Vincenz Leuschner, Rebecca Bondü, Marc Allroggen, Herbert Scheithauer

• Threats and announcements of homicidal violence at schools may have massive consequences like evacuations, police searches, criminal investigations, or loss of the sense of security by students, teachers, and parents. However, there is a lack of systematic studies about that phenomenon. The present article would like to contribute to closing the research gap. It presents results about the frequency and structure of threats and announcements of homicidal violence in schools in Berlin. The study is based on an official dataset from school administration reports of violent acts in Berlin schools which has been studied within the Berlin Leaking-Projekt. The sample consists of 427 threats and announcements of homicidal violence between 1996 and 2007. The study is an exceptional analysis of the phenomenon: it presents crosscutting results about frequency and characteristics of threats and the threatening students as well as results of a longitudinal analysis about the development of threats and announcements. Results show a rate of 0,3Threats and announcements of homicidal violence at schools may have massive consequences like evacuations, police searches, criminal investigations, or loss of the sense of security by students, teachers, and parents. However, there is a lack of systematic studies about that phenomenon. The present article would like to contribute to closing the research gap. It presents results about the frequency and structure of threats and announcements of homicidal violence in schools in Berlin. The study is based on an official dataset from school administration reports of violent acts in Berlin schools which has been studied within the Berlin Leaking-Projekt. The sample consists of 427 threats and announcements of homicidal violence between 1996 and 2007. The study is an exceptional analysis of the phenomenon: it presents crosscutting results about frequency and characteristics of threats and the threatening students as well as results of a longitudinal analysis about the development of threats and announcements. Results show a rate of 0,3 threats and announcements per 1 000 student and year. During the observation time span a steady increase of threats and announcements - year by year, influenced by imitation effects after school shootings - has been observed.…