Gendered dimensions of legislative turnover in Latin America

The proposed paper will focus on the analysis of gendered aspects of legislative turnover, understood as the rate of renewal of legislators, in 17 Latin American countries through quantitative analysis. While legislative turnover has received ample attention in research on Western Europe, other regions including Latin America have received far less scrutiny, materialized in some case studies. Seeing as the Latin American region has numerous countries utilizing various types of gender quotas, it lends itself to an analysis of the effects of institutional settings onto gendered turnover rates. The aim of the paper is to analyse to what extent gendered differences in turnover rates exist in the studied countries, and to what extent they are related to institutional settings. The first part of the paper will analyse the cross-country variation of representation of female legislators in Latin America. It will focus on how the share of female legislators has changed with the implementation of gender quotas, as well as whether or not female legislators have less experience compared to their male counterparts, given the historical underrepresentation of women in legislative bodies. While the implementation of gender quotas significantly increased the share of women elected to parliament in the region, the question of political experience and gendered turnover has so far not been explored in terms of cross-country comparisons. Secondly, the paper will use multilevel hierarchical regression models to analyse which institutional settings increase or decrease the legislative turnover gap between male and female incumbents. Multilevel hierarchical regression is used due to observations being nested in countries, and in some cases branching further into lower and upper chambers. The aim of this analysis is to find out whether or not gender quotas in particular increase legislative turnover rates both in the short-term perspective, marked by their implementation, as well as in the long-term, seen as the upholding of gender quotas through multiple legislative elections. Another focus will be on electoral system design, considering that female candidates and incumbents may benefit more from certain types of electoral systems. This is due to not only voter behavior likely being different in various electoral settings, but the electoral result of converting votes into seats for individual candidates likely also differring greatly, given that district magnitude in Latin America varies from single-member districts such as those used in the elections to the Senate of the Dominican Republic, up to the nationwide electoral constituency used for electing 120 members of the Peruvian Chamber of Deputies in 1995 and 2000. To study the aforementioned questions, a unique dataset of Latin American legislators is utilized. This dataset includes data on elected members of unicameral chambers, as well as lower chambers and senates in 17 Latin American countries that allow consecutive reelection, thus excluding Costa Rica. The dataset spans almost 40 years in some countries and includes data on over 20000 individual observations of elected legislators, thereby lending itself to more advanced quantitative modelling methods such as multilevel hierarchical regression.