Belgium is probably the world’s best known case of where caretaker governments reside. Yet a clear scholarly definition and measurement of this concept is missing. Based on a detailed analysis of the Belgian federal cabinets, this research note explores the main characteristics and measures the length of the various caretaker periods. We find that Belgium was governed for no less than 1,485 days by a caretaker government between 2007 and 2020, which equals more than four full calendar years. This research note also presents a novel typology of caretaker periods based on the institutional and political practice within the Belgian legislative and executive branches. This typology can be used to assess caretaker periods at other levels of government as well as in other countries in order to improve our understanding of the many ‘faces’ that a caretaker government can take on.

This Research Note investigates party advertising in newspapers and on social media (Facebook) during the 2019 general elections in Flanders, the largest region of Belgium. The 2019 elections saw a marked increase in the use of social media advertising by parties, whereas newspaper advertising saw a decline. Prior research that compares multiple types of advertising, particularly advertising on social and legacy media remains limited. As such, based on a quantitative content analysis we investigate not just the prevalence of party advertising on both types of media, but also compare the level of negativity, presidentialisation, and issue emphasis. Our analysis reveals substantial differences: we find that not only the type of advertisements varies across the platforms, but also that social media ads tend to be more negative. Finally, parties’ issue emphasis varies substantially as well, with different issues being emphasized in newspaper and Facebook advertisements.

This research note investigates the role of paid digital campaigning in the 2018 local elections in Flanders. We make use of the official declarations which candidates are legally required to submit. In these declarations, candidates indicate whether and how much they invested in online campaigning tools during the four months preceding the elections. We collected data on a sample of 3,588 individual candidates running in the 30 municipalities of the Leuven Arrondissement. A multilevel logistic regression model shows that the odds of spending on digital campaigning increases among incumbent aldermen and local councillors. The latter finding supports the normalization thesis of digital campaigning. The results also show that scale is important – the more potential voters a candidate has, the higher the odds that the candidate invests in digital tools.

Sub-constituency campaigning occurs when parties focus their campaign resources on specific geographical areas within an electoral district. This behaviour was traditionally thought to occur only in single-member plurality elections, but recent research demonstrates that proportional systems with multi-member districts can also elicit sub-constituency campaigning. However, most studies of sub-constituency campaigning rely on self-reported measures of campaigning, not direct measures of campaign intensity in different regions and communities. We present novel data on geographical variations in the intensity of Flemish parties’ campaign advertising during the 2014 general elections in Belgium, which provides a direct measure of sub-constituency campaigning. Our findings show clear evidence of sub-constituency campaigning: parties campaign more intensely in municipalities where they have stronger electoral support and in municipalities with greater population density.