The Perils of Social Media Politics

In the APSA Public Scholarship Program, graduate students in political science produce summaries of new research in the American Political Science Review. This piece, written by Maria Nagawa, covers the new article by Sumitra Badrinathan, University of Oxford: “Educative Interventions to Combat Misinformation: Evidence from a Field Experiment in India”.

Remember when on a visit to Kenya, President Barack Obama pushed an elderly lady out of his way? Well, he didn’t. Here in the United States, such a false story might be tagged or taken down if it appeared on Twitter or Facebook. But in developing countries, where information is shared on encrypted social media applications like WhatsApp, users have to fact check stories themselves. And when these stories are incendiary or divisive, they can have devastating consequences, including violence and death. However, as author Sumitra Badrinathan has found, combating privately shared misinformation, particularly in a politically contentious environment, is difficult. Even after an intervention training of social media users on how to verify information, users were still largely unable to correctly identify misinformation.

The author investigates the spread of misinformation in India during the 2019 elections, when Internet campaigning was permitted for the first time. Politicians exploited this opportunity to propagate political information through WhatsApp messaging groups that targeted religious, socioeconomic, caste, and other identities. Indian citizens are particularly vulnerable to misinformation. Low levels of literacy and education, high levels of Internet access but naivete about its use, access to a private and encrypted messaging application, and the graphic and salient nature of information contribute to this vulnerability.

The author conducted an experiment with a randomly selected group of 1,224 respondents in Bihar, India where literacy rates are among the lowest in the country and Internet access is at 35 percent. The participants who received the intervention were trained to identify false information using reverse image searching and navigating a fact-checking site. To account for the role of partisan politics, the group was further divided into two: one that received corrections to false pro-ruling party messaging and another to false anti-ruling party messaging. Two weeks later, after voting in national elections but before results were out, respondents were tested on whether they could accurately identify a false news story about the ruling party.

The author finds that people generally do not believe in false news stories. Less than 10 percent fall for half the false stories, but when they do, they tend to believe in pro-ruling party misinformation regardless of party affiliation. Small differences in beliefs between ruling party supporters and non-supporters underscore this point. The largest difference was estimated at 9 percent, in sharp contrast to the United States where the differences between Republicans and Democrats can be as large as 80%.

“Developing countries, with the bulk of the world’s population and rapidly increasing access to the Internet and social media, are going to become the primary sources and recipients of political misinformation. “ Surprisingly, ruling party supporters were more likely to correctly identify false news stories when those stories are negative. This is attributed to their party’s advantage in disseminating information.

Conversely, participants who were trained to identify misinformation did not increase their ability to do so. This is exacerbated by partisanship. Ruling party supporters who were trained became worse at identifying false pro-party stories (and not anti-party stories) while non-supporters improved, demonstrating that partisanship plays a role in moderating beliefs.

The study is novel in showing how attempts to correct false stories in developing countries might fare. Developing countries, with the bulk of the world’s population and rapidly increasing access to the Internet and social media, are going to become the primary sources and recipients of political misinformation. By showing that misinformation is sticky, and the success of corrections is moderated by party affiliations, Badrinathan forces us to contend with the challenges of battling misinformation in the world’s largest political battlegrounds.