest that they'd be less sensitive to purely. ideological messaging. The second thing you said, which I think is pretty profound, is this idea that people are used to and dependent on processing the world in binaries. You mentioned the left versus right binary, which I think is the way that certainly people like me, political commentators, are used to analyzing candidates? Are you on the left? Are you on the right? Or are you tacking

toward the center. But there are other binaries that voters, maybe especially non-ideological voters, might prefer to analyze the world through. Another somewhat overlapping binary would be moderate versus radical. That's not left versus right. That's almost like a perpendicular axis of, are you closer to the middle or closer to the edge on the left or on the right? And there's another axis that, without getting too crazy on access metaphors, is sort of the three-dimensional access that's cutting through these, which we're hearing about a lot right now, weird versus normal. And it does seem to me like Kamala and obviously Walls are trying to reorient, whether or not they intend to or not, are having the effect of reorienting some voters and certainly the political

media's view of this election through a new lens of normal versus weird, which is a little bit like moderate versus radical, but with less overtly political vocabulary. Do you have, with your expertise and your background here, any way of evaluating whether or not this new axis, weird versus normal, can have any purchase in the 2024 election? I'm a little skeptical. I'm a little skeptical of it. Again, I don't want to say, like, true or false, right or wrong. But my skepticism comes from two places. So, first of all, getting back to

what I said before about just persuadable voters just don't think about politics that much. They don't care about it that much. And it actually takes some political knowledge and then some mental work to hear a political kind of statement and then kind of infer what it means. So, like, on my last point,

for example, in this recent experiment we did where we tested lots of messages in favor of Harris, the messages that we wrote to be kind of send the signal she's moderate versus send the signal she's liberal doesn't actually change how people perceive her overall as moderate versus liberal. People just like don't connect the dots to know that, oh, if you say that you want to be like tough on the border, that means you're moderate, especially swing voters. They just don't know what goes with what. And so I suspect it could be the same sort of way with this weird thing. And consistent with that, back in 2020, when we did this,

kind of prior version of this experiment with testing all these messages about Trump. One of the things we found about the messages in favor of Biden, which did work, is the more specific and less vacuous, if you will, the messages, the better. It's interesting because one more point on this before we move to the third conventional wisdom, the political press loves to give advice. We love nothing more than to give advice. but if the most persuadable voters are the least ideological voters, then ideological political commentators might be the worst possible judges of what sort of messages will persuade,

right? Like, for example, if Kamala Harris presented an advertisement for the op-ed desk of the New York Times and the Washington Post that said, I support social security, you get a lot of slack-jawed responses like, yeah, duh, you're a Democrat. Every Democrat support Social Security. That's the most capital D democratic thing to support is the thing that FDR invented. But in fact, you're saying for a persuadable voter, the idea that you support social security is incredibly, I don't know what the word is motivating, informative might be the right word in a way it's totally uninformative for an op-ed writer.

And maybe just dial it on this point that the group most likely to give advice might be the worst possible judge of what kind of political messages actually work. Yeah. I mean, this I think is kind of a gets back to just like a classic finding in public opinion research that people who are interested in politics and, you know, write about politics, like study politics, right? You know, think about the two of us. We're just, we are actually, if you want to use the phrase weird, like we are

weird ones, right, in terms of just thinking about politics all the time. And I think we will just, you know, never be able to, you know, really properly internalize, like just how little most voters and especially persuadable voters care about politics and how much they often know about it. You know, questions like, which is the more conservative of the two parties? which of the two parties, you know, wants to, you know, restrict access to abortion. There's still plenty of people who don't actually know this stuff. And those voters are the voters who tend to be more persuasive.

