Talking politics
How do I get out of my political echo chamber?
By the Agora desk · Published · How we count player splits
Add thoughtful voices from the other side, read the strongest version of their argument, and talk to real people rather than only reading about them. Research suggests most people's news diets are more mixed than the echo-chamber idea implies, so the bigger trap is often a distorted picture of what the other side actually believes.
In short
- Only a small minority are in a true news echo chamber; the US has more than most countries.
- The bigger bubble is in your head: people overestimate how extreme their opponents are.
- Following the loudest voices on the other side can harden your views, so choose quality over outrage.
- Try to state the other side's best argument so well that they'd agree with your summary.
- Explaining how a policy works, step by step, is a quick test of how well you really understand it.
How big is the echo chamber problem?
Smaller than you might think, at least for news. A 2022 review of the research by the Reuters Institute at the University of Oxford found that in the UK only about 6 to 8% of people are in a partisan news echo chamber. The US is the main outlier: in one comparison of seven countries, it was the only one where more than 10% of people relied only on partisan news.
The same review found that search engines and social media generally lead people to slightly more varied news than they would choose themselves, the opposite of the "filter bubble" idea, because algorithms throw up things you weren't looking for. The people who seal themselves off are mostly a small, highly partisan and highly engaged minority.
So if you're asking the question, you're probably not trapped. But your picture of the other side may still be badly off.
The real bubble is in your picture of the other side
More in Common's 2019 "Perception Gap" study found that Democrats and Republicans each imagine almost twice as many of their opponents hold extreme views as really do. People who consumed the most news, and people who posted about politics on social media, had the most distorted pictures. The politically disengaged were about three times more accurate than the most committed partisans.
That points to a different problem than the one most people worry about. You may well see the other side's views every day, but mostly through the most extreme examples: the clip that went viral, the post your side is mocking. That's not exposure; it's a highlight reel of the worst moments.
Why following the loudest opponents can backfire
In a 2018 experiment led by sociologist Christopher Bail, Republicans on Twitter who agreed to follow a bot retweeting liberal accounts for a month became measurably more conservative. Democrats who followed a conservative bot moved slightly more liberal, though that shift wasn't statistically significant.
And a large 2023 Facebook experiment found that cutting how much like-minded content people saw by about a third didn't change how polarised they were. Together, these studies suggest that simply adjusting the volume of content you see isn't the answer. What you see, and how you engage with it, matters more.
A one-month plan to widen your view
Treat this like a diet change: small, deliberate, and focused on quality.
- Week 1, audit: list where your political news actually comes from (apps, accounts, podcasts, group chats). Bias-rating tools such as AllSides, Ad Fontes Media or Ground News can help you see the lean of each, as a starting point rather than the final word.
- Week 2, add: follow two or three serious writers or outlets from a different perspective. Look for people who explain and argue, not people who dunk. A useful test: would someone on their own side respect them?
- Week 3, compare: take one big story and read it in two or three outlets that lean differently. Note what each one leads with, leaves out, and calls things. Then find the original document, data or video if you can.
- Week 4, talk: have one real conversation with someone who sees an issue differently, aiming only to understand. Our guide to talking politics with family has questions to ask.
- Throughout: when you feel a jolt of outrage at a post, pause before sharing. That jolt is often a sign you're seeing a highlight reel.
Test your own understanding
Two quick tests show whether you've really left the echo chamber. First, try to steelman: write down the strongest case for a view you disagree with, in terms its supporters would recognise. If you can't, you've found a gap. Agora's issue guides, such as tariffs explained, set out each side's case in its supporters' own words before asking what you think.
Second, explain the mechanics. In a 2013 study by psychologist Philip Fernbach and colleagues, people who were asked to explain in detail how a policy would work (not just why they liked it) realised they understood it less well than they thought, and their views became less extreme. This happened among Democrats, Republicans and independents. Try it on your own strongest opinion: how, step by step, would it actually work?
Signs you're getting out
You don't need to change your mind to have escaped an echo chamber. You know it's working when:
- You can explain the other side's best argument without sneering.
- News occasionally surprises you, or complicates something you believed.
- You can name an issue where your own side's case is weaker.
- You know, and like, at least one person who votes differently.
- You notice when a story is designed to make you angry.
Common questions
Are algorithms trapping me in a filter bubble?
Mostly not, according to the research reviewed by the Reuters Institute: search and social media tend to give people slightly more varied news than they'd pick themselves. The bigger effect is self-selection by a small, very partisan minority, and the tendency of viral content to show the most extreme examples of the other side.
Are AllSides and Ground News unbiased?
They're tools, not referees. They label outlets by lean and put coverage side by side, which is useful for spotting differences in framing. But their ratings involve judgement calls, and an outlet's lean says nothing about whether a particular story is accurate. See how to spot media bias.
Do I have to give every view equal weight?
No. Escaping an echo chamber means understanding other views, not treating every claim as equally true. Questions of fact can be checked against evidence; questions of values (what matters most, what's fair) are where reasonable people genuinely differ.
Where do I find good-faith people on the other side?
Look for writers who engage with opponents' best arguments, long-form podcasts or interviews rather than clips, local groups and community meetings, and organised cross-partisan conversations. Offline, ordinary neighbours and co-workers are often more representative than anyone online.
Sources and further reading
- Reuters Institute (2022): Echo chambers, filter bubbles, and polarisation: a literature review
- More in Common (2019): The Perception Gap
- Bail et al. (2018), Exposure to opposing views on social media can increase political polarization, PNAS
- Nyhan et al. (2023), Like-minded sources on Facebook are prevalent but not polarizing, Nature
- Association for Psychological Science: Extreme political attitudes may stem from an illusion of understanding (Fernbach et al., 2013)
- Ground News: compare how outlets cover the same story (a tool, not an arbiter)
Read next
- Why is politics so polarised?What polarisation means, the causes researchers point to (sorting, media, geography, primaries, social media) and where the evidence is still mixed.
- How can I tell if a news story is biased?Bias hides in story choice, sources, wording and what's left out. A practical checklist for spotting it, and what bias charts can and can't tell you.
- Is there such a thing as unbiased news? Where to look insteadNo outlet is perfectly neutral, but some try hard. Wire services, public broadcasters, bias-rating tools and how to build a balanced news diet.
Put it into practice: latest stories
Written by the Agora desk from the sources above. See our editorial standards. More Learn articles.
