If you really love nothing. Evaluating second-order factors in the case of Italian constitutional referendum of 2020

The paper evaluates several second-order factors to explain the outcomes of the 2020 constitutional referendum held in Italy. This particular referendum serves as a relevant case study to assess the significance of second-order factors, given its unique characteristics, including the simplicity of the referendum question regarding the reduction in the number of parliamentarians (MPs), the minimal impact on public finances resulting from this reduction, and the subsequent decrease in territorial democratic representation. While it might have been expected that a significant majority would vote in favour of the "No" option, thus preserving the current levels of territorial coverage (and democratic representation) of MPs, the actual results saw a substantial majority in favour of the "Yes" vote (69%). Our argument posits that the overwhelming prevalence of the "Yes" vote (to reduce the number of MPs), especially in poorer areas of the country, can be attributed to specific factors that influence individual evaluations in a direct-democratic setting. In greater detail, by using cross-sectional data, the paper tests the role of socio-economic condition, trust in institutions and political participation in affecting the referendum outcome. The results of our empirical analysis confirm our hypotheses, demonstrating that second-order factors indeed influenced the referendum's outcome. Specifically, our analysis reveals that: (i) a higher socio-economic condition could generate a higher share of "No" votes; (ii) a higher trust in institutions could lead to an increase in the share of "No" votes; finally, that (iii) an increase in political participation could produce a decrease in the share of "No" votes. In the concluding section of the paper, we discuss how this analysis contributes new insights to the study of voting behaviour in direct-democratic contexts.