Search engines play a central role in routing political information and news to citizens. The algorithmic personalization of search results by large search engines like Google has long been suspected to contribute to the emergence of `filter bubbles.' The resulting informational segregation may fuel political polarization and have far-reaching negative social consequences. However, measuring the causal effect of algorithmic personalization on informational segregation in search results is challenging and the link between personalization and filter bubbles is poorly understood. To test whether and how Google's algorithmic personalization promotes filter bubbles, we set up a population of 150 synthetic internet users ("bots") who are randomly located across 25 US cities and are active for several months during the 2020 US Presidential Elections and their aftermath. These users differ in their browsing preferences and political ideology, and build up realistic browsing and search histories. Daily experiments in which all users enter the same election-related queries reveal that search results differ substantially between users. Results are tilted towards previously visited websites and biased towards the locally prevalent ideology. It is by these two channels that Google's algorithmic personalization of search results promotes filter bubbles.
Language
English
HSG Profile Area
SEPS - Quantitative Economic Methods
Event Title
5th International Conference on THE POLITICAL ECONOMY OF DEMOCRACY AND DICTATORSHIP
