Abstract

Contrary to popular narratives, it is not clear whether using social media for news increases belief in political misinformation. Several of the most methodologically sound studies find small to nonexistent effects. However, extant research is limited by focusing on few platforms (usually Facebook, Twitter or YouTube) and is heavily U.S. centered. This leaves open the possibility that other platforms, such as those that rely on visual communication (e.g., Instagram) or are tailored to strong-tie network communication (e.g., WhatsApp), are more influential. Furthermore, the few studies conducted in other countries suggest that social media use increases political misperceptions. Still, these works use cross-sectional designs, which are ill suited to dealing with omitted variable bias and temporal ordering of processes. Using a two-wave survey fielded in Mexico during the 2021 midterm elections (N = 596), we estimate the relationship between frequency of news exposure on Facebook, Twitter, YouTube, Instagram and WhatsApp, and belief in political misinformation, while controlling for both time-invariant and time-dependent individual differences. I n contrast to political discussion, information literacy and digital skills, none of the social platforms analyzed exhibits a significant association with misinformed beliefs. We also tested for possible indirect, moderated, and reciprocal relationships, but none of these analyses yielded a statistically significant result. We conclude that the study is consistent with the "minimal media effects" paradigm, which suggests that efforts to address misinformation need to go beyond social platforms.