Why Fear Spreads Faster Than Facts in a Crisis

The Signal That Wasn’t

In the spring of 2020, a friend forwarded me a voice memo. A nurse, she claimed, had recorded it in a Manhattan ER. The details were specific: body bags stacked in hallways, staff running out of oxygen, a death count the hospital was hiding. I forwarded it to three people before I checked. The hospital had issued a denial. The voice was never verified. The memo was a composite of rumors from three different cities.

I did not forward the correction.

That is the puzzle at the heart of a 2022 paper by Roger Kasperson, Thomas Webler, Bonnie Ram, and Jeannette Sutton, published in Risk Analysis (Kasperson et al., 2022). They revisit a framework called the Social Amplification of Risk Framework, or SARF, first proposed in 1988. The framework was built to explain a stubborn fact of crisis behavior: people routinely overreact to small risks and underreact to large ones. But the 2022 paper argues something sharper. It is not just that people miscalculate. It is that fear has a structural advantage over fact. The architecture of how we share information, especially on social media, acts as a signal booster for dread.

The authors do not say we are irrational. They say we are rational in a system built to amplify the wrong signals.

Why a 34 Year Old Framework Still Explains Your Twitter Feed

SARF was born in an era of fax machines and three television networks. Kasperson and his colleagues originally wanted to understand why some hazards, like nuclear waste, triggered massive public outcry while others, like radon in basements, barely registered. The framework proposed that risk events pass through "amplification stations": media outlets, community groups, government agencies, and informal social networks. Each station can amplify or attenuate the signal. A small leak at a chemical plant becomes a national story. A thousand deaths from air pollution remain invisible.

The 2022 paper updates this for the age of algorithmic feeds. The authors write that social media platforms act as "continuous and instantaneous amplification stations" (Kasperson et al., 2022). But they add a crucial twist. Traditional media had gatekeepers. Editors could decide a story was not worth running. Social media has no gatekeepers. It has engagement metrics. And fear outperforms fact on every metric.

The paper reviews three decades of empirical research testing SARF. The authors acknowledge that critics have long argued the framework is not testable in a strict scientific sense. It is more a "post hoc" tool for explaining why things went wrong than a predictive model (Kasperson et al., 2022). But they push back. The value of SARF, they argue, is not prediction. It is organization. It gives researchers a way to map the pathways through which risk information travels and mutates. And in a crisis, that mapping is what tells you where to intervene.

The Asymmetry of Attention

Here is the core finding that should worry anyone who has ever shared a warning without checking it. The authors describe a mechanism called "social amplification." It works like this: a risk event occurs. It is small. But it gets picked up by a group with strong emotions. They share it with vivid language. That version gets picked up by media. Then by politicians. Each step adds emotional weight. The original risk is now a crisis.

The paper cites a classic example: the 1989 Alar scare in the United States. A chemical used to regulate apple growth was linked to cancer in animal studies. The risk was real but tiny. A single advocacy group released a report. The media ran with it. Schools pulled apples from cafeterias. Parents poured out juice. The chemical was eventually banned, but the economic damage to apple growers was enormous. The actual risk, the authors note, was far smaller than the public response.

But here is the part that is not a simple story of mass hysteria. The authors argue that amplification is not irrational. It is a side effect of how humans process threats. We evolved to treat potential dangers as urgent until proven safe. That heuristic worked on the savanna. It does not work in a viral information ecosystem where every threat is presented as urgent.

The paper does not give a neat number for how much faster fear spreads than fact. But it offers a structural explanation. Facts require context. They require caveats. They require time to verify. Fear requires none of those things. A rumor can be shared in the time it takes to type a sentence. A correction takes a paragraph, a source, and a reader willing to slow down.

How Social Media Turned Up the Volume

The 2022 paper is especially interesting when it addresses what SARF could not have predicted in 1988. The authors write that social media channels introduce "new dynamics" that the original framework did not account for (Kasperson et al., 2022). Specifically, they point to three features.

First, speed. Information travels faster than any verification system. By the time a fact checker has published a correction, the original false claim has been seen by millions. The correction reaches a fraction of that audience.

Second, echo chambers. The authors note that risk information is now filtered through personalized algorithms. If you already fear vaccines, your feed will show you more stories about vaccine injuries. If you trust them, you will see stories about disease outbreaks. Two people can experience the same crisis with completely different information diets.

