Traditional or Gender-Neutral Leadership Templates? A Meta-Analysis

• In Event: Gendered Attitudes: Origins & Implications

Abstract

Women are still under-represented in government relative to their proportion of the population, even in advanced industrialized democracies that are viewed as consolidated. Much gender politics research explores reasons for this on-going under-representation, focusing on the impact of electoral institutions or party ideology. Other research focuses on preferences of party leaders, to maintain the male power, or party leader beliefs about how voters will react to female candidates. Little work, however, particularly outside of the US, explores when and how voter attitudes are gendered. Yet we know that people have mental templates of what a leader “looks like”. This paper will use data from parallel experiments conducted in 8 countries (10 study sites) to carry out a meta-analysis of leadership templates yielding a cross-national analysis of when people have gender-neutral leadership templates. With NSF funding we conducted an experiment with young people in Canada (Alberta and Quebec), Chile, Costa Rica, England, Israel, Sweden, Uruguay, and the U.S. (California and Texas) resulting in a total of 6,318 study participants. Analysis of individual countries allows us to explore which types of people (based on party preference, ideology, socioeconomic background, media attentiveness, and gender role beliefs) view female candidates as leaders vs. who has traditional leadership templates that view leadership as masculine. Cross-country analysis allows us to explore how national contexts (the government policy agenda, experience with women in positions of power in government, electoral rules) are associated with gender neutral vs. traditional leadership templates. Analysis will include both different levels of posts (legislator, cabinet minister) and different types of policy areas (stereotypically masculine and feminine). The countries included in this study were selected to include national policy agendas that are expected to be favorable to women’s leadership (i.e., an emphasis on social welfare policy) or very challenging for women rising to top leadership posts (i.e., a focus on national defense and security). The countries included in the study also vary based on whether or not women have held many and diverse posts in national government, and in the type of electoral rules used (closed- and open-list PR, SMD, whether there is a national gender quota law). Meta-analyses typically use aggregate data from already published studies that use different methods and study designs. We are able to use individual participant data, clustered in their particular study location, because all 10 experiments use the same experimental design and comparable research material, based on Matland (1994). This allows us to evaluate the magnitude of the response to candidate gender compared to the magnitude of the response to our baseline control: candidate party.

Authors

• Michelle M. Taylor-Robinson, Texas A&M University

• Nehemia Geva, Texas A&M University