ResearchHow Personal Commonalities Foster Closeness of Political Views

Press Release No. 15/2022
11 February 2022

Study shows that divergent opinions need not necessarily lead to polarisation

We naturally feel close to people who are similar to us, who share our interests or partake in related activities, for instance. This natural feeling of closeness can be funnelled to reduce political differences and increase consensus on conflictual political topics. That is a discovery made by an international team of researchers led by Dr Stefano Balietti, a social scientist from Heidelberg University who uses computer-based methods to study human group behaviour. In an online study with participants from the United States the scientists investigated the circumstances under which people change their own attitude towards a certain topic. According to the study, having points in common helps people to move towards one another, even in the case of extremely divisive political issues and despite strong personal convictions.


“Scholars have long assumed that the informal, respectful exchange between individuals with differing political views contributes to strengthening trust in democracy and avoiding societal division. That is particularly true regarding discussions between people who are friends and come from a similar social background. Conversely, when people are suddenly confronted with the political view of strangers, as particularly happens in the social media, the resultant discussions often take a negative turn,” explains Dr Balietti, a researcher at the Alfred Weber Institute for Economics of Heidelberg University and a fellow at the Mannheim Centre for European Social Research (MZES). With their study, Dr Balietti and his colleagues explored the question of whether individuals change their attitude towards polarising political issues on the basis of personal commonalities, even if they are not acquainted.

The researchers drafted an online experiment on the focus topic of inequality and redistribution of wealth. The first phase began with asking the participants about their personal characteristics, their political leanings and the topic of the experiment. They then composed a short, argumentative essay in order to persuade “a friend” to assimilate their attitude regarding inequality and redistribution. A second phase was carried out around six months later with a new group of participants. With the aid of an algorithm they were allocated to a partner from the first cohort of the experiment based on incidental similarities like age, gender, interests or characteristics, as well as their attitude to the focus topic. They received a computer-generated social media profile of their partner, showing what they had in common, and then read their essay. In the last stage the researchers again asked the two partners about their view of the focus topic and how close they felt towards one another.

“Surprisingly, both participants with strong convictions and those with mild convictions changed their opinion about the focus topic in the direction of the opinion of their partner – independently of their political leanings. This led to a decline in polarisation and, overall, an increase in support for redistributive policies,” Dr Balietti reports. If the two partners felt close based on shared personal features, the probability of assimilating the opinions of their partner rose by 86 percent. In the long term, the researchers want to discover whether this mechanism based on random similarities can also be used to design social media platforms to counteract hate and disinformation and to foster a respectful exchange of opinion and subsequent consensus-building.

The study, conducted with the participation of US researchers from the University of California in Santa Cruz, Microsoft Research in New York and the University of Pennsylvania, was published in the journal “PNAS”.

Original Publication

S. Balietti, L. Getoor, D. G. Goldstein, D. J. Watts: Reducing opinion polarization: Effects of exposure to similar people with differing political views. PNAS (28 December 2021)

Further Information