Title: Retweets Distort Exposure to Polarized Information

Abstract: The growing prominence of social media in public discourse has led to greater scrutiny of the quality of information spreading online and the role that polarization plays in this process. However, studies of information spread on social media platforms like Twitter have been hampered by the difficulty of collecting data about the social graph, specifically follow links that shape what users see in their timelines. As a proxy of the follower graph, researchers use retweets to construct the diffusion graph, although it is not clear how these proxies affect studies of online information ecosystems. Using a dataset containing a sample of the Twitter follower graph and the tweets posted by users within it, we reconstruct the retweet graph and quantify its impact on the measures of exposure. While we find that echo chambers exist in both networks, they are more pronounced in the retweet neighborhood. We compare the polarization of information users see via their follower and retweet graphs to show that retweeted accounts systematically share more politically extreme content and misinformation. This bias cannot be explained by the activity or polarization within users' own social neighborhoods but by the increased attention they pay to more polarized sources. Our results suggest that studies relying on the follower graphs underestimate the polarization of information users pay attention to online.