Representation of Women in Second-order Elections: The Czech Republic and Slovakia Compared

Abstract

Elections to the European Parliament (EP) fall into the category of second-order national elections. In these types of elections voters behave differently when casting their vote. But what about political parties: do they behave differently in EP elections than in national parliamentary elections? This article draws on available electoral data related to women's representation in the Czech Republic and Slovakia to explicitly link the second-order election (SOE) model to independent actions of political parties. The results show that women candidates have (1) more positions and better ranking on ballot lists; as well as (2) better chances of getting elected when running in SOEs in Slovakia but not in the Czech Republic. In order to explain this variance in results, the most similar system design (MSSD) was applied. The findings from the MSSD indicate that three variables have the potential to explain this variance in representation of women in EP elections between the countries: parties’ positions on European integration dimension, the numerical size of national legislature, and preferential voting. The variable parties’ positions on European integration have the highest explanatory power.

Key words:

• European Parliament
• second-order elections
• political parties
• women's representation
• first-order elections

Notes

1 In the countries under analysis (i.e., the Czech Republic and Slovakia), this condition is satisfied, as both types of elections were fought within almost identical party systems.

2 In elections to the EP as well as NP, Czech voters may use as many as two preferential votes (Parliament of the Czech Republic 1995, 2003a ). In EP elections, Slovak voters may use as many as two preferential votes, whereas they may use up to four preferential votes in national parliamentary elections (National Council of the Slovak Republic 2003b, 2004 ).

3 We also performed the statistical tests of the differences. In case of Slovakia, we can strongly reject (with p-values <0.05) the null hypothesis of equality of the means in favor of alternative hypothesis that the mean is higher in SOEs. For Czech Republic the null hypothesis cannot be rejected. Nevertheless, the reader should keep in mind, that due to the structure of our data, the statistical results cannot be taken strongly.

4 Given that we are concerned only with the comparison of aggregate (country) levels of women representation across SOEs and FOEs, we do not report individual party-level results here. Nevertheless, Appendices 1 and 2 provide an overview of party-level results so that an interested reader can get some knowledge of party-level trends in women representation across SOEs and FOEs in both countries.

5 The only notable difference between the electoral systems used concern the number of constituencies in the Czech Republic (14 in national and 1 in EP elections) and method for allocating seats (Droop formula for Slovak EP elections and Hagenbach-Bischoff method for national elections and D'Hondt method form both elections in the Czech Republic.

6 Both parties however use their informal gender policies for national and EP elections.