Social Media and Political Behavior: Evidence from an Electoral Autocracy

• In Event: The Political Economy of COVID-19 (Virtual Session)

Abstract

Growing access to affordable mobile data represents one of the most significant technological changes over the last decade in the Global South. The contemporaneous spread of social media has created new opportunities for citizens to share information, persuade others, and coordinate their behavior; and for politicians to reach citizens directly. On the other hand, a growing literature---overwhelmingly focused on developed settings---points to the detrimental effects of social media’s spread, with viral misinformation muddying public discourse and citizens becoming increasingly polarized. In spite of this, we know little about the effects of social media on the spread of political (mis)information and its effects on citizens’ political behavior in the Global South, where differing levels of competition, press freedom, and education suggest different channels of impact. In the aftermath of Uganda’s recent contested national elections, we implement a field experiment to study how increased access to social media affects political knowledge and behavior. Across a sample of 1,500 individuals in peri-urban localities across the country, we randomize the payment of mobile data bundles and citizens’ social media taxes. These taxes, introduced by the government in 2018 and necessary to access any social media platform inside the country, otherwise act as a significant financial barrier. Leveraging rich behavioral and survey data sources, we consider how greater social media access affects measures of (i) information consumption and knowledge; (ii) political engagement, attitudes, and views about democracy; (iii) attitudes and behaviors in relation to COVID-19 restrictions and vaccination decisions; and (iv) subjective welfare and mental health.

Authors

• Jeremy Bowles, Stanford University

• John Marshall, Columbia University

• Pia Raffler, Harvard University