User preferences and editorial content selection on Facebook

Abstract

This study investigates the relationship between user demand and news selection on Facebook. I compile an original data set of over 40,000 Facebook posts and 300,000 website articles between January 2009 to June 2017 from Germany’s leading news magazine Spiegel Online. The results suggest that editorial news selection on Facebook is driven by prior user engagement (i.e., likes, shares, and comments) with similar posts: a one standard deviation increase in likes per post in a topic category in the last week raises the number of posts in this topic category in the current week by 9.33%. The estimates are even larger for shares (9.69%) and comments (30.79%). This finding can be best explained by editors catering to consumer preferences. While a strong impact of the demand side implies that the chances of supply-driven media are small, consumer sovereignty in news markets might also undermine the media’s role as the Fourth Estate.

Keywords: news selection, user preferences, user engagement, media bias, social media

JEL Classification: L19, L82