Addressing the Selection Bias in Media Coverage of Strikes: A Comparison of Mainstream and Specialty Print Media

Abstract

Despite an increase in research that examines the media's selection of protest events for coverage, two areas of study have been left undeveloped. First, the type of protest examined is limited to common forms of the demonstration (march, vigil, rally). A second drawback of this literature is its focus on mass audience newspapers. The goal of the current study is to address these two issues by comparing coverage of a previously ignored form of protest, the strike, across two different media sources, the mass audience New York Times and the Daily Labor Report, a newspaper which targets industry and labor leaders and garners its revenue from subscriptions, not advertising. Due to specific differences between the two newspapers (primarily readership and revenue base), it is expected that certain strike characteristics (industry) will play a greater role in the New York Times’ selection of strikes than in the Daily Labor Report. Using government data to construct the population of events, I find that both newspapers select strikes in a manner that resembles coverage of other forms of protest. Important variables include size, length, and disruptiveness. The main difference between the two newspapers is the New York Time's attention to strikes in industries that affect the public and consumers and its strong regional bias. These findings indicate that not only do similar media selection processes work for both protest and strikes, but also that, despite some differences, media type did not affect selection greatly.

Citation

Martin, A.W. (2005), "Addressing the Selection Bias in Media Coverage of Strikes: A Comparison of Mainstream and Specialty Print Media", Coy, P.G. (Ed.) Research in Social Movements, Conflicts and Change (Research in Social Movements, Conflicts and Change, Vol. 26), Emerald Group Publishing Limited, Bingley, pp. 141-178. https://doi.org/10.1016/S0163-786X(05)26005-6

Download as .RIS