The main objective of this thesis is to contribute to a more systematic understanding of how mainstream news media in liberal democracies report about protests. Existing research indicates that when mainstream news media report about demonstrations, protesters often face delegitimising coverage. This phenomenon, known as the “(journalistic) protest paradigm, is thought to be a default mindset that leads journalists to emphasize the method of protesters over their message — restricting the impact of one of few tools citizens have to raise important issues. More recent studies, however, suggest a more mixed picture, indicating both that the protest paradigm is used more conditionally than previously thought and that there have been overall changes in protest reporting in recent decades. There are limitations to the existing literature, however. The scope of studies has been rather narrow, focusing only on single, often radical, protest events or scrutinize the coverage surrounding a specific issue or movement. Furthermore, there are limitations to the theoretical foundation of the protest paradigm. Consequently, operationalisation of the paradigm and the way results have been interpreted differ substantially across studies, which has even led to contradicting findings regarding one protest event in the past. The thesis uses a novel dataset of all articles published in eight national UK newspapers between 1992 and 2017 about domestic protests and demonstrations (N = 27, 496). To analyse coverage in this large corpus, I use an innovative approach to framing analysis that combines best-practice manual coding techniques with supervised machine learning. Using this approach provides a strong methodological and theoretical foundation for the analysis of protest coverage: the operationalisation of frames is more explicit than in existing studies of the protest paradigm and frames are found inductively from the data, rather than being derived from decades old theory. The analysis shows that a stable majority of articles uses frames linked to the protest paradigm throughout the time frame. At the same time, a substantial and growing number of articles employ legitimising frames — either on their own or co-existing with delegitimising framing. Specifically, I find seven distinct frames: four that follow the delegitimising patterns of the protest paradigm, two frames that legitimise protests and their message and one that is neutral. The results show that patterns of reporting about protest are not static and that the circumstances and features of protest events shape their coverage. Specifically, I find four main determinants for the use of the different media frames: (1.) violent protests get more delegitimising coverage, and less legitimising coverage; (2.) the goal of a protest matters for the kind of reporting it receives, yet relationships between frames and goals are complex and goals overall matter more for legitimising frames; (3.) protests receive less legitimising coverage from tabloid newspapers than from broadsheet outlets and one of the legitimising frames is used less often by right-wing media — which means that differences between outlet categories exist but are less pronounced than expected; and (4.) reports published more recently and longer after the start of an event have a higher chance of containing legitimising framing. Overall, the thesis adds to existing knowledge on how the media frames protest over time and provides insights into the conditional logic with which journalists use different frames. Moreover, it develops a new approach to framing analysis combining manual and automated content analysis.
See the accompanying blog post for a summary of the main findings of the thesis.