Social Media & Election Integrity

How do contemporary information ecosystems threaten democracy and election integrity? University of Michigan’s Associate Professor Jojoyeet Pal helps to answer this question. Using qualitative and computational methods from the field of Natural Language Processing, Pal analyzed social media discourse.

He and his collaborators measured influencer and politician polarization via their social media posts. Their research utilized word embeddings, a machine-learning method that translates words into numerical vectors. Through this method, they found that the Indian political information ecosystem tends to reward polarized speech with greater social media engagement.

In another work published in South Asian History and Culture, he and his collaboirators evaluated misinformation that circulated in the 2019 Indian elections. Their research revealed that major political parties not only employed misinformation as part of their campaign strategies, but were also frequent targets of such misinformation. In other words, misinformation became a fundamental part of election discourse.

You can read more from Pal on social media and election integrity, like his thoughts on the use of insults in the 2019 Indian elections, or the role of social media more broadly in those elections. Pal also maintains a more complete collection of his research’s coverage in the media.