Estimating the trade-off between higher turnout and a more representative election result

Abstract A growing body of research has used exogenous shocks in voter turnout to analyze the relationship between turnout and election results. However, the findings of these studies with regard to the legitimacy of election results remain unclear because an increased turnout does not necessarily translate into an election result that is more representative of the entire electorate. We developed a new three-step method for elucidating the trade-off between turnout and the representativeness of election results and applied it to a referendum held in the Netherlands in 2018. First, we used regression analysis to decompose election results into the voting preferences of two voter groups: those who voted irrespective of any turnout shock and those who voted because they were affected by the shock. Second, we obtained an estimate of a fully representative counterfactual election result using the results of a survey. Third, we estimated how the exogenous shock affected the representativeness of the election result and analyzed the trade-off between turnout and representativeness.