Third, emotional contagion. The paper reviews research showing that emotional content spreads more widely on social media than neutral content. Anger and fear are the most shareable emotions. The authors do not say this directly, but the implication is clear: the platforms are designed to amplify whatever gets the most engagement, and fear gets the most engagement.

The Experiment Nobody Ran

The 2022 paper is a literature review and a theoretical update, not a controlled experiment. The authors do not claim to have proven that fear spreads faster than fact in every crisis. They are offering a framework for understanding how risk communication works in the wild. That is both a strength and a limitation.

What the paper does not do is quantify the gap. It does not say, "fear spreads 10 times faster than fact" or "corrections reach 30 percent of the audience." Those numbers would be useful, but they are hard to produce in a way that generalizes across crises. The authors are honest about this. They note that SARF has been criticized for being "useful, at most, as a post hoc analysis" (Kasperson et al., 2022). They do not fully refute that criticism. Instead, they argue that a framework does not need to predict outcomes to be valuable. It just needs to help you see the system.

This is a reasonable position, but it leaves a gap. If you are a public health official trying to decide whether to correct a viral rumor or focus on distributing masks, the framework tells you that the rumor will spread fast. It does not tell you how fast, or whether correcting it is worth the effort. The authors acknowledge this as an area for future research.

What the Research Does Not Prove

It is tempting to read this paper and conclude that people are stupid. That is not what the authors say. They argue that amplification is a feature of the system, not a flaw in human cognition. People are not bad at evaluating risk. They are responding to signals that have been distorted by the time they reach them.

The paper also does not claim that all amplification is bad. Sometimes a small risk really is worth a large public response. The Alar case, for example, led to a ban on a chemical that had some evidence of harm. The authors are careful to say that amplification can be appropriate. The problem is when it is disproportionate.

What the paper does not address directly is the question of intent. Some amplification is organic. People share out of genuine concern. But some is strategic. Bad actors know that fear spreads faster than fact, and they exploit that. The framework does not distinguish between the two. That is a limitation the authors acknowledge.

How This Changes What You Should Do

If you are a communicator, a journalist, or just a person who wants to be less wrong in a crisis, the 2022 paper offers a clear lesson. Do not fight the amplification system. Work with it.

The authors suggest that effective risk communication must "design strategies that account for the social dynamics of risk perception" (Kasperson et al., 2022). That means accepting that people will be emotional. It means not leading with statistics. It means telling a story that competes with the fear story.

Here is a concrete example. In the early days of COVID-19, a rumor spread that the virus was a bioweapon. The factual response was to explain the genetic evidence showing natural origin. That response was correct. It was also boring. The fear story had drama, villains, and a clear narrative. The factual story had a lab in Wuhan and a complicated paper about bat coronaviruses.

A better response, based on the SARF framework, would have been to tell a different emotional story. Something about the scientists who discovered the virus. Something about the global effort to sequence it. Something that matched the emotional intensity of the fear story but pointed in a different direction.

The paper does not give a template. But it gives a principle. Facts do not compete with fear on a level playing field. Fear gets a head start and a faster lane. If you want to correct misinformation, you have to play in the same lane.

What This Actually Means

• ▸Lead with narrative, not data. In a crisis, people do not need a spreadsheet. They need a story that makes sense of the threat. The SARF framework shows that emotional content travels faster. If you want your message to spread, give it emotional weight. That does not mean lying. It means finding the human story inside the data.

• ▸Correct early and correct often. The paper shows that amplification happens fast. By the time you have a polished correction, the damage is done. The authors suggest that risk communication must be "timely and transparent" (Kasperson et al., 2022). In practice, that means releasing information as you have it, even if it is incomplete. A partial truth released early is more useful than a complete truth released late.

• ▸Map your amplification stations. If you are managing a crisis, identify the key nodes where information flows. Is it a community Facebook group? A local news station? A WhatsApp chain? The SARF framework is useful for mapping these pathways. Once you know where the signal is being amplified, you can inject your message there.

• ▸Do not assume corrections work the way you think they do. The paper suggests that corrections often fail because they do not reach the same audience as the original misinformation. If a rumor spreads in a closed Facebook group, a correction on Twitter is useless. You have to enter the same channel.

• ▸Accept that some amplification is permanent. The authors note that once a risk has been amplified, it cannot be fully attenuated. The memory of the rumor persists. The correction is always playing catch up. This is not a reason to give up. It is a reason to be faster, more creative, and more emotionally intelligent than the people spreading fear.