Mayoral Selection and the Demand and Supply of Women Mayors

Abstract

Few studies consider the effect of selection procedures on women's attainment of mayoral office. This article begins to fill this gap through analysis of more than 100 cities in California. The results show that selection method of mayor by popular vote exerts a substantial negative effect on whether a woman holds the office of mayor. The study then explores the negative effect of popular vote on women's attainment of mayoral office. There are two possibilities for exploring this effect: (1) popular vote creates more opportunity for the negative influence of stereotyping and role socialization on women's recruitment and election to mayoral office or (2) alternatively, popular vote increases the negative influence of sex differences in political motivation on women's attainment of mayoral office. The analyses support the second interpretation of the negative effect of popular vote on women's mayoral office-holding.

Keywords:

• women
• mayors
• mayoral selection
• popular vote
• candidate motivation
• California

Notes

Amy C. Alexander is now at Leuphana University, Lüneburg, Germany.

1. I compare appointive and rotating selection procedures as one category rather than two separate categories to selection by popular vote. The differences between these two procedures are minor compared to the differences between these and selection by popular vote. When collecting data across the cities in California, city codes and city Web sites were often unclear as to whether the procedure operated according to rotation, appointment, or some combination. Moreover, my primary information source for this variable, the California Roster, supplied information only on whether the mayor was or was not a popularly elected office.

2. I also obtained data for the percentage of Republican registered voters in my sample of cities. However, these two variables correlated above .90, and I therefore had to choose only one to use in the regression model. I used the strength of the correlations between the two measures of partisanship and the dependent variable to determine the best measure to include in the model.

3. I dummy-code an election selection procedure one and an appointive selection procedure zero.

4. For this variable, a woman in the office of mayor was coded one, a man zero. One important concern in coding sex is how the researcher handles gender-neutral names. When I came across a gender-neutral name, I ran an Internet search of the name, the office-holding position, and the city. This was a very successful approach. In almost every case a local article, profile, or picture was uncovered that indicated the sex of the individual in question.

5. With descriptions of experience as mayor from just ten women, this information is gathered from a very small number of cases. Twenty mayors in total were contacted, ten elected and ten appointed in spring 2007. These women were selected with the goal of capturing the diversity of cities in California for the narrative data. The ten women who responded represented cities that reflected a wide range of diversity. The population levels varied from 200,000 to 10,000 residents. The average median incomes of these cities varied from $104,000 to $34,000. The diversity of the populations of these cities also varied. Some cities were dominantly white, two had high Latino populations, and one city had a high population of Asian residents. In addition, half of the cities were predominantly Democratic in their partisanship and half were predominantly Republican. The cities also varied geographically, including Northern Los Angeles County, San Diego County, Fresno County, and South Orange County.