Partisan Filter Bubbles Let the Worst Through

Selective permeability is a property of cellular membranes that only allows certain molecules to enter or exit the cell. This is important for the cell to maintain its internal order irrespective of the changes to the environment. – BiologyDictionary.net

Social media critics have long worried that users’ interactions with platform algorithms sort them into “filter bubbles” — siloes of news, commentary, and conversation concordant with users’ existing worldviews. Others fear that partisans self-sort into similarly polarizing “echo chambers.” Whether the result of users’ choices or algorithmic selection, anything that might complicate a user’s tidy reality never makes it into their feed. By insulating the left and right from each other’s views, partisan bubbles can increase partisanship, polarization, and political animosity.

However, to the extent they exist, partisan bubbles are not impenetrable. While the image of the sealed partisan bubble is intuitively appealing in a polarized world, an alternative view might better explain both what we experience online and why we are increasingly polarized. Our bubbles seem to allow the most extreme content from the other side through. Thus, rather than being walled off from opposing views, partisans observe only the most absurd of the other sides’ beliefs, often stripped of accompanying context. 

To borrow a term from biology, partisan bubbles are a “selectively permeable” barrier. By allowing some outgroup speech through — but only the most radical — selectively permeable bubbles skew the ingroup’s perception of their rivals. The resultant partisan caricatures prevent the recognition of common causes, and even obscure common humanity.  

In a recent Atlantic article titled “Left-Wing Activists Are Bringing Back MAGA Twitter,” journalist Kaitlyn Tiffany follows the rise of PatriotTakes, a twitter account that republishes wild speech from Parler for a liberal audience, and similar organizations:

“The account’s stated mission is to ‘research, monitor, and expose the extremism and radicalization of the far right across the darkest parts of the internet.’ In practice the account shares screenshots with snarky captions, all day and every day, of conversations from social-media sites that an ostensibly mainstream liberal audience might never see. Sometimes they’re of conversations that no one really needs to see.”

By compiling and republishing archives of fringe, right wing speech, PatriotTakes and similar groups help to make partisan bubbles selectively permeable. Adele Stan of the more established Right Wing Watch defends this model, arguing that it provides both entertainment and understanding:

“People either want to mock the right or understand it, and often it’s a combination of both.” She added, “We got a great boost to our own Twitter account when my colleague started embracing a strategy that was really similar, uploading the crazy stuff that we see, clipping it to a digestible length, and directly putting it into a tweet.”

Of course, if you’re viewing the right through a pinhole lens pointed at just the “crazy stuff,” any resultant understanding will be warped. Stan’s comments also illustrate the demand for outrageous right wing content on the left. Right Wing Watch’s twitter followers rewarded them for confirming their worst suspicions of the right. With this incentive in play, it is easy to see how a cottage industry of partisan outrage-finders has emerged. Each provider can gain an advantage over their rivals by highlighting even more extreme outgroup speech, ensuring a race to the bottom. 

NBC’s Brandy Zadrozny called PatriotTakes “incredibly performative,” saying, “It doesn’t shine a light on anything. It just takes videos from one place and puts them in another.” Whether motivated by genuine concert about context collapse or fear of competition, this concern illustrates that PatriotTakes’ audience is motivated by something baser than understanding. 

Make no mistake, this is not a problem exclusive to one side or another. Tiffany’s account just offers an unusually clear illustration of how particular middlemen contribute to selective permeability. @LibsofTikTok serves a similar role on the right, bringing liberal rants from TikTok to a conservative twitter audience. Campus Reform provides a college-centric analog to Right Wing Watch. Broadcast media entities play a part as well, often by magnifying particular outrages discovered by accounts like PatriotTakes rather than initially bringing them to light. 

While selective permeability somewhat complicates the bubble hypothesis, it has a similar intuitive appeal. The desire for a comforting reality that fosters partisan bubbles can also explain their permeability. Rather than recognizing members of rival parties as fellow Americans who hold radically different, perhaps upsetting, beliefs, it might be less taxing to dismiss them as crazy. 

More worryingly, these partisan-misbehaviour-mongers may provide a useful function to their respective political cohorts. To extend the biological comparison, the outgroup speech allowed into the bubble helps to maintain “internal order,” or a united front, in the face of external persuasion and propaganda. When you can view your opponents as dangerous racists pursuing apartheid, or communists hell-bent on the destruction of the family, it becomes easier to maintain internal discipline. It also feeds the most extreme wings of either faction; if the median voter of the opposing party is out to destroy you, perhaps extraordinary action is necessary. 

There is not a simple solution to the selective permeability of partisan bubbles. This phenomenon exists largely because we reward it as media consumers. Titillating examples of the outgroup’s misbehavior grab our attention, but make us miserable. And so perhaps the best course of action is to simply stop rewarding this focus. Individually, we can take care to avoid thinking less of our neighbors because of the archived transgressions of anonymous strangers (or at least recognize when we’re doing so). And whenever possible, we should strive to form personal connections with those different from ourselves, so that our perceptions of them are not reliant on outrage peddlers.


Will Duffield is a guest contributor to the Narratives Project. Will serves as a policy analyst in the Cato Institute’s Center for Representative Government, where he studies speech and internet governance. His research focuses on the web of government regulation and private rules that govern Americans’ speech online.
August 19, 2021