Let's move on to number three. The third conventional wisdom, which is maybe just a conventional wisdom that is my own personal theory, is that people don't change their minds. Like, think about it. If somebody asked you at a job interview, when was the last time you asked you, actually changed your mind about something important, you'd probably have to think about that for a while. I think everyone listening to this podcast might have to think about that for a while. And that's because I think persuasion, true persuasion, is very, very, very rare. And for the most part,

nobody changes their mind about anything they consider important. Is that true or false? And do you have any recent research to help us understand why that might be true or false? Yeah, absolutely. So yeah, this is this is one where I think, you know, coming out of like, you know, pop social science over the last few decades, we get these really extreme views. One is like, oh, look, you know, here's this one weird trick where if you say these magic words, then you're going to persuade everyone who doesn't agree with you to start agreeing with you. And on the other hand, I think also extreme statements like, you know, people never change their minds. You can't persuade people. And as usual, I think the truth is probably somewhere in the middle. Same. again with my colleague, Josh Callow, we've done some work. A really fun study we did a few years ago. During the run-up to the 2020 election, we recruited a lot of Fox News viewers to watch CNN instead for a month. And part of what we were trying to get at here is, in fact,

actually, even more pessimistic view than the idea that people don't change their minds, which is a view that actually trying to change people's minds backfires, that if you were to take a Fox News viewer and have them watch CNN, that you're going to kind of just make them dig in more. And so that's part of what we were trying to get out with that. And we actually found that having these Fox News viewers, I mean, these are really conservative folks at baseline on a zero to 100 scale. The rating they gave of CNN was one.

The average ratings of Donald Trump were, you know, in the high 80s, right? Most people giving in ratings in the 90s are, you know, at 100. So it's a very conservative group. Nevertheless, watching CNN for a month, it did change their views. again, it's not most people changing their views on most things, but we saw some meaningful effects on people's knowledge about the world. They, on average, you know, there's a subset of people who went from saying, you know, I love Donald Trump to not I don't like Donald Trump, but, you know, I like the guy,

but he has some flaws. So I would count that as a form of persuasion. And we didn't find any evidence of this kind of backlash. So I think that's the truth here is that, and Alex Koppaget at Yale, you guess at Princeton and others have done some great work on this topic, too, that generally these kinds of persuasive interventions have small effects, but they do have small effects. They don't tend to backfire. And when you think about in a competitive election where, you know, our national elections now are decided by these razor-thid margins, those kind of small

effects can matter. You're reminding me of a paper from several years ago, maybe more than a decade ago, that I believe was called the Fox News effects, which found that depending on how low in the channel dial Fox News was versus MSNBC or CNN, people were more likely to watch Fox News, and that higher likelihood of watching Fox News corresponded to a larger Republican vote share in that district, which suggested, I think, kind of eerily,

that Fox News was so persuasive that if it had a lower-cheworthy, channel number on the dial than CNN or MSNBC, that could have a meaningful effect on the Republican or Democratic representative from that district. One aspect of this, I find very subtle and important about persuasion is the idea that persuasion can happen through the pure channel of learning. That is, let's say I'm a moderate progressive who never watches Fox News. And I'm staying with a cousin or a friend. And because they're obsessed with Fox, I just have to sit there and absorb prime time for a few nights while we have dinner. Fox is going to cover topics where I would otherwise

never learn about those topics and where I don't hold strong views. So, for example, I know that there was a period where Fox was really interested in covering Canadian euthanasia law, and a policy that I know nothing about. If I watch segment after segment about the downsides and not just the downsides, the worst possible stories coming out of new Canadian euthanasia law, all of a sudden, these facts and interpretations about Canadian euthanasia law start to affect the way that I think about Canada and Canadian politics. And from this model, I might infer or derive some theory of Canadian politics that's like, you know, Canada's kind of crazy. I didn't know a whole lot about Canadian politics beforehand, but now that I see what they're doing with euthanasia, I can't trust

anything that they're doing. And so from the seeds of mere information about or mere segments, Fox News segments, about Canadian law, I might infer an ideology about Canada. Is this something like a stylized description of what you found essentially that these little morsels of learning that happen from being forced to watch CNN essentially allowed these views. viewers to spin up new theories about politics. Yeah. So, I mean, in our study, because we're just comparing people who watch CNN to those who stayed with their current Fox News habit, we don't sort of like control, like, what aspect of CNN coverage, right? Like, or, you know, we can't get in their head to know, like, exactly why it persuade them. With that said, I, you know,

