Elections under the umbrella: examining the effect of weather on turnout in first-order and second-order elections

Elections are one of the cornerstones of democratic societies that offer an insight into how many people want to actively participate in the political process. Meanwhile, early studies dealing with political behaviour showed that a factor seemingly unrelated to politics - the weather – also affects turnout. Moreover, it has become a common theme of discussions of election results in the media and among ordinary people. Yet, there is still little evidence about the political conditions under which the weather impacts participation. This paper aims to fill the gap and answer whether the type of election could mediate the effect of weather on turnout. More specifically, is the weather effect in first-order elections smaller than in second-order elections? The question stems from the idea that when comparing two different types of elections in the same geographical area, the weather effect should be stronger in elections where there is 'less at stake'. The benefits of voting are not as high, and it is easier for the voter to forgo the benefits and not participate in the election. Using turnout data from 2,000 municipalities, 31 elections held in Slovakia between 1994 and 2023, and ERA5 weather data, I investigate the role of the election type in the weather-turnout hypothesis. Slovakia is a parliamentary democratic republic with the most important parliamentary elections and four others' (presidential, local, regional, and European) less at stake elections. Notably, the country exhibits significantly different turnout rates in first-order and second-order elections, with discrepancies repeatedly exceeding 40 %. The findings of this study aim to illuminate the potential differential impact of rainfall, temperature, and "real" temperature on turnout based on the perceived importance of the election. In conclusion, the research contributes to understanding the nuanced interplay between weather conditions and voter behaviour, highlighting the mediating role of election types in shaping the weather-turnout hypothesis. The outcomes offer valuable insights into how varying levels of election importance may moderate the impact of weather on voter turnout.