Constructing Crises, 1785-2015

Crises appear to be ubiquitous. But what is a crisis? A crisis is a sudden negative event or development concerning something important and that requires urgent action. This definition is not fixed; and what is regarded “negative”, “important”, “urgent”, and “reaction-craving” in the public’s eye is subject to social construction and social change. Digitalized newspaper archives are an invaluable data source that can provide insight into how societal notions of public “crisis” have evolved. What constitutes a crisis is highly consequential. Public communication and political opinion formation work differently in times of “crisis” (Perse, 2001), when society switches into a sort of “emergency mode”—where swift and unanimous action is required. But this “emergency mode” can be exploited politically (Bennett et al., 2007; Entman, 2004). It is therefore highly important to understand how societies define the ambiguous term “crisis” and apply it over time. Towards that goal some basic questions must be answered: (1) Has the share of “crisis coverage”, compared to total coverage, increased over time? (2) Which topic areas are often covered in “crisis mode” and which are not? (3) How has the prevalence of different topic areas shifted over time? This study traces the news coverage in the Times (London) from 1785-2015 that features the term “crisis” or related terms (n=128 861). The share of “crisis” coverage (compared to all coverage) has increased from less than 1% in 1800 to roughly 3.5% in 2015 (Figure 1). The 120 topics in a structural topic model (Roberts et al., 2019) were classified into fourteen topical areas of crisis coverage. The largest topic area is geopolitics (20%). The next two issues that are often covered in “crisis mode” are “illness/epidemics/health” and “disasters/accidents/events”, that account for 12% each. These are classical areas of “crisis mode” where sudden negative happenings require a quick response. Economics topics (ECO, IND, FIN) reach 19%. Over time, we see that many issues exhibit a non-linear development, e.g. geopolitics has seen an increase up until the 1950s before a sharp decline. Disasters and illness/epidemics/health had their high-tide during the 19th century and have lost in prominence since. Only wars have seen a steady decrease in prominence and not even the two World Wars did reverse the downward trend. Macro-economic crises and financial crises exhibit a steady increase in prominence (Figure 3). Limitations of the data and the analysis are discussed in detail.