Time Trends in Printed News Coverage of Female Subjects, 1880–2008

Abstract

Former studies have suggested that female subjects are substantially underrepresented in most media. However, no study has systematically investigated historical trends in women's coverage over the last century. We use data collected by the Lydia text analysis system to assess the development in the coverage of female subjects in 13 daily newspapers, dating back to 1880. Our trend analysis shows that female subjects' representation has remained relatively low throughout modern history. Following a temporary rise during the 1930s, it declined until the 1960s, and then showed a modest uptake in recent decades. The data further demonstrate that no newspaper section comes even close to equality between male and female names.

KEYWORDS:

• gender
• inequality
• media coverage
• news
• press
• time trends
• women

Notes

1. For example, the first GMMP took place on January 18, 1995, the day following the Kobe Earthquake in Japan and news coverage around the world during that day was dominated by this tragic event (Spears and Seydegart 2000). This potential methodological instability is also demonstrated in specific data reported by the GMMP over the years. For example, in North America the percentage of female subjects in news classified as “Economy” was 22 percent in 2005, but then soared to 37 percent within a period of only four years (Gallagher 2010).

2. Note that the ratio of female names out of all names is higher in this table than in our analysis, as the table presents distinct names rather than all references. The former means that we only count each name once, even if that name appeared many times during a given year. In the latter, however, the same name (e.g. Barack Obama) can appear many times during a year and all of them will be counted. We preferred to use this latter measure (all references) in our analyses as it accounts not only for how many male names and how many female names appeared in the news, but also for how often these names have appeared.

3. Our analysis of the coverage in Time Magazine, starting in 1920, reveals a very similar pattern to that presented here for the New York Times. The trends for Time Magazine are available from the authors.

4. The Lydia system recognizes and distinguishes between these sections according to the main themes appearing in the analyzed text.