Social media posts were polarised leading up to elections in 31 countries in 2024

Publicly released:
International
CC:0
CC:0

Italian researchers analysed over 4 million Facebook posts by news outlets and political parties last year, and say right-leaning organisations focussed more on politics, religion, and migration, while left-leaners engaged more on education, health and technology. They say they looked at the social media engagement from 508 news agencies and 336 political parties across 31 countries that held elections in 2024 and found specific leans towards the political poles in the most discussed subjects. The authors say that this polarisation on topics for left- or right-leaning organisations was not absolute, as they found a few issues where left and right had common ground.

Media release

From: PLOS

Global social media engagement trends revealed for election year of 2024

Right-leaning news outlets and political parties engaged more on politics, religion, and migration; left-leaning news and political groups engaged more on education, health and technology

An analysis of more than 4 million Facebook posts created by news outlets and political parties in 2024 highlights global social media engagement trends and political polarization during a year which included elections for many countries. Giulio Pecile and colleagues at Sapienza University of Rome, Italy, present these findings in the open-access journal PLOS One on February 5, 2025.

Prior research has suggested that the tailored content presented on social media platforms may reinforce political polarization and influence voter behavior. In 2024, a substantial portion of the world’s population participated in elections, providing an opportunity to examine global-scale social media engagement during a major election year.

Pecile and colleagues analyzed how audiences engaged with content posted on Facebook by news agencies and political parties between September 2023 and May 2024. The dataset included audience reactions (for instance, likes, comments, or shares) as indicators of engagement for around 4.2 million Facebook posts made by 508 news agencies and 176,000 Facebook posts by 336 political parties, across a total of 31 countries that held elections in 2024. To uncover audience engagement patterns, the researchers applied a method known as principal component analysis, which is commonly used to analyze large datasets.

Focusing on a broad selection of topics—such as education, sports, and climate change—the researchers found that engagement on different topics varied between countries, between news outlets versus political parties, and between right- versus left-leaning groups.

Most notably, engagement levels for certain topics differed according to political orientation. Right-leaning news outlets and political parties had more engagement on politics, religion, and migration, while left-leaning news and political groups had more engagement on education, health and technology. The researchers note, however, that this divide is not absolute and their data suggest common ground can be found for specific issues.

The researchers also note that this study, while underscoring polarization risks of social media, could also inform efforts to narrow political divides and develop evidence-based policies on broadly important topics—such as climate change, public health, and war—that may be more likely to succeed.

The authors add: “This study offers a global perspective on how political and news-related content resonates with audiences during an election year, and underscores the influence of social media on public opinion and its potential to both deepen echo chambers and foster constructive dialogue on critical global issues. Exploring the topics most covered and best received by political parties and news organizations, and analyzing the substantial difference between the reception of news and party content by users, can inform those groups on how to better connect to their public during the democratic process.”

Journal/
conference:
PLOS ONE
Research:Paper
Organisation/s: Sapienza University of Rome, Italy
Funder: The author(s) received no specific funding for this work.
Media Contact/s
Contact details are only visible to registered journalists.