[Paper Review] Is 'Everyone is Angry' a Lie? The Truth About the 10% Who Write 97% of SNS Content

When browsing social media, do you ever feel like, 'Is the world really this chaotic?' or 'Everyone is so angry'? In fact, according to the latest academic research, that feeling might not be accurate. A theoretical paper integrating insights from psychology, political science, and cognitive science has gained attention for describing social media as a funhouse mirror factory." Multiple empirical studies, including surveys of over 8,000 people across 8 countries and analyses of social media post data, reveal a surprising mechanism: the voices of an extreme minority appear to be the majority." In this article, based on the paper's content, we will explain why and how social media distorts reality, supported by concrete data.

About the paper introduced here

The paper we are reviewing is a theoretical review paper that integrates insights from multiple academic fields, including psychology, political science, and cognitive science. This paper does not collect new primary data; instead, it reviews multiple existing empirical studies to systematically organize 'why social media distorts the perception of social norms.'

The central metaphor of the paper is the 'funhouse mirror factory.' Like the distorted mirrors in a carnival funhouse, social media does not reflect reality as it is, but acts as a device that magnifies extreme opinions.

Shocking Data ①: 97% of political posts come from 10% of users

Among the survey data cited by the paper, the most shocking fact is that 97% of political-related posts on Twitter (now X) are written by the top 10% most active users.

In other words, most of the 'political opinions' we see on our timelines are posted by a very small group of "loud voices." The remaining 90% of people—presumably the moderate majority—hardly post at all. They are either just reading or do not use social media much to begin with.

This is not limited to politics. Multiple studies have shown that for any topic, those with extreme opinions are more likely to post actively, while those with moderate views tend to remain silent.

Shocking Data ②: The truth about 'online hostility' shown by a survey of 8,434 people in 8 countries

Another study emphasized by the paper is the cross-national survey by Bor & Petersen (2022). The total sample size reaches 8,434 people.

What this study revealed is that 'people who are aggressive online are just as aggressive offline.' In other words, social media does not 'change' people to be aggressive; it is simply a place where naturally aggressive people are more visible.

Moderate people refrain from posting, while extreme people speak out actively. As a result, only aggressive opinions become visible on social media.

Why do only extreme posts stand out? — The problem of algorithms

What makes the problem even worse is the platform's algorithm.

Social media algorithms are designed to maximize 'engagement' (likes, comments, shares). So, what kind of posts are most likely to gather engagement?

According to the research by Brady et al. (2017, 2021) cited by the paper, those are posts containing 'moral outrage,' 'hostility,' and 'negative emotions.' Analysis of social media post data reveals that posts containing 'hostility toward out-groups (groups different from one's own)' are particularly likely to be shared.

As a result, moderate and calm posts are buried, and the timeline becomes dominated by extreme and aggressive posts.

The brain's 'averaging' function backfires

Here, insights from cognitive science are important.

Goldenberg et al. (2022) conducted 9 experiments (with a total of 1,583 participants) to verify a function in the human brain called 'ensemble encoding.'

This is a mechanism where, when we see many stimuli, we do not remember them individually but process them collectively as an average impression of "what it's generally like." For example, when we see a photo of a crowd, we grasp "what the overall expression is like" rather than the faces of each individual.

The research showed that there is an amplification effect where, if extreme emotional expressions are mixed in, the average emotion of the whole is overestimated.

When applied to SNS, it works like this. When only extreme opinions flow through your timeline, your brain processes them as the "average." Even though they are actually minority extreme opinions, you perceive them as "this is the average opinion of society."

"Overperception of anger"—the other party is not as angry as you think

Even more interesting is the research by Brady et al. (2023). This study compared the actual anger of the poster and the degree of anger estimated by those who saw it regarding posts on SNS that contained moral outrage.

The results were clear. Observers estimate that the poster is "angrier" than they actually feel.

This is a phenomenon called "overperception of anger," and it is a direct cause of "false polarization"—that is, believing that the opposing party is much more radical and hostile than they actually are.

Two social problems: "Pluralistic ignorance" and "false polarization"

The paper points out two major social problems caused by these mechanisms.

① Pluralistic ignorance

This is the misunderstanding that a minority behavior is actually a majority behavior.

For example, when college students frequently see photos of parties or posts about being drunk on SNS, they assume that "it's normal for all college students to drink this much." In reality, only a small portion of students are engaging in dangerous drinking, but it looks like the "norm." This has a serious impact on health behaviors.

② False polarization

This is the feeling that those who are politically opposed are "more extreme and unreasonable" than they actually are.

What stands out on SNS are the most extreme people from both camps, even though there are actually many moderate people on both sides. Despite this, both sides assume that "the other side is radical," making dialogue difficult.

Why fact-checking cannot solve it

There is an important point that the paper emphasizes.

This problem is different in nature from simple "fake news" or "misinformation." What is written in extreme posts is, in many cases, a "fact." There are people who actually hold those opinions, and there are people who actually performed those actions.

The problem is that those "facts" do not represent the actual distribution. It is that a very small number of extreme cases appear as if they are the overall trend.

This is a problem of "biased sampling of facts," and it is not the type of thing that can be pointed out as a "lie" through fact-checking. That is precisely why it is difficult to address.

User Desires and Platform Reality

What is interesting is the gap between user perception and reality shown by the study by Rathje et al. (2023).

According to the survey, many SNS users respond that they want "positive, accurate, and nuanced content" to spread. However, at the same time, they recognize that what actually "goes viral" is negative and divisive content.

There is a significant gap between what users want and how platforms are designed.

Limitations and Cautions of the Paper

This paper also has its limitations.

First, many of the cited studies use data from North America and Europe, and it has not yet been sufficiently verified to what extent the same processes hold true in Asia, including Japan.

Also, this is a narrative review (a review where the authors select and summarize relevant studies), and compared to systematic reviews or meta-analyses, there is a possibility that contradictory studies not mentioned exist.

However, the individual studies cited were published in major peer-reviewed academic journals such as Nature Human Behaviour, American Political Science Review, and Science Advances, and are considered to have a certain level of reliability.

How Should We Take This?

There are several implications from this study.

1. Do not take the "public" you see on SNS at face value The opinions and behaviors that flow through your timeline are highly unlikely to reflect the actual distribution. Even if you feel like "everyone is saying that," it might be an illusion.

2. Do not overestimate the "enemy" People with opposing political views may not be as extreme as they appear on SNS. If you actually meet and talk to them, you will often find that you can communicate surprisingly well.

3. Be aware of the existence of the "silent majority" Because people with moderate opinions do not speak up, the structure where only extreme opinions stand out is reinforced. It is important to be aware of this structure.

Summary

This paper, which integrates multiple empirical studies, reveals the mechanism by which social media functions as a "funhouse mirror."

• 97% of political posts are generated by 10% of users

• Posts containing anger or hostility are favored by algorithms

• Due to the brain's "averaging" function, extreme opinions appear "normal"

• As a result, pluralistic ignorance and false polarization occur

If you feel like 'the world is coming to an end' after looking at social media, please take a moment to pause. What you are seeing might not be reality, but a 'distorted mirror'.

#PaperReview #SNS #SocialMedia #FilterBubble #EchoChamber #Polarization #MediaLiteracy #InformationLiteracy #CognitiveBias #Algorithm #FactCheck #Psychology #SocialPsychology #NetLiteracy #DigitalWellbeing #ScientificPaper #Evidence