our best guess is, is, I think somewhat consistent with what you're saying. First of all, we do find that there is pretty meaningful learning. So, for example, at the time, there was a lot of coverage on CNN of the fact that Kyle Rittenhouse, the 17-year-old who shot and killed two people at the Kenosha protest was a Trump supporter. There was a lot of coverage that Trump had admitted privately that he knew COVID was more deadly than the flu. There was a lot of coverage of just the fact at that time on CNN in the late summer of 2020

that COVID was still killing a lot of Americans. And the problem wasn't over. And people were more likely in our experiment to know all those facts. They were also, when we took them away from Fox, less likely to know the stuff Fox would have told them as well. So we don't know for sure whether or not that's what kind of caused people to change their minds. But at the end of the study, people went back to watching Fox News. They didn't really keep watching CNN as best we can tell.

I think that goes to show that, you know, people will kind of learn information when you expose people to them, when they see messages that, you know, and, you know, informative arguments about particular topics. But then even those people who change, my sense from our work is that you don't see these like follow on effects. You know, just like with the same thing we talked about in Harrison moderation, hearing that Harris is moderate on the border is not going to, in our best guests from our data, lead people to see her as, oh, that must mean she's kind of a moderate on everything. I think it's important to reiterate here that almost automatically, if people are persuadable, that means they can be persuaded back to their original opinion or to some third opinion. So some people in the business of political persuasion think, you know, my job is to take people and persuade them of one thing and then they'll just stay with that position forever. But the very fact that they changed their mind once might make them more likely to change their mind again and again and again. You know, when I think about, I'm trying to apply this to my own life and think about something that I generally follow, but don't really know that much about,

maybe like aspects of working out, like what kind of weightlifting is good for my legs. I don't really research this thing, so I'm basically going to go by whatever the last thing someone told me was. If my personal trainer tells me one thing, I'll say, that's true. If I read an article about the New York Times that says, nope, my personal trainer was lying to me. I'll say, that's true. If I read a Wall Street Journal article the next week that said, The New York Times article was absolutely baloney. I'll say, that's true. So this is me somewhat caring about an issue

and essentially being willing to give over my mind to whatever mildly reasonable take I hear. And that in a way might be a better model of a sort of persuadable voter than this idea that people are just waiting to hear a perfect argument and then change to that position forever. I want to remind people where we are in the show before we continue with the last two

myths or possible myths. Myth number one was that Democrats should focus on attacking Trump. You said no, they should focus on defining Kamala Harris. Number two, I said the way to win is to convince voters that Kamala is moderate, move to the center. You said, no, voters don't think about left, right, and center the same way the political commentators do. You want to anchor people to an idea that Kamala Harris supports policies that they like,

even if that means reminding them of something. that is fairly basic to a political commentator, like she supports Social Security. And number three, I said, people don't really change their minds. And you said, maybe for the most part, they don't, but some of them do. And in a 4848 electorate,

it's incredibly important to get the final 4% of the population on your side. The fourth myth that I want to go over, the fourth conventional- I'm sorry. I'm on one. Absolutely, yeah. Okay.

One thing I want to add to what you were saying that I really agree with is what you're saying about the fact that a lot of persuasion kind of decays very rapidly. That, to me, is the real, like, elephant in the room of all of the research on persuasion is, you know, the political practitioners, the social scientists who study this up, we almost always look at kind of persuasion right away. Like, some news story happens, like, okay, what do the polls look like the next day? Or some of that happens.

Or in, you know, social science, okay, we show someone a message, we ask them a survey question, 10 seconds later. We know when we go back and try to study the long-term impacts of this stuff, the vast majority of it kind of decays. And we don't really understand why. And because sometimes it doesn't. That, I think, is just to call out like a real limitation to all of this and it's something we really need to understand more about it. Let's hit number four. And this is not so much a conventional wisdom, but maybe a plausible inference from everything that you've said so far. And that is that there is a science of political persuasion in the first place. You are a professor of political science.

