Does gender matter for political leadership?

Responsible researcher: Viviane Pires Ribeiro

Article title: DOES GENDER MATTER FOR POLITICAL LEADERSHIP? THE CASE OF US MAYORS

Article authors: Fernando Ferreira and Joseph Gyourko

Location of intervention: United States

Sample size: 5,500 elections

Sector: Gender

Type of Intervention: Analysis of the impact of female participation in the executive branch

Variable of main interest: Female participation
in executive power

Evaluation method: Discontinuous Regression

Assessment Context        

Although women continue to be underrepresented in most economic and political positions, recent decades have seen a significant increase in the number of women taking on leadership roles in the public and private sectors in many countries. This shift has attracted the interest of economists and other social scientists who want to understand the implications of female leadership (or lack thereof, as the case may be) for public policy outcomes.

Local governments in the United States (US) have experienced a significant increase in female participation in politics. Until the 1970s, an insignificant number of women participated in local mayoral elections. However, female participation rose to about a third in mayoral elections before leveling off around 1995. The percentage of women who won mayoral elections increased from 2% in 1970 to over 15% in recent years. Female candidates typically had less than a 50% chance of winning in 1965 until the mid-1990s. After that period, that probability aligned close to the 50% line.

Intervention Details

Ferreira and Gyourko (2014) investigate the impact of female participation in the executive branch of US cities, using information from more than 5,500 elections held in 575 cities between 1950 and 2005. The study differs from existing research because it is the first to focus on women in executive leadership positions in the local public sector, not in legislative participation. The fact that mayors have executive power can make it easier to reallocate resources within a city to suit someone specific's political preferences. Legislators, on the other hand, have to negotiate with other representatives (possibly the executive) to pass legislation, so the additional impact of a legislator may not be as effective or may be noticeable only when large turnout shocks are observed.

The electoral data for mayors used in the study by Ferreira and Gyourko (2014) is an updated version of the sample described in Ferreira and Gyourko (2009). Most of the information is based on responses from a survey sent to all cities and counties in the United States with more than 25,000 inhabitants in the year 2000. The following information was requested: the timing (year and month) of all mayoral elections since 1950; the name of the elected mayor and the second-place candidate; the vote total and vote totals for each candidate; party affiliation; type of election and some additional information regarding specific events, such as runoffs and special elections.

Information on local public finance variables was combined with electoral data. Public finance data cover fiscal years 1950-2005 and were obtained from two different sources: the Historical Data Base of Individual Government Finances (1970-2005) and the Census Bureau City Finances Series (1950-1969). Local public finance variables include measures of revenues and taxes, expenditures (on current operations and capital goods), employment (full-time and part-time), as well as distributional data relating to shares of expenditures on labor, public safety, and parks and recreation.

Crime rates were combined with election data to estimate the potential effect of mayoral gender on the effectiveness of police enforcement: murder and assault (violent crimes), as well as robbery and theft (property crimes). Crime data is available at the police district level in "Uniform Crime Reporting" reports issued by the FBI and the Department of Justice.

Methodology Details

To investigate the impact of female participation in the executive branch of US cities, Ferreira and Gyourko (2014) use a regression discontinuity (RD) design to deal with the endogeneity of female candidacy to city characteristics. More specifically, the authors compare short- and long-term results in elections in which a female candidate barely beats a male candidate with those in which a female candidate barely loses to a male candidate.

A two-step procedure was employed to assign gender to mayoral candidates. First, all given names were matched to a Census list of common given names. If the given name was estimated by the Census to be of a specific gender more than 99% of the time, then the name was assigned to be of that gender. For example, Census data shows that more than 99% of all those with the name “Robert” were male. Therefore, any candidate whose first name is “Robert” is considered male. Approximately 80% of candidates had a distinctly masculine or feminine name based on this criterion. For candidates with ambiguous names (e.g., Casey, Pat, Leslie), the authors looked for evidence of the candidate's gender. This second stage was carried out through internet searches, emails and telephone calls. Internet searches were generally local government websites and local newspaper archives with articles and photos. Similar searches were carried out in cases where data on first names was missing.

Results

In contrast to most research on the influence of female leadership, Ferreira and Gyourko (2014) found no impact of gender on a variety of local outcomes, such as the composition of municipal spending and municipal employment, the size of municipal government represented by the total spending or employment, or local crime rates. These results are valid both in the short term and in the long term.

The authors also investigate whether the incumbent effect differs by gender of the mayoral candidate. Even if city characteristics are quite similar for closed elections, voters may still have predetermined (i.e., discriminatory) views about women, which could harm a female candidate's chances of winning the election. In this case, women who take office “randomly” according to the regression discontinuity approach should have greater unobserved skills than the corresponding candidate. This, in turn, could translate into greater relative probabilities of reelection, as these leaders would have the chance to demonstrate their superior political ability while serving as mayor. On the other hand, discrimination can still be powerful while women are in office, hurting performance and potentially leading to lower re-election rates. In this sense, the results suggest that female mayors are more politically skilled compared to male mayors. This gender gap in incumbent effect is considerable: at least 5 percentage points at an average unconditional female re-election probability of 56%.

Public Policy Lessons

According to Ferreira and Gyourko (2014), the election of female leaders can still generate important political repercussions, even in the absence of any impact on policy results. For example, it can increase other women's chances of success in the future. However, the authors' analysis concludes that randomly electing a woman as mayor does not produce higher success rates for other women in the short or long term. In the immediate future, the high re-election rates of women in office will naturally cancel out the participation of other candidates. No additional effects are evident a decade or two after the initial election. The authors also tested whether a female mayor affects female success rates in other elections, such as local electoral districts, but found no evidence of such repercussions.

Although female mayors have not implemented different policies, they appear to have greater unobserved political skills, as they have an incumbency effect at least 5 percentage points greater than a male candidate. However, the authors found no evidence of political repercussions: the exogenous election of a female mayor does not change the long-term political success of other female mayoral candidates in the same city or of candidates in local congressional elections.

These results suggest that the environments in which women have political power influence the relevance of gender to policies and political outcomes. For example, it can be more difficult to change policy when women take on leadership positions without the benefit of political quotas or reservations. Furthermore, the nature of the political and economic environment in which cities compete in the United States does not offer much room for redistributive policies, and local politicians may be more sensitive to the preferences of the median voter.

Intense competition among local governments can prevent gender from influencing policy outcomes, as it prevents political parties from implementing divergent policies. Furthermore, the types of policies relevant to local government can also play an important role in mitigating the potential political changes that result from the election of a female leader. Economic responsibilities such as local taxation and provision of basic services are the responsibility of municipal government, while more controversial social issues such as abortion and gun control are not.

References

FERREIRA, Fernando; GYOURKO, Joseph. Does gender matter for political leadership? The case of US mayors. Journal of Public Economics, vol. 112, p. 24-39, 2014.