Biden calls for regulating big tech, but how is it driving polarization?

President Biden published this op-ed in the Wall Street Journal last Wednesday, calling for increased regulation of tech platforms in several areas: privacy protections, competition, and “extremism and polarization.”

Could regulation reduce polarization? Whether we can write a good law depends on whether we correctly understand what is already happening. A large review of “digital media” (which is much broader than “social media”) suggests that there is some connection to polarization, probably even a causal connection. But old ideas about filter bubbles are being challenged by new models.

New research by Petter Törnberg shows that social media can increase polarization not through less contact between the sides, but more. This happens through a mechanism called partisan sorting, which is when people who identify with one political party tend to align themselves socially, economically, and politically with people who share that same affiliation. It’s not (necessarily) that anyone is getting more extreme, but that all of the axes of the conflict are aligning.

Törnberg’s simulations show that when we have less contact with people far away (left image) our politics tend to align at the local level, preserving diversity at the national scale. But when we have lots of contact with people far away (right image) we can efficiently sort into only two large blocs.

X = one side ● = other side. Color visualizes ideology across 10 different issues.

For example, national sorting is causing Mississippi Republicans to think of themselves as Republicans, rather than Mississippi Republicans. If this is the mechanism behind increases in polarization, then the problem isn’t filter bubbles or echo chambers. This has fundamental implications for how social media might be redesigned for better conflict.

Pairing facts with personal experiences increases tolerance & empathy

This research in political conflict interventions suggests that when political or moral opponents clarify their differing views using direct personal experiences they appear more credible and reasonable than when they use facts alone. For example, someone could defend gun ownership by recounting an experience where they defended themselves from an intruder, rather than emphasizing that guns are used in self-defense (mostly without firing) by millions each year.

This paper recounts four experiments in the United States and Germany, exposing people to journalism articles and social media posts written for this study. It shows that when personal anecdotes are accompanied by facts regarding a contentious political issue, it humanizes political opponents and increases tolerance. This is a useful discovery for both conflict mediation settings and family gatherings!

Red-on-Red criticism on January 6th anniversary day.

Noah Rothman, a seasoned Red commentator, cautions Red not to dismiss the importance of January 6th. He offers a historical parallel: after 9/11, Blue accused President Bush of weaponizing the event. The problem was, most citizens were genuinely worried about further terrorist attacks, and supported Bush’s reaction to 9/11 despite Blue disapproval. Today the nation rightfully regards the Capitol riots with “horror,” yet Red media and politicians have expressed ambivalence toward doing the same.

Noah argues that Red are “contributing to [people’s] anxiety with their too-clever efforts to convince the public to focus on how the attack’s legacy is being weaponized against them. As a tactic, it has backfired.”

Quote of the Week

Digital media do not isolate us from opposing ideas; they throw us into a national political war, in which we are forced to take sides. The suggestion is, in short, that polarization on digital media is driven by conflict rather than isolation.

- Petter Törnberg

Better Conflict Bulletin

Biden calls for regulation of platforms that…