MIRERC 019/2025: Joint study of social media and affective polarization in Kenya

Abstract

This study aims to develop a comprehensive methodology for measuring polarization on social media
platforms, focusing specifically on affective polarization—where users distance themselves from others
due to identity or group affiliation rather than issue-based disagreements. The research will measure the
degree of polarization and polarizing behaviors across different platforms, providing a comparative
ranking of platforms based on their polarization footprint. By revealing numeric representations of
polarization, the study intends to encourage user reflection and potentially prompt behavioral changes.
The study will consist of three parts:
1. Field Study: A methodology for measuring affective polarization on platforms through
observational methods, by scraping and classifying content that appears in the feeds of users,
and the structure of the social graph.
2. Experiment: An survey-based experiment to explore the interplay between individual-level
polarization and types of polarizing content, aiming to understand the causal impact of such
content on political attitudes.
3. Survey: A survey examining user experiences on social media, adapted from the Neely Social
Media Index, to understand the perceived negative and positive effects of social media use.
By comparing the findings from all three parts, the study seeks to draw insights into the correlation
between perceived experiences and observable polarization, as well as the impact of polarizing content
on attitudes.
Kenya has been selected as the study location for several reasons:
● The country is at a critical juncture in terms of platform regulation, offering an opportunity to
influence policy discussions.
● The study aims to diversify the research on digital conflict beyond the USA and Europe,
contributing to the decolonization of the field.
● Build Up has established links with Kenyan policy bodies such as the National Cohesion and
Integration Commission, Media Council, and Communications Authority of Kenya.
The study will also explore how platforms can be incentivized to reduce polarization through league
tables, as well as investigate potential policy solutions, such as taxing platforms based on their
polarization footprint. Ultimately, the research will provide a globally replicable template for
measuring polarization and inform policies to address the growing impact of affective polarization on
society.