Respondents interviewed in the months preceding election day report significantly lower levels of life satisfaction; once voting has taken place, aggregate well-being immediately returns to its regular average

Changes in Well-Being Around Elections. Nicolas Schreiner. WWZ, Working Paper 2021/03, January 21, 2021. https://edoc.unibas.ch/81295/1/20210126100126_600fda669dc58.pdf

Abstract: Elections constitute the essential element of democracy, yet surprisingly little is known about their immediate consequences for individual well-being. Cross-country empirical evidence is particularly absent for the campaign period leading up to elections. While elections as a process allow citizens to contribute to democratic quality, they are also intrinsically conflictual and require voters to exert effort to make informed decisions. To measure the aggregate changes in well-being along the entirety of the electoral process, I use survey data from before and after 148 national elections in 24 European countries between 1989 and 2019. Respondents interviewed in the months preceding election day report significantly lower levels of life satisfaction than their compatriots asked the same calendar week but in years without elections. Once voting has taken place, aggregate well-being immediately returns to its regular average. Exploratory analyses suggest that partisan conflict and social pressures regarding democratic participation may play a role in explaining the reduction in life satisfaction before elections.

JEL classifications: D72, D91, I31

Keywords: elections, well-being, life satisfaction, election campaigns, electoral systems, political polarization, eurobarometer

6 Conclusion

This chapter contributes to the study of well-being and elections by providing the first systematic cross-country and long-term evidence for a link between the two. The absence of research on satisfaction in the public during election campaigns so far seems to be a particularly striking void in the study of democracy. As public involvement and interest in politics arguably peak during this period, it would appear vital to know how the populace is affected thereby. My empirical analysis shows that the apex of the campaign season is, in fact, the single span of time around elections during which aggregate well-being systematically and significantly changes. During the final month(s) before election day, average life satisfaction is substantially lower than it would otherwise be. These negative effects do not persist after voting takes place, however. Thus, elections appear to reduce aggregate well-being before the day of the vote without producing similarly systematic short-term gains afterward. These findings should also be taken into consideration in future research regarding the consequences of elections, as the time span before election day likely represents an unsuitable control period to compare outcomes after the election against. Global dissatisfaction with democracy is at or near historic highs (Foa et al., 2020). Hence, now more than ever, it would be vital to understand the mechanisms for why the electoral process seems to regularly herald a season of public unhappiness instead of being a time of civic celebration. Based on some very preliminary exploratory analyses, social conflict and polarization could prove to be promising starting points in this regard. Designing electoral institutions that curtail the formation of stark political cleavages and affective polarization, such as greater proportionality in legislative allocation, could thus potentially play a role in halting or even reversing the aforementioned trends. Furthermore, institutions that influence the costs of voting and political information (including surrounding social norms), may offer fruitful opportunities for research into the causes of lower well-being during election campaigns. However, given the suggestive evidence so far, it does not appear that voters become unhappy because they are overwhelmed by the complexity of the choices they face in elections. To this point, based on the existing research on the effects of direct democracy on civic engagement (see, e.g., Benz and Stutzer, 2004), it would also be interesting to study whether my findings for representative elections translate to ballot measures. Yet, even if it turns out that the negative consequences of election campaigns are unavoidable, this would not speak against elections as a mechanism for societal decision-making. Rather, the short-term reduction in life satisfaction during the months immediately preceding an election may be understood as the investment costs required for reaping the long-term profits of democracy in terms of well-being.