Abstract
We present the first macro-level evidence regarding the Werther effect by exploring the association between suicides in books and actual completed suicides in the United States between 1950 and 2000, using suicide-related words and phrases in the Google Books Ngram corpus. To rule out the possible influence of alternative mass media, we make use of the New York Times corpus and the Internet Movie Database (IMDb) to construct measures for suicides in news and electronic media. Our findings reveal that suicides in nonfiction books during the second half of the twentieth century significantly predict actual suicide rates, extensively controlling for suicides in other media as well as context factors. Suicides in fiction alone do not predict suicide rates due to the replacement effect.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
