Towards a modern gender gap in Europe?

Abstract

This article examines electoral gender differences in Western Europe. It describes the electoral gender gap in a longitudinal cross-national design (1974–2000, 12 countries) with EuroBarometer data. The analysis shows that gender differences in voting in Europe follow a similar movement to what has been described for the USA. Women tended to vote more for conservative parties in the 1970s, while in the new millennium they have given higher support to left parties. The speed of this development differs cross-nationally and not all countries reached the state of a modern gender gap (where women lean left).

At the national level the driving force for the emergence of a modern gender gap is mainly increased levels of female labor participation. At the individual level structural variables explain traditional gender differences in voting, but fail to account for the gender gap in recent years.

Notes

1 In this paper the terms “traditional gender gap” and “women's conservatism” are used synonymously.

2 In this paper the terms “modern gender gap” and “gender realignment” are used synonymously.

3 In a most-similar-system design it is reasonable to believe that the questions are understood in an almost identical way by all respondents.

4 EuroBarometer is a survey conducted on behalf of the European Commission at least two times a year in all member states of the European Union. The survey provides regular and varied information on social and political attitudes in the European publics, but also includes some permanent questions such as, for example, the vote intention.

5 The countries included in this analysis are: Belgium, Denmark, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain and the United Kingdom.

6 Newer comparative data on vote intentions is provided by the European Social Survey (ESS) 2002 and 2004 and the International Social Survey Program (ISSP) 2002 and 2003, but not on a yearly basis.

7 The category left includes the socialist, left-socialist and green parties as coded by Armingeon, Gerber, Leimgruber and Beyeler (2008).

8 Another possibility would be to employ left–right self placement or an expert placement of parties on the left–right axis instead of party families. However, the author is convinced that the vote choice captures the ultimate outcome better than a self placement on the left–right axis. Expert placements of parties on the other hand do not provide more than the party family affiliation which is used in this study (McDonald, Medes, & Kim, 2007).

9 Several tests with the data for Spain revealed that there is a problem with nonresponse, particularly of women. Therefore, the data should be interpreted very cautiously.

10 Due to nondemocratic periods in their history prior to 1981 (Greece) and 1985 (Portugal and Spain).

11 Tables with the coefficients for all independent variables can be obtained from the author on request.

12 The country dummies are significant as a whole. According to Kittel (2005) a fixed effects model is the only solution in this case as the model is biased without the inclusion of country dummies. The country dummies are not reported in the table.

13 The exact operationalization of the independent variables can be obtained in Appendix A.

14 The correlation between the variables is 0.78 (Pearson's correlation coefficient).