Two Tribes

By Jonathan Stray

The arrival of social platforms was supposed to liberate us from the monopoly of the media, but data scientist and journalist Jonathan Stray asks if it has simply handed more power back to the old guard?

For more than two decades I’ve watched the transformation from analog to digital media up close. I’ve watched the American media first as an outsider and then an insider, and I’ve been asking the same questions you have. What influence has modern media over our lives and what influence has money over the media? Has the explosion of social media liberated us or just divided us? How do powerful people use media to manipulate and oppress, even if they can no longer censor the truth?

Here is what has changed over the two decades I’ve been watching.

We used to say that freedom of the press is guaranteed only to those who own one. In America, this meant three TV networks and a handful of national newspapers. In Britain the BBC reigned. The distribution monopoly was so iron-clad that pirates built radio transmitters to challenge it.

That’s all over now. Successive generations of social media have rippled through society, each one easier and more popular the last. You used to have to write code to put your words online; now there are a dozen apps that let you broadcast live video from your phone. But somehow, media power is still extraordinarily concentrated. For a very long time I thought the march of technology would “democratise” media. I was naïve. It turns out you can hold an audience without a distribution monopoly, and the news is dominated by more or less the same big players as 20 years ago.

Social media really does make it possible for anyone to make something viral — although that doesn’t mean you’ll go viral twice. Being “in media” means you can can be heard repeatedly. You can get a segment on CNN, or a tweet from The New York Times’ account, and millions of people are guaranteed to see it. It’s not distribution that makes you part of the media, it’s the ability to control attention. And that’s still a scarce resource.

We used to talk about powerful journalists as “gatekeepers,” but that doesn’t quite fit now. Today it’s essentially impossible to stop someone from publishing a story, and we’ve witnessed the rise of a strange and powerful intermediary: the social media platform. I cheered the rise of platforms, starting with Slashdot in 1997, the prototype for Reddit. Anyone could submit a link for consideration, and it would live or die on the votes. This was a pretty great way of bubbling up the good stuff, and many versions of this flourished on the early web. Collaborative filtering, it was called, and it gradually got more sophisticated and opaque.

“It’s not distribution that makes you media, it’s the ability to control attention. And that’s still a scarce resource.”

Then personalisation hit. Rather than recommend the best items, why not recommend the best items for you? Seems ideal for Netflix, but I’m less sure what it means for news. If you’re under 40, you probably see most of your news on Facebook first. You can sorta kinda control what Facebook shows you by who you friend, follow, and like, but not really. In the end, Facebook’s filtering algorithm is proprietary and inscrutable.

An insane amount of media power now rests in the hands of a small number of technology companies, and we only barely understand what they’re doing with it. Journalism has a long tradition of caring about something other than advertising revenues, but these companies don’t come from that tradition. They may or may not be engineering their recommendations in the public interest. And if personalisation shows us only what we like to hear, maybe we’re reduced to living in an echo chamber of like-minded people, the so-called “filter bubble.” I haven’t seen good evidence that personalisation is driving our divisions, but it is true that we are increasingly divided.

Partisanship is on the rise. In America, both voting patterns and survey attitudes have drawn farther apart and clustered more tightly into right and left. You can see this in all sorts of data: voting patterns of both the citizens and senators, responses to the General Social Survey which measures attitudes on social issues, and yes, the widening divide between America’s “red” and “blue” media.

At the same time, Americans have never been so involved in the political process. In 2016 voter turnout was down a few points, but still near historic highs, and the past few years have seen massive street protests on both the left and right. The irony of polarization is that it also drives participation: those who hold stronger views are more likely to be politically active.

I had hoped the internet would lead to greater political participation, and it has. I just didn’t realise that this would also mean greater tribalism. It’s very hard to get momentum behind moderate ideas.

“The media is no longer monolithic. Instead, along with its citizens, it is splitting into two tribes.”

Americans’ trust in their media has been declining for decades, and is now at an all-time low. No one really knows why, though there are lots of theories. Some have suggested that the internet’s destruction of the newspaper business model has triggered a race to the bottom, with desperate journalists writing clickbait to stay alive. That’s a real problem, but it’s not the whole answer: trust was falling for two decades before the web. Others have suggested that quality has fallen, saying that the media make more mistakes and publish more falsehoods than ever before. Yet media research shows that error rates have been stable for decades. Maybe this is all driven by the loss of trust in institutions generally. We wanted to bring down the Man, and we did. So now what?

In 2016, around January, when Trump was just a precocious joke, I set out to find out how much influence the media actually had. I did what might be the simplest comparison between media and influence: mentions versus poll numbers, across the primary candidates for each party. What I found showed that media exposure and primary polling numbers are near perfectly correlated. And this doesn’t account for whether what was said was good or bad — it’s pure attention that matters.

In fact this is nothing new, and political scientists have known for decades about the correlation between coverage and polling in the primaries. But the relationship between media and elections isn’t quite as simple as it seems. First of all, this chart doesn’t show anyone being being persuaded to switch sides; it shows Democratic voters’ preference among democratic candidates and GOP voters’ preference for Republicans.

And if media and popularity are in lockstep, that doesn’t necessarily mean it’s the media that drives votes. Attention and polling are synchronised partially because journalists feel obliged — and incentivised — to cover popular candidates. It’s quite difficult to say whether attention drives popularity or popularity drives attention, even with sophisticated statistical analyses. Media is still quite influential, but it’s more accurate to say that it’s locked into a complex system of interests and ideologies. In the 21st century media power is less about authority, and more about resonance.

“Show the same facts to two different people, and they will walk away believing two different things. This isn’t merely a problem of education, it’s a problem of diverse world views, tribal interests, and the inherent ambiguity of life.”

Thirty years ago, Noam Chomsky and Edward Herman described the “propaganda model” of the mass media, showing how commercial incentives aligned with government control to turn the media into invisible champions of the status quo. It was an incisive critique, because it moved past scheming villains and painted a more realistic picture of how hegemony emerges. But “the media” is no longer monolithic. Instead, along with its citizens, it is splitting into two tribes. Each one responds to news that supports their world view, creating a natural incentive to provide a certain type of news. Our social media platforms further amplify this self selection.

I’ve heard that we are now in an era of “post-truth” politics, but this is a glorious age for facts! Never before has so much basic knowledge been available to all. Ancient scholars dreamed of great libraries, but it was 20th century engineers who finally built the web. We live in a time of plenty, from free encyclopaedias to online courses, to open scientific publications. The problem is, the rationalist dream was all wrong.

Show the same facts to two different people, and they will walk away believing two different things. This isn’t merely a problem of education — it’s not always the case that one person is clearly ‘wrong’ and needs to learn. It’s a problem of diverse world views, tribal interests, and the inherent ambiguity of life.

While the media is still controlled by a small number of large companies, it is also dependent upon us — what we know, what we share, who we trust — and more than ever it’s actually quite easy to get a counter-narrative going. But which narrative?

To understand 21st century media power, first you need to ask a more personal question: why do you believe what you believe?

This is an extract from Weapons of Reason’s fourth issue: Power. Subscribe to the next four issues of Weapons of Reason.

Illustrations by Timo Meyer