New research: The internet does not turn people into trolls – it just makes real-life trolls more visible

Many of us feel that the internet is not a safe place to discuss politics. If we want to discuss contentious issues, we would much rather do it face to face with others. But why is this so? An often-used idea in both media and research is that the problem is the internet and social media as such: That feelings of anonymity behind the computer screens turn us all into trolls, with little empathy for our discussion partner. Yet, this argument is simply not true according to new research just published in the journal American Political Science Review.

"There are many psychological reasons why we might have a harder time controlling our temper online. We do not see the faces of those we are arguing with, and the fast-paced written form of communication can easily lead to misunderstandings. Yet, we also know from psychological research that not everyone has a personality that is equally disposed to aggression. In the end, these personality differences turn out to be a much stronger driver of online hostility," says Alexander Bor from the Department of Political Science at Aarhus University and first author of this new research.

The results of the study show that those who tend to be hostile in political discussions on the internet report to be just as hostile in political discussions face-to-face. These individuals have dispositions that make them crave recognition and status and motivate dominant and aggressive behaviour both online and offline to not lose a discussion.

The research is based on studies with more than 8,000 Americans and Danes, who were surveyed about their experiences and behaviour in political discussions that occurred either online or offline. Despite the differences in political institutions and levels of political polarisation, status seekers in both countries were the main culprits behind political hostility, both online and offline.

The visibility of internet trolls
The research also documents that people in both Denmark and the United States do in fact feel that online political discussions are worse than offline discussions, but points to a new explanation for this phenomenon.

"Our research shows that the reason why many people feel that online political discussions are so hostile has to do with the visibility of aggressive behaviour online. Online discussions occur in large, public networks and the behaviour of an internet troll is much more visible than the behaviour of this same person in an offline setting," reports co-author Michael Bang Petersen, professor of political science at Aarhus University.

In both online and offline settings, few people feel that they personally are being attacked or harassed. Yet in an online setting, there is a marked tendency for people to observe others being attacked and harassed.

Digital education is not enough
The research suggests that the internet is not responsible for making people aggressive. Rather, aggressive people utilise the features of the internet for their own purposes. According to the authors of the study, this suggests that online hostility is not an accident but rather a deliberate strategy.

"We cannot remove online hate through education because it is not born out of ignorance. Hostile people know that their words hurt, and that is why they use them. Our research suggests that it is necessary to clearly describe what is okay and what is not okay for each specific discussion page and to police those norms, for example by using moderators. To end online hate, we need to decrease the visibility and reach of those who are hateful. The alternative is that many people will be deterred from participating in online discussions. This is a democratic problem, given that social media play a larger and larger role in political processes," says Alexander Bor.

###

About the study:

• The study is financed by the Aarhus University Research Foundation and the Carlsberg Foundation.
• The research is based on a combination of cross-sectional surveys, survey experiments and behavioural experiments conducted in the United States and Denmark with over 8,400 participants in total.