A New Year's (Media) Diet / How Can We Heal?

Americans are growing apart and have been for several decades now. We can sense this in our daily lives—the awkward Thanksgivings, the partisan segregation in our communities, the vitriol on social media.

Political scientists measure division using something called “affective polarization.” This is usually done with a ‘feeling thermometer” question. A survey respondent would assess how they feel about the Democratic Party on a thermometer scale of 0 to 100, and then how they feel about the Republican Party. The absolute value of the difference between those two numbers is the polarization level for that individual. If I hate the Democrats (10 out of 100) but love the Republicans (90 out of 100), my score would be an 80, which is quite high. You can average these values at the population level, like across all American voters, to get a rough sense of polarization within the country.

This index has some interesting properties. If a person loathes both parties—say 20 out of 100 for both the Democrats and Republicans— they would have a polarization score of zero. Ditto for if they love both parties (though almost nobody does). A person’s polarization score can go up if they come to like their own party more (“in-group love”), dislike the opposing party more (“out-group” hate), or any combination of those two things.

The general story of the American political system that affective polarization has been on the rise for some time, and it is primarily driven by a rise in outgroup hate. The graph below, created by political scientist Sean Westwood using longitudinal data from the American National Election Studies (ANES) project, shows this trend. Since the 1980s, our feelings towards our own parties have stayed roughly the same—about 70 degrees on the thermometer scale. Feelings towards the outgroup have plummeted from about 50 degrees in 1980—almost neutral—to a chilly 20-25 degrees by 2020. We increasingly loathe and distrust people from the other side.

This shift seems to be driven in large part by changes in the media environment. Decades ago, we all consumed the same news. This shared set of facts grounded the conversation and helped facilitate understanding across groups of Americans. With the rise of cable news, Americans came to consume facts that they mostly agreed with, and each side of the political aisle came to have its own set of facts.

This was in turn compounded by the rise of the internet and social media platforms like Facebook, Twitter, Instagram YouTube, TikTok, Twitch, Substack, and all the others, as algorithms feed users content that they are inclined to click on, watch, and agree with. We are increasingly stuck in echo chambers of our own making, and it only seems to be getting worse.

Some degree of affective polarization is normal, healthy and expected, but extremely high levels of polarization are probably not a good sign for the political system. Polarization may preclude opportunities for policy cooperation and weaken support for democracy itself, paving the way for authoritarian political movements. This review by Jamie Druckman is helpful for those interested in learning more.

I’ve got some new data on polarization from a project with Kevin Arceneaux on the 2024 election that I’ll hopefully be sharing in a few weeks. One preliminary result we have is that polarization has fallen substantially since the election—from 60 to about 50 or so— but it remains quite high in general, even higher than it was in 2020. Which begs the question—how can we heal?

A good start would be thinking more about what we consume. Think of your media use as a diet. A good diet is balanced and diverse. Getting your news from the same one or two networks, newspapers, or websites every day, or even worse, the same voices from those places, is the dietary equivalent of eating from only one food group. While it might taste good going down, it won’t be good for you in the long run.

Recent research by the David Broockman and Joshua Kalla—two of the best political scientists in the game— shows how sustained consumption of “cross-cutting media” leads voters to moderate their attitudes. They recruited a group of Fox News viewers and incentivized them to instead watch CNN for a month. They found that these folks are indeed persuadable and malleable, and the experience caused substantial learning and moderation of views. I suspect that if they had done the same experiment in reverse—recruiting CNN viewers to watch FOX—they would have found the same effect.

It’s a strange time to be telling you to consume “cross-cutting media” when Fox news has people on literally making the case for going after Greenland. The counterargument— why should I consume news from the other side, when we think the other side’s stuff is just propaganda and disinformation? Maybe we should be polarized? Maybe I like being polarized! After all, it’s cognitively easier.

At a minimum, consuming counter-attitudinal news might make you a stronger thinker, allowing you to understand the arguments made by the other side. It exposes you to new information. You may be more reified in your own thinking afterwards, or you may concede that some of the points made are valid. You will cut through the filter of your own partisan bubble and hear counterarguments straight from the source.

Anne Marie-Slaughter, the CEO at New America, had a nice a LinkedIn post to this effect about her resolution for 2025. I’ll leave you with her thoughts:

Whenever I am talking to someone with whom I disagree on a subject that is important to me, I am going to try to shift my response from objection to curiosity. I'll ask myself WHY I am disagreeing. Do I dislike the speaker? Am I feeling threatened in some way? If so, why? Do I have other emotions toward the speaker that are shaping how I am reacting?

If I'm really responding to the content, what assumptions am I making about what the speaker means? Are those assumptions correct? How can I be sure? Are there assumptions that the speaker is making about me or my views that is shaping their position? How can I find that out?

I can ask questions, of myself and the speaker. Questions that are not rhetorical but genuine, asked in the spirit trying to understand. I can also ask whether it is possible that my position and the speaker's position are both true and both false, at least in part.

I resolve to take this approach in conversations of many different kinds - with family and friends, colleagues, people I meet for the first time. It may be that once I have asked those questions and heard the answers, I will feel the same way I did at the outset. But I may also have learned something and changed the nature of the interaction. The more I am certain that I am right, the more I will try to shift from certainty to curiosity, and perhaps from there to at least a measure of uncertainly, enough to learn and change and grow.

Thanks for reading, and please do share/restack/email/comment/post/all that if you find it useful.

And for fun, leave your affective polarization score in the comments.

Rory