You study persuasion. You have shared several findings from the literature on political persuasion. And one could infer from all of this that there is a fixed set of known truths about how people change their mind, like a ten commandments of persuasion that we could slap on the wall of every political outfit. in the world at the national level for Senate elections, House elections, state Senate elections. And if campaigns follow these things, then they would be able to persuade people better. Is this true?

I think generally no. When it comes to the work on voter turnout and what turns people out to vote, to vote, I actually think that is true to a decent extent that there's just been now decades of experimentation to refine what works to turn people out to vote. And critically, you find that what turns people out to vote in one election also works in the next election. And it's stuff that would be really boring to you and me, like letting people know, as I was saying earlier, your ballot is secret, whether you vote is public. The most hardest version of this that works the best is actually mailing people and everyone on a block, the voter turnout of them and all their neighbors to say,

your neighbors know whether you vote, and we're going to let your neighbors know whether you voted. that just keeps working again and again and again. When it comes to persuasion, though, I think that, you know, although I think there might be these kind of general principles, like they're sort of obvious, like don't tell voters you disagree with them about important issues. I do think that we find that it's much more variable from election to election. And I kind of analogize this a little bit to, you know, people say, okay, if you're political scientists, tell me about what persuades people.

it's kind of like going to an economist and saying, okay, you're an economist, like, tell me what stocks to buy. And part of why an economist can't tell you that is that there's all these competitive players in the stock market that mean that, like, the price of a stock, like, reflects all that wisdom. And in the same way, I think there's these kind of obvious things that political campaigns could do that because they're obvious, they're already doing them. So, for example, you know, Democrats did the work in 2015 and 2016. and throughout Trump's presidency of informing voters about all of these, you know, terrible facts about Donald Trump. And I think that was probably the right thing to do at the time, right?

That doesn't mean that's the right thing to do forever, right? Now voters know a bunch of that stuff to the extent they can. And so now you have to say something else. It's just interesting to think that turnout is a solvable game and maybe even to a certain extent, a solved game from a strategy standpoint. But what I'm taking from you is that persuasion is not a solvable game, and it will never be a solvable game because it is necessarily dynamic. You were talking about the analogy of asking economists, how do I pick tech stocks? Another analogy would be, how do you score a touchdown in football? Well, the answer in the 1980s might have been, you know, running back counters. All right, defenses adjust. And then,

the answer becomes West Coast offense, short passes, and matriculate the ball down the field, then defenses adjust. And then you say, well, no, it's actually about play action and long passes, and defenses adjust. And then you adjust it. Because, especially in American politics, but certainly in all politics, I just say especially in America, because we have a two-party system, persuasion is always going to be zero-sum, right? That voter is either going to vote for Democrats or Republicans except for the whatever, 3% that vote for Jill Stein and RFK Jr. So it's impossible for persuasion to become a solved game unless one political party somehow deports itself from any possible knowledge of what worked in the previous election.

And that will never happen. So I take your point that to be interested in political persuasion is necessarily to be interested in an infinite game that will never fully be solved and where the strategy might change the optimal strategy might change election cycle after election cycle. Is that fair to say? I think that's right. And I think what that implies, though, is not like throwing up your hands and giving up or just navel-gazing or trusting gurus. It's learning. I think that implies basically like a meta-strategy, right, which is you have to constantly be figuring out what's persuasive in any given moment. So one study that I recently worked on together with a bunch of collaborators,

including Luke Hewitt, Ben Tappen and Alex Koppik. We got some data from this vendor called Swayable, and it's a vendor that Democratic campaigns and Progressive Causes Work with to test their ads. And we looked at their data from their ad testing in 2018 and 2020, all their internal data based on over 500,000 people of all these ad tests that Democratic campaigns had done. And we got to look at all this data and says,

well, what does this say about, like what works in persuasion? And we coded up all these ads for a bunch of hypotheses, right? You know, does it talk about issues? Does it, you know, use a messenger? Does it, you know, is it really pushy, all this kind of stuff? And what we find is that what works, you know, consistent with what you were saying, there is no such thing as the thing that always works in advertising. Different ads are, some ads are much more effective than others, but the ads that, the kind of ads that work well, one year don't work well the next year. For example, in 2018, when Democrats were running following the big Trump tax cuts and the Obamacare repeal vote.

