What Causes Political Polarization?
What causes political polarization?
Political polarization has structural causes rather than moral ones. Media deregulation removed friction that once rewarded restraint, cable and digital economics made audience capture more profitable than broad appeal, and algorithmic curation optimized for engagement. Together these changes made emotionally intense partisan content the most rewarded form of political communication.
- Broadcast scarcity once forced outlets to serve overlapping mass audiences.
- The Fairness Doctrine and ownership caps imposed friction that deregulation removed.
- Cable and digital economics reward a loyal narrow audience over a broad one.
- Engagement optimization rewards outrage, novelty, and moral intensity.
- Fewer than 30% of Americans regularly rely on a shared set of mainstream outlets.
The regulatory changes
For several decades, three conditions shaped American broadcast news: spectrum scarcity limited the number of channels, ownership caps limited consolidation, and the Fairness Doctrine required stations to present contrasting views on controversial issues. None of these made journalists virtuous. They made partisan capture economically unattractive. When the Fairness Doctrine ended in 1987 and the Telecommunications Act of 1996 removed the national radio ownership cap, those constraints came off. Clear Channel Communications grew from roughly 40 stations in 1995 to more than 1,200 by 2005.
The economics of narrow audiences
Three broadcast networks reaching more than 90% of households had to avoid alienating large parts of that audience. A cable channel needs a fraction of that reach to be profitable, and a loyal ideological audience is more valuable per viewer than a casual broad one. The same math intensified online, where a creator can be fully funded by a small, highly engaged following. Under these conditions, appealing to everyone is a worse business than appealing intensely to some.
Algorithmic amplification
Recommendation systems do not choose to promote conflict. They promote what holds attention, and conflict holds attention. Vosoughi, Roy and Aral found in Science (2018) that false news on Twitter was about 70% more likely to be retweeted than true news and reached the first 1,500 people roughly six times faster, largely because it was more novel and more emotionally charged. Applied continuously across billions of impressions, this produces feeds in which the most emotionally activating version of any political event is the version most people see.
The result: fragmentation, not just disagreement
Combined, these changes did more than move opinions apart. They removed the shared factual baseline that made disagreement resolvable. Gallup measured trust in mass media falling from roughly 68% in 1972 to 28% in 2025, and fewer than 30% of Americans now regularly rely on a common set of mainstream sources. Two people can follow the news carefully, in good faith, and end up with incompatible accounts of the same week. That outcome is a product of system design, not character.