Measuring Media Content Concentration at a Large Scale Using Automated Text Comparisons

Abstract

ABSTRACT The structural crisis of the news media and the according lack of resources for journalism has led to several types of concentration in most Western media markets. Not only is concentration of news media ownership increasing, but there is also more editorial cooperation between and within media companies and more reliance on external sources for content, such as from news agencies. Editorial cooperation and external source reliance have led to concentration processes on the level of content — a phenomenon insufficiently considered in existing research. This paper suggests an innovative method to measure media content concentration with automated one-by-one text comparisons. The procedure allows identifying similar media articles that appear in multiple media outlets and therefore adds to the current literature, which mostly measures media concentration on the basis of readership or audiences of outlets and their ownership. The longitudinal study combined automated text comparisons using Jaccard similarities with manually coded data using a text corpus of seven newspapers between 2012 and 2018 in Switzerland — a small and disrupted media market. The results showed an increase in media content concentration, especially in the coverage of international events with editorial cooperation and reliance on news agencies as the drivers.