Title: The Returns to Viral Media: The Case of US Campaign Contributions (with Johannes Böken, Arianna Ornaghi and Nicola Mastrorocco)

Abstract: This paper provides estimates of the impact of social media attention on US campaign contributions. Our setting is a daily dataset of campaign contributions and Twitter activity for Members of Congress over the 2019-2020 period. Our average elasticity of 0.01 for the Contributions-Likes relationship is driven entirely by the top tail (90th percentile and above) of `viral' Tweets. In turn, both negative sentiment and member ideology are strong predictors of virality. The relationship with regard to ideology is U-shaped - members on the far right and far left systematically receive more Likes relative to those in the middle. As part of our identification strategy, we develop a `news pressure' strategy based on the overlap of followers between political and non-political media accounts on the network.