Contemporary election campaigns seldom involve candidates praising rivals. Widely documented in Western democracies, negative campaigning is an entrenched campaign tactic in Britain. As such, scholars have extensively explored what drives negative campaigning and its impact on prospective candidates and the electorate. This paper adopts a novel approach to unpack the conditions, dynamics and impact of campaign negativity in scarcely researched multi-party second-order elections. Specifically, I comparatively examine the primary drivers, intra- and inter- party behaviour and electoral impact of negative campaigning across traditional and digital media channels in British by-elections.
While recent studies have looked beyond the United States to explain the consequences and conditions which motivates candidate attacks in European multi-party systems (Maier, Nai and Verhaar, 2024; Martin and Nai, 2024; Walter, 2013), four substantial research limitations persist. First, studies prioritise first-order elections to examine the determinants of negative messages (Mendoza, Nai and Bos, 2024), glossing over second-order elections as bagatelle to more politically meaningful general elections. This yields weaker insights into explaining party behaviour and dynamics in multi-party systems with different election contests. Second, most research use single case-studies as the focal point (Haselmayer, 2019), with fewer comparatively examining this phenomenon over many elections to develop greater systematic insights (Maier and Nai, 2022; Maier, Nai and Verhaar, 2024). Third, most content analyses of negative campaigns rely on a single data source, such as leaflets (Duggan and Milazzo, 2023), social media content (Auter and Fine, 2016), or candidate self-reports (Maier and Nai, 2021). To my knowledge, no studies have hitherto triangulated data sources across traditional and digital communication channels to identify negative messages. Finally, extant studies frequently treat candidates as unitary political actors in elections, or shift focus to party leaders to measure negativity. Previous research has thus failed to account for the intra-party dynamics and variations of party behaviour both within and across competing parties in elections.
Stemming from these limitations, this paper fills these lacunae by conducting a comparative longitudinal analysis of negative campaigning across 54 contested British by-elections held since 2010. Drawing on an original dataset of approximately 80,000 X posts, 36,200 Facebook posts, 5900 Meta adverts and 1500 candidate leaflets used in 54 contested British by-elections since 2010, this paper provides the first study to systematically examine the drivers, dynamics and impact of party behaviour in multi-party second-order elections in Britain. All major and minor British parties are included (n=18), and incorporates data from national party and local (constituency) party pages, as well as candidate (n=277) and party leader (n=62) online pages across Facebook and X. In doing so, this work contributes to several strands of literature. It extends studies on negative campaigning beyond first-order elections, emphasising the unique contextual factors impacting party behaviour during by-elections. Methodologically, I triangulate data sources to provide greater insights into the development of party communication strategies within the fourth-era of political campaigning, map variances across traditional and digital technologies, whilst shedding light on intra-party campaign dynamics in multi-party systems to address a significant gap in the electoral studies literature.
