Does social media use polarize political opinions?

When does social media polarize political opinions?

Social media tends to polarize when users are exposed mostly to content that matches their existing beliefs—a phenomenon called selective exposure. A systematic review of 94 studies found that pro-attitudinal media consistently exacerbates both ideological and affective polarization [6]. In Nigeria, an analysis of 30,000 posts found that 68% contained polarizing language like ethnic appeals, war metaphors, or derogatory labels, which reinforced in-group solidarity and out-group hostility [2]. Similarly, a cross-national study of 157 countries found that more time on social media and elite use of platforms to spread disinformation were linked to increased political polarization, which in turn raised the risk of civil conflict [8].

The effect is also amplified by algorithms. A field experiment showed that Facebook's algorithm was less likely to supply users with posts from counter-attitudinal outlets, even when users subscribed to them, meaning the platform itself can limit exposure to opposing views [5]. This algorithmic filtering, combined with users' tendency to follow like-minded accounts, creates echo chambers that deepen divides.

When does social media not polarize—or even depolarize?

The evidence is mixed, and several studies find that social media does not always polarize. A systematic review of 27 studies on climate change discussions found that only 13 reported polarization, while 6 found none and 8 were inconclusive [1]. The review noted that polarization was more common in studies using digital trace data (like actual posts) and focusing on affective polarization (emotional dislike), but less common in studies using self-reported surveys and examining ideological polarization (policy positions).

In China, a survey of 802 social media users found that news exposure alone did not directly polarize opinions about the three-child policy. Instead, the act of discussing and elaborating on news actually depolarized views—meaning that thoughtful engagement can reduce division [3]. A field experiment in the U.S. also found that exposing people to counter-attitudinal news (from the opposing side) decreased negative attitudes toward the other party, though it did not change their own policy opinions [5].

Platform matters too. In India, a survey-based study found that WhatsApp use strengthened the polarizing effect of social media news consumption, while YouTube use actually weakened it [4]. This suggests that the design and culture of each platform shape whether it polarizes or not.

Who is most affected by social media polarization?

The effects are not equal across all groups. Young adults in Tanzania, for example, are particularly vulnerable because social media has become their primary source of political information, often surpassing traditional media, and algorithms reinforce their existing beliefs [9]. In heterogeneous urban communities, polarization on social media translates into real-world social tensions, declining trust, and fragmented social networks [7].

Emotional factors also play a role. A study of 203 American Democrats and Republicans found that feeling hurt, dismissed, or offended was linked to higher affective polarization and lower ability to see complexity in political issues [10]. This suggests that people who already feel emotionally wounded by politics may be more susceptible to social media's polarizing effects.

Cross-cultural differences matter too. While U.S. studies often find polarization, research in China found the opposite—that social media can depolarize [3]. This may be due to different political systems, media environments, or cultural norms around public discourse.

Sources used in this answer

Related research questions