Ads down ballot that talked about issues in 2018, Democratic ads that year did really well. Then in 2020, we find that the ads in 2020 down ballot that talked about issues actually did less well than average. So the question, like, should politicians talk about issues in their ads? Like, it doesn't have an answer. It depends on, well, like, what issues are you talking about? And so I think that's just one example of the broader phenomenon that, yes, the context is changing as you're saying what your opponents are changing, what voters already know is changing based on what you and your opponents have said before. And so I think what that means is not that you give up, but you have to constantly be gathering data and think about building a system, you know, these systems like Swable, for example, to try to figure out what's working at any given time.

I think that's kind of the frontier of where campaigns are right now. I'm so glad that we got there in this interview because this is really going to change the way that I write and think about political persuasion. I have historically written about political persuasion by looking at history, by saying here's what worked in 1996. Here's what worked in 2004. And I'm just now realizing that's a bit like saying, here's what worked for NFL offenses in 1982. Therefore, this is what the San Francisco 49ers should try to do in 2024. That doesn't make any sense. You'd be. laughed out of the room or the ESPN studio if you said the 49ers should do something this year because something kind of like it worked 42 years ago. But the same principle somewhat holds for political persuasion in that if what we're talking about is a dynamic equilibrium, just because something worked a few cycles ago doesn't mean we should have a tremendous amount of confidence that it will necessarily work in a new electorate, often with new voters, and of course with new candidates.

Very last question for you, in the open to this episode, I talked about my fondness for moral foundations theory and how I have become fond or at least habituated at saying that I think no one is actually persuaded.

Nobody changes their mind if you want to get somebody to agree with you, your goal shouldn't be to tell them that they're wrong. It should be to tell them that they're right, but for different reasons than they intuited.

So for example, you're talking to a conservative there against immigration, rather than say, I'm a progressive, you should support immigration because refugees are people to, and equality, equality, equity, and justice, you should say, you know how in your area, there's a shortage of housing. Well, 25% of construction workers are immigrants,

if there are more immigrants, you would have more homes. You know how you have a mother who needs to be taken care of, and right now there's a shortage home health workers. Well, if there were more home health workers, your mother could be taking care of, and it would cost you less, potentially. So you can frame, essentially, the issue of immigration away from equality and more toward a sense of family security, economic selfishness, a sense of patriotism. I wonder what you make of moral foundations theory, because given everything that you just said,

maybe moral foundations theory is like the dink and dunk west coast offense of the 1990s, nearly 2000s. It worked for a while. It works in some cases. It works for some offenses in some context. But it's not the answer, capital T, capital A for political persuasion. How do you think about moral foundations theory? Yeah. I mean, I like moral foundations theory a lot. But at the same time, I think you're right that the standard for what makes a theory useful, right, can't be that it explains like 100% of the cases. You know, social science is not like physics where we can say, you know, I'm not a physicist. Maybe this isn't even true that like, you know, every electron will behave the same way or something

like that, right? And then, you know, all electrons are made the same. You know, social science is not like that, right? Not every person, not every context is made the same. So at the one hand, I really like Moral Foundation's theory. I teach it in my classes. I think it's a great framework through which to kind of understand part of the world. At the same time, it's an impossible standard to think that's going to be the best thing to do every time, right?

In part for the reasons we talked about earlier that maybe having a campaign where the first three ads use moral foundations theory is going to be great. But then you'll have picked off the voters that could be persuaded that way. And there's going to be other voters who could be persuaded with some other way, right? So I don't think a theory has to bat a thousand to be useful. Same thing with these kind of historical analogies, right? You don't have to like slavishly, you know, do exactly what a historical campaign does. But at the same time, it might, like that knowledge of history probably is helpful, right? And so I think, again, in so many response to these myths, right, there's a healthy middle ground.

David Brockman, thank you very much. Sure, my pleasure. Thank you for listening. Today's episode was produced by Devin Baraldi. our summer schedule for plain English for the next few weeks will be one episode a week on Fridays. We'll see you next week.