Policies for women participation in politics: assessing different territorial impacts

The paper addresses the issue of gender equilibrium in politics and how the outcomes of gender rebalancing electoral mechanisms may be assessed. It relies on data and results of an evaluation of the effects produced by regulatory measures introduced in Italy for gender rebalancing in the elective offices. Women still encounter huge difficulties in being included in political decision-making. Even when they have the chance of participating to political competition as candidates, this is not a guarantee that they are elected and assigned to important political positions. Many countries have introduced rules and mechanisms to accelerate the rebalancing process. The general feeling is that their effectiveness may vary a lot, because they need to be well designed and effectively implemented to achieve good results, so asking an important evaluation question. Different types of mechanisms may be designed to ensure gender balance in politics. The main adopted options (that can also be paired) are: Gender shares in the lists of candidates; Double/triple gender preference; Alternate list. Their effectiveness may vary a lot, because they need to be well designed and implemented to achieve good results. Also, effectiveness is conditioned by general electoral rules and by local cultural determinants. Although a statistical recognition of gender shares among candidates and elected representatives is possible for all institutions, the impact evaluation is limited to the case of municipalities. In the other cases, the lack of a credible counterfactual forces to limit to statistical analysis of observed change. On the other hand, in the depicted situation, the voting regulation for the administrative elections is a perfect context where to evaluate the effectiveness of gender rebalancing mechanisms. The paper will also discuss the "dynamic equality index", able to better assess the impact of measures, also considering a great heterogeneity in starting conditions. The evaluation is based on RDD approach around the threshold present in the law (5.000 inhabitants). The dimension of the window around the discontinuity is established through different techniques. The paper includes also falsification tests on covariates. In a more advanced version of this paper, spatial dimensions of the impact will be assessed as-well. The preliminary analysis shows very robust results: the impact is significant and positive for all the adopted approaches, even though the size of the impact varies.