New University of Limerick research has revealed how political polarisation can extend to previously unpolarised or non-political topics.
A ground-breaking study published in the Journal of Artificial Societies and Social Simulations, researchers in Professor Mike Quayle’s team at UL’s Social Dynamics Lab, Department of Psychology, Centre for Social Issues Research, has shown how initially small differences in preference on a topic can be quickly amplified, often ultimately leading to a stark divergence in world views.
The study reveals two fundamental conditions that, when both are met, significantly increase the likelihood of polarisation spilling over to new topics.
The research team used both laboratory experiments and a novel methodology called agent-based modelling which explores social processes in simulated ‘mini-societies’.
The research marks a significant step forward in comprehending the dynamics of political polarisation and its impact on diverse societal issues.
“Diversity of political opinions is a sign of a healthy democracy. In contrast, a splitting of societies into competing camps that progressively dislike each other often prevents agreement and can cripples political processes. A key challenge for researchers who are interested in the dynamics of public opinion is to understand the underlying principles of polarisation,” explained Dr Dino Carpentras, a former Marie Curie Fellow at UL now based at ETH Zürich in Switzerland who is a researcher in Computational Social Science and the main author of the study.
“The COVID-19 pandemic provides a vivid example for such processes, where, in some countries, the compliance with safety rules, for example wearing a mask, became a matter of political partisanship rather than one of health-related concerns.
“Obviously, finding reasonable common ground to coordinate solutions that are backed up by large parts of societies becomes difficult if not impossible when polarisation results in perpetual partisanship,” added Dr Carpentras.
The group analysed more than two thousand interactions in an online experiment to uncover the basic principles of polarisation.
“We wanted to understand how people change their minds about politically neutral topics when they interact with strangers from which they know only their preferred political affiliation,” explained Dr Adrian Lueders, who is now working at the University of Hohenheim, Germany and was previously a post-doctoral researcher at UL’s Social Dynamics Lab.
“A general pattern that we observed in our experiment was that people tend to be influenced by others regardless of whether they belonged to the same or to different political groups.”
The so-called ‘social influence effect’, however, was significantly stronger if people learned about the opinion of someone who had a similar political leaning, according to the researchers.
“This seemed to suggest that polarisation cannot extend to new topics, but then, how can we explain the partisanship that emerged around public health interventions during the COVID-19 pandemic? The important difference is that in the experiment people interact only once, while in the real world a topic is shaped by continuous interactions and discussions,” clarified Dr Paul Maher, a lecturer in UL’s Department of Psychology.
To explore this difference further, the researchers used agent-based simulations.
“This is a methodology which allows us to simulate the interaction between thousands of fictitious people, almost like in a video game,” Dr Carpentras explained.
“This also allows us to explore different scenarios and isolate which conditions produce specific outcomes, such as the appearance of ideological division.”.
The group found that two conditions were necessary for polarisation to appear in a new topic.
Firstly, people need to be aware of the political identity of their interaction partners - a process that is common on social media platforms such as Twitter where users frequently use cues such as certain hashtags or keywords to signal political alignment.
Secondly, people need to interact more with people of the same political party, a phenomenon that has sometimes been described as an ‘echo chamber’.
“Social media has become progressively more politicised, offering people a greater capacity to interact with large audiences of like-minded others. This might be a part of the reason why political polarisation sometimes expands to aspects of people’s everyday lives,” Dr Maher concluded.
The study, ‘How Polarization Extends to New Topics: An Agent-Based Model Derived from Experimental Data’, by Dino Carpentras, Adrian Lüders, Paul Maher, Caoimhe O’Reilly and Mike Quayle, has just been published in the Journal of Artificial Societies and Social Simulations. Read the full article here: https://www.jasss.org/26/3/2.html.