Herd journalism: Investment in novelty and popularity in markets for news

- Past work on news focuses on bias. Our focus: (1) novelty of coverage and (2) popularity in other outlets.
- Analysis based on major newsweeklies cover stories (2001-2009) using keywords and MTurk.
- In game theory model, news firms invest due to consumer preferences for novelty/popularity.
- Novelty and popularity increase sales; popular stories have a lifespan of one to two weeks.
- Impact of novelty/popularity declined post 2006; lifespan has declined to under one week.

Consumers of news care both about the novelty of the news they read, as well as how popular that news topic is with others. Editors choose what to report on based on consumer preferences and the coverage of their competitors. We build a continuous time model that predicts whether news providers invest in covering novel news stories or instead report on popular. We construct a dataset of cover stories to test the model and find that both novelty and popularity are associated with increased sales, but popular stories only have a lifespan of one to two weeks. We find that the impact of novelty has declined since 2006, and the lifespan of a story has shortened. While theory predicts that editors strategically alternate between reporting on novel stories and popular stories, we find evidence of positive serial correlation in the popularity of cover stories. Finally, news outlets often cover the same story. Theory predicts that competitive news outlets are more likely to pool on the stories they cover while news outlets with market power are more likely to separate.