Electoral Systems and the Legislative Behavior of Women – Using Parliamentary Questions to Measure the Substantive Representation of Women's Interests

Numerous empirical studies on women’s political representation show that female legislators more strongly focus on the representation of women specific interests than their male colleagues. Thus, it is often argued that a greater presence of women in parliaments (descriptive representation) should automatically lead to a higher representation of women’s issues (substantive representation). By now, however, our understanding of the causal mechanisms and the theoretical links between descriptive and substantive representation of women is severely limited. In particular, we know very little about the effect of institutional variables on the parliamentary behavior of women and under which conditions female representatives actually act more strongly in the interests of women. In order to take a first step to fill this research gap, this Paper analyzes the effect of the electoral system and the electoral incentive structure on the substantive representation of women by individual female legislators. Based on the assumptions of the Competing Principals Theory, it is theoretically expected that the substantive representation of women’s interest should be higher in proportional electoral systems and lower in single member district system (SMD). In SMD, a female legislator has to represent the interests of the majority of the people in a geographic district to increase her reelection chances. An exclusive representation of women’s interests and a simultaneous disregard of the male constituency could therefore be a dangerous strategy. The empirical analysis uses the German mixed electoral system as a quasi-experimental design and analyzes the substantive representation of women in the German Bundestag between 2005 and 2013. Due to possible contamination effects in mixed electoral systems, I use the reelection safety of a legislator in the two different tiers to operationalize the actual electoral incentive structure. A serious problem in the analysis of the individual behavior of female legislators is the operationalization and the measurement of substantive representation of women issues. Due to a high party discipline – especially in European national parliaments - previous measurements (e.g. bill sponsorship, speeches, committee membership of legislators) have been criticized to rather measure the ideological position of the political party than the preferences and the behavior of individual legislators. To reduce the impact of party discipline, I use parliamentary questions to measure the substantial representation of women by individual legislators. Because oral and written questions can be submitted by individual legislators (without much party oversight), these data provide a promising source for a valid measure of the individual behavior of female legislators. For each legislator, the number of submitted questions with a women specific concern is recorded. It is argued that legislators submit the more women specific questions the more they are interested in the substantive representation of women. Based on this original data set, the results of a zero-inflated beta regression show a significant interaction effect between gender and the electoral incentive structure. Female legislators only act more strongly on the behalf of women if their reelection is secured and if they do not depend on the representation of local interests from the constituencies in their districts.