Centre for Development Economics
Department of Economics
Delhi School of Economics
ANNOUNCE A SEMINAR
Curated Truths: A Model of Media Bias under Social Media Algorithms and Fake News Competition
by
(IGIDR)
(Thursday, April 9, 2026, at 2:00 PM IST )
Venue: Amex Room
Abstract:-
We examine the reporting strategy of a mainstream media outlet that competes for user endorsements on a social media platform against a bot disseminating fake news. Platform users comprise quasi-rational voters who value both accurate information and the endorsement of reports consistent with their ideologies, and who are, on average, nonpartisan. The bot generates right-leaning reports with an exogenously given probability, while the platform employs a personalized algorithm designed to maximize user engagement. Our model predicts that, in equilibrium, the media outlet adopts a left-leaning bias to differentiate itself from the bot. This bias diminishes as (i) the algorithm places greater weight on content previously endorsed by users and (ii) voters become polarized towards the right. The relationship between media bias and the right-bias of the bot follows a U-shape (is positive) for higher (lower) degrees of personalization.