Social media is often blamed for turning political debates into toxic battlegrounds, but a new study provides evidence that offline societal inequalities actually play a role in shaping this digital aggression. The findings suggest that users in less democratic and less economically equal countries experience significantly more political hostility online than those in more egalitarian nations. This research was published in Nature Human Behaviour.
When the internet first emerged, many observers hoped it would create a more equal and inclusive public square. This optimism extended to autocratic regimes, where early social media platforms were sometimes viewed as tools for democratic uprisings. Today, however, political discussions on social media are widely seen as aggressive and hostile.
People tend to blame the technology itself for this shift. They point to the anonymity of the internet or the way social media algorithms reward outrage and divisive content. While platform features do play a part, researchers wanted to know if the broader socio-economic and political environment also shapes online behavior.
Previous studies provide evidence that offline inequalities tend to breed social conflict. When a society has extreme economic or political disparities, people often face intense competition to improve or defend their social standing. This competition can increase a psychological trait known as status-seeking motivation.
Status-seeking motivation is the drive to acquire power, wealth, or prestige, sometimes through dominance, fear, and intimidation. People with high status-seeking motivations tend to be more aggressive in general. Prior research linking status-seeking to online hostility has mostly focused on a few Western countries.
Alexander Bor, a researcher at the Central European University’s Democracy Institute, noticed this geographical gap in the scientific literature. He and his colleagues decided to look beyond the Western world to understand how digital hostility works on a global scale.
“We have been studying the topic of online political hostility for several years,” Bor said. “Our initial work focused on the U.S. and Denmark led us to the surprising conclusion that the same people are hostile in online and offline political discussions.”
Bor noted that expanding the scope of the research was necessary to see if these patterns held up worldwide. “But our theory implied that similar mechanisms may be at work, and there is very little research on online hostility from the Global South,” Bor said. “Hence, we set out to test our hypotheses in a large sample of 30 very different countries.”
To explore these ideas, the researchers set up an observational study involving 15,202 participants from those 30 different countries. These countries spanned six continents and were selected to provide a wide variety of political systems and economic realities. The researchers used quota sampling to recruit participants in each nation.
Quota sampling is a statistical method where the people chosen for a survey are balanced to reflect the broader population in terms of age, gender, and education. Participants completed online surveys asking about their experiences with political discussions over the previous 30 days. The survey defined political discussions as any conversations about societal issues, including local problems, national politics, or international conflicts.
The researchers measured online political victimhood by asking participants how often they saw content that ridiculed, cursed at, humiliated, or harassed people like them. To measure hostility perpetration, the researchers asked participants how often they personally engaged in those same aggressive behaviors. They asked about these actions in both online settings and face-to-face offline interactions.
The survey also included a specialized questionnaire to measure status-driven risk-taking. This questionnaire identifies individuals who are willing to take extreme personal risks solely to gain money, power, or prestige. The researchers found that experiences of online political hostility vary systematically and strongly across different countries.
Specifically, people living in less democratic countries report experiencing significantly more online hostility. The study used an established metric called the Liberal Democracy Index to rank countries by their political freedoms. Nations scoring lower on this index had noticeably higher rates of self-reported online victimhood.
A similar pattern emerged for economic inequality across the studied nations. The researchers used the Gini coefficient, which is a standard statistical measure of income distribution within a nation. People living in countries with higher economic inequality experienced more online political hostility than those in more egalitarian societies.
The United States provides an interesting benchmark for these economic findings. “We hear so much about the problem of online political hostility in the U.S., yet according to our data, the U.S. is pretty typical for a high economic inequality country,” Bor said.
The survey responses also provided evidence about the psychological roots of this hostility. Across all 30 countries, the researchers found that people who act hostile online also act hostile in offline face-to-face conversations. This suggests that digital aggression is not simply a product of internet anonymity or platform algorithms.
“Most people assume that there is something in the internet that brings out the worst in people,” Bor said. “Yet, our research shows that for some people, those who have a strong status-drive, the internet is just another domain where they can be hostile.”
The researchers noted that these status-driven individuals are more prevalent in less democratic countries. This higher concentration of status-seeking people might help explain why less democratic societies see higher overall levels of digital hostility.
The authors also analyzed demographic factors to see which groups drive this behavior. They found that young men are consistently the most hostile group in online political discussions around the world. Because less democratic countries often have younger populations, this demographic reality also contributes to the higher average levels of hostility observed in those nations.
To ensure their survey tools were accurate, the researchers ran a series of follow-up experiments involving 4,294 participants in the United States. One experiment used a virtual dice-rolling game to test for social desirability bias. Social desirability bias is the tendency for survey respondents to answer questions in a way that makes them look good, such as hiding their own bad behavior.
The dice game provided evidence that participants were answering honestly about their hostile actions. Another follow-up experiment tested whether frequent exposure to online hostility makes people numb to it. The researchers showed participants a series of highly aggressive social media posts and then asked about their general experiences.
They found no evidence that this repeated exposure caused people to under-report their own victimhood. A third experiment tested whether hearing narratives about online hostility influences how much hostility people report experiencing. The researchers found that reading a statement about high levels of internet toxicity slightly increased self-reported victimhood.
However, this suggested that the main international survey likely underestimated the true amount of hostility. This gave the researchers confidence that their findings were mathematically sound. While the study provides a detailed global snapshot, it does have limitations that require attention.
The data is observational, meaning it captures a specific moment in time and cannot definitively prove cause and effect. It remains possible that the relationship works in both directions. Highly hostile online environments might actually worsen offline inequalities and weaken democratic institutions.
Another limitation is the reliance on self-reported survey data rather than digital tracking. “We study self-reports using survey questions and not actual behavior on the internet,” Bor said. “We think this is the best way to compare countries, but admittedly, we are talking about subjective perceptions and objective hostility.”
This means the researchers measured how people felt they were treated or how they remembered acting, rather than analyzing computer logs of their exact words. Automated behavioral data from social media platforms would provide an additional layer of evidence. However, algorithms designed to detect hostility are not yet reliable across dozens of different languages and cultural contexts.
The authors suggest that future research should use longitudinal data to track these trends over time. Longitudinal studies follow the same groups of people for years, which helps determine whether changes in a person’s offline environment directly cause changes in their online behavior.
Bor is taking his own research in a slightly different direction moving forward. “I am now getting ready to do a large, five-year project on a different topic, how our moral intuitions underlie support for democracy,” Bor said.
Still, the authors hope this study encourages other scientists to broaden their geographic focus. “Our research offers a stark reminder that we should focus more on the Global South when we are studying problematic internet behavior,” Bor said.
Interestingly, public perceptions of social media reflect these complex dynamics. The researchers found that people in democratic countries tend to view social media primarily as a source of societal turmoil. People in less democratic countries are more likely to see social media as an instrument for liberation and free expression, even though they experience more hostility on those platforms.
These differing viewpoints suggest that efforts to fix social media cannot rely on a one-size-fits-all approach. Platform regulations might help curb immediate harms, but the findings indicate that digital hostility is deeply intertwined with broader societal struggles. Reducing online aggression may require addressing the offline political and economic inequalities that drive status competition in the first place.
The study, “Social media users experience more political hostility in less economically equal and less democratic societies,” was authored by Alexander Bor, Antoine Marie, Lea Pradella, and Michael Bang Petersen.