The present dissertation examines the phenomenon of political polarisation on social media.
Specifically, the dissertation addresses the question of how the intensity of polarisation and
the ideological lines along which it occurs might vary between different national contexts.
First, it explores the differences in the intensity of political polarisation on Twitter in 16
democratic countries (Article 1). Second, it examines the ideological lines along which
polarisation occurs in two non-Western contexts, specifically among Russian (Article 2) and
Ukrainian (Article 3) users of Vkontakte – a social media platform popular among users
from post-Soviet states. The dissertation demonstrates that the levels of political polarisation
differ dramatically between countries. In democracies, polarisation tends to be lowest in
multi-party systems with proportional electoral rules (e.g., Sweden), and the highest in
pluralist two-party systems (e.g., United States). It also shows that, in non-democratic non-
Western contexts, polarisation does not necessarily run along the left–right spectrum or
party system lines. In authoritarian regimes or those with less stable party systems,
polarisation runs along the lines of other issues that are more politically relevant in a given
context. In Russia, polarisation manifests itself along pro-regime vs anti-regimes lines,
whereas in Ukraine, polarisation happens around geopolitical issues. Polarisation on social
media thus tends to reflect existing political cleavages and their intensity, in line with the
theory of political parallelism. The major implication of this dissertation in the context of
research into polarisation on social media is that findings on the topic from single-country
studies that come from Western democratic contexts should be interpreted with caution, as
they are not necessarily generalisable. To make generalisable inferences about the
relationship between social media and political polarisation, more comparative studies are
needed, as well as studies that take into account platform affordances and the causal
mechanisms that might drive polarisation.