An interdisciplinary team of sociologists and physicists has developed a computer model to analyze the steady rise in political polarization. The multi-agent model is particularly intended to clarify which role social media play in this. The study, which the researchers have now published in the journal PLOS ONE paints a pessimistic picture: According to the model, there are tipping points at which social polarization becomes irreversible.
Possible effects and political identity
So far it has been the case the echo chamber effect As a common explanation for political polarization in social media: Because users in social networks prefer to consume content that corresponds to their worldview, their political opinion would further solidify – according to the theory. This approach is controversial, however, because social networks are usually not the main sources of news for many users. Your effect is therefore much smaller than initially assumed. In addition, there are studies that show a paradoxical finding: If users are deliberately exposed to foreign, contrary opinions as far as possible, their political positions will not become more open and flexible, but rather they still harden.
The sociologist Petter Törnberg from the University of Amsterdam and his colleagues from Germany, Italy and Sweden assume that political polarization is an effect that can be explained by the increasing influence of political identities. According to this approach, political debates are no longer rational discussions about political disagreements. Rather, they resemble a struggle between warring tribes, in which it is more important who belongs to the in-group and who belongs to the “others” than who makes which arguments.
Software agents simulate polarization
How strong the respective political identity is, therefore, depends on the interactions with like-minded people. If, on the other hand, you have a lot to do with people from different groups, the bond with the group identity becomes weaker. To test the theory, the researchers developed a computer model to study the social dynamics of identity and political polarization. In it, software agents interact with randomly selected other agents from their immediate neighborhood and with other agents whose political identity was similar to theirs – who they had contacted according to their preference from the total pool of all agents. From the sum of all interactions, it is then calculated in each time unit of the model whether the bond with the political identity is strengthened, weakened, or even whether the identity changes completely.
While investigating the dynamics of the model, the team came across tipping points – degrees of polarization that trigger feedback loops that lead to escalating permanent political polarization. In addition, the model shows so-called “hysteresis” effects – that is, even if the conditions were to change in such a way that it would be much more difficult to find radical groups on social media, the polarization would not fall below a certain value again. “Just like climate change, political polarization can also react in unpredictable and dangerous ways,” explains Törnberg, the lead author of the study.
In order to reduce the polarization again, the separate groups would have to agree to work towards a common goal. “In the past, this task was fulfilled by the mass mobilization that characterized the Fordist society of the modern age, for example through large-scale wars,” writes Törnberg. “Today, in a postmodern, fragmented society, it is less clear how such a gathering could be implemented. About two years ago, a common response from researchers was that a large, shared problem – like a global pandemic – might offer a solution Unfortunately, that doesn’t seem to have helped. “Even if he was not relying on a technical solution to the problem in principle, he was certain” that it would be entirely possible to introduce a de-polarizing form of social media. “