False Information Spreads Faster and Further Than Truth
Vosoughi, Roy & Aral — MIT (2018) · The spread of true and false news online (Science, 2018) (2018)
False news travels 6x faster than truth and is 70% more likely to be shared. A 2018 MIT study published in Science analyzed 126,000 news cascades spread by 3 million people. The driver is novelty — false news feels more surprising, and humans share what surprises them. Bots aren't the primary problem; people are.
Core Concepts
The Problem
The assumption is that misinformation spreads because of bots, bad actors, or platform algorithms. The MIT study showed something more uncomfortable: humans are the engine. We share false information because it's more novel, more surprising, and triggers stronger emotional responses than truth.
The Claim
The study by Vosoughi, Roy, and Aral at MIT analyzed every verified true and false news story distributed on Twitter from 2006 to 2017 — approximately 126,000 cascades tweeted by 3 million people more than 4.5 million times.
**Speed**: False news reached 1,500 people about 6x faster than true news.
**Reach**: Falsehoods were 70% more likely to be retweeted than true stories.
**Depth**: False stories were shared in longer chains — they penetrated deeper into social networks.
**Domain**: The effect was strongest for political false news, but held across terrorism, natural disasters, science, urban legends, and financial information.
**Mechanism**: The researchers tested a 'novelty hypothesis' — false news was more novel than truth, and people are more likely to share novel information. False stories triggered replies expressing surprise and disgust; true stories triggered sadness and trust. Surprise and disgust are higher-arousal emotions that drive sharing.
**Not bots**: The data showed humans, not bots, were the primary spreaders. When bots were removed from the analysis, the pattern held — and in some cases became stronger.
Key Evidence
- •Vosoughi, Roy & Aral (2018): published in Science — 126,000 news cascades, 3M people, 4.5M tweets, 2006-2017
- •False news reached 1,500 people ~6x faster than true news
- •Falsehoods 70% more likely to be retweeted
- •Novelty drives sharing — false news is more surprising than truth
- •Effect strongest for political news but holds across all domains
- •Humans, not bots, are the primary spreaders — pattern holds with bots removed
Practical Implication
Truth has a structural disadvantage in information markets. Knowing this changes how you consume and share information: assume surprising claims are more likely to be false precisely because they're surprising. The novelty that makes something shareable is the same quality that makes it likely wrong.
Nuance & Limits
The study was conducted on Twitter (now X) — results may differ across platforms. The researchers studied verified true/false stories, which may not capture the full spectrum of misinformation (much of which is partially true). The 'novelty hypothesis' explains why but doesn't fully account for motivated reasoning, tribal identity, or confirmation bias as additional drivers. Also, the study measured sharing, not belief — people share things they find interesting, not necessarily things they believe.
Source Material
Citation Density
Very high — one of the most cited social media studies. Referenced across tech, politics, psychology, and business podcasts. Tenev (KP #270) experienced it firsthand during GameStop.
Related Ideas
Information environments shape belief. The 'nudge' applies to what people believe, not just what they do.
You can't control what false narratives spread about you. You can control your response.