6. Concluding remarks
This contribution to the literature on the analysis of electoral systems and corruption hypothesizes and tests that, besides a direct effect of the degree of proportionality of electoral systems on corruption, there is an important indirect effect: political competition is a channel through which electoral systems affect corruption. Italy represents a very interesting case study because, during the period under analysis, both the diffusion of corruption and the degree of proportionality of the electoral system varied widely. Constructed on the Senate electoral results, the Gallagher disproportionality index (GDI) was used as a measure of the degree of proportionality of the electoral system. The degree of political competitiveness among political parties is measured by the normalized Herfindahl index of votes concentration in the hands of political parties, always at the Senate elections. Multiplying the two political indexes an interaction variable has been created that captures what has been called the indirect effect of the degree of proportionality of electoral rule on corruption. Estimations show that both the effects matter in explaining corruption.
If it is assumed that political competition moves in the same direction as the degree of proportionality of electoral rule, the beneficial effect on corruption of an increase in the degree of proportionality of the electoral system is reinforced by an increase in political competition. This only holds true if the GDI is below a certain threshold, otherwise when the value of the GDI goes up to that threshold and the effect of political competition on corruption is inverted, the direct and indirect effects drive corruption in opposite directions.
Summing up, the final effect of the degree of proportionality of electoral systems on corruption depends on how political competition reacts to a change in the degree of proportionality. This last statement is, to the best of our knowledge, still unexplored. Therefore, to assume a variation only in the degree of proportionality is not sufficient to establish what is the consequent trend of corruption. Duverger's Law links an electoral system's change to the number of political parties, not to the competition among them. The latter, as discussed, depends on the number of parties and the votes’ market power.