Title: Sharing political news online: A network model of information spread on Facebook

Abstrac: Social media plays a crucial role in online political campaigns as political parties can reach, inform, and mobilize voters through these platforms. Political campaigns share information on social media to mobilize support, and prior research shows that sharing content on social media correlates with the offline popularity of political parties. In this paper, we model the spread of political content on the internet. We start by exploring popularity and sharing behavior related to posts by Hungarian politicians on Facebook. We utilize this analysis to build an agent-based model. Within this, we test how echo chambers, homophily, and network structure affect the number of shares that contribute to information diffusion on social media. Our simulation compares spreading in different network structures and shows that preferential attachment models are not the most efficient for fostering diffusion in networks with relatively low density or when a filtering mechanism is present. Our model confirms that homophily generally has a positive effect on diffusion, especially within echo chambers. Echo chambers enhance the diffusion of political news with a limited potential audience. Furthermore, the results of our agent-based simulation indicate that homophily and echo chambers can significantly influence the spread of political content on social media, with echo chambers particularly enhancing diffusion in networks where overall diffusion is low.