Women in Executive Office: Variation Across American States

• Author: Zoe M. Oxley,Richard L. Fox
• Published date: 01 March 2004
• Date: 01 March 2004
• DOI: http://doi.org/10.1177/106591290405700109
• Subject Matter: Articles

Why do some political systems elect more women

than others? Addressing this question is impor-

tant for a number of reasons. First, alleviating

gender inequities in both symbolic and substantive re p r e-

sentation is only possible if we know the cause of the

inequities. Second, wide disparities in the election rates of

women (across nations and within the United States) are

v e ry common. Among state executive offices—the focus of

our study—only fourteen states have ever elected a woman

g o v e rn o r, while in Arizona women were elected to all five

top executive offices in 1998 (CAWP 2001). A number of

social, structural, and political factors have been identified to

explain differing election rates of women t o American state

l e g i s l a t u res (Ford and Dolan 1998; Nechemias 1987; Nor-

rander and Wilcox 1998; Rule 1981, 1990, 1999; Sanbon-

matsu 2002; We r ner 1968). We employ a conceptual frame-

work of candidate re c r uitment developed by Pippa Norr i s

(1997) to summarize the findings from these legislative stud-

ies. Norris developed her framework to explain why cert a i n

candidates run for and win election to national parliaments

and legislatures. While not necessarily gender-specific, her

framework can and has been used to examine disparities in

the election of women across nations (Norris 1993, 1997).

Using Norris’ framework, we then derive several expla-

nations for the varying election rates of women to state

executive offices (such as Governor, Attorney General, State

Treasurer, Superintendent of Education).1In the modern era

of devolution, these offices have a growing power and

prominence in the implementation of public policy, yet they

have been relatively unexamined by both state politics and

gender politics scholars. Using an original data set that

encompasses twenty years of executive office elections (from

1979 until 1998), our analyses illuminate the experiences of

women pursuing these offices. The application of theory

and findings from the comparative and state legislative liter-

atures also allows us to assess whether prior explanations

for variation in the election of women are generalizable

across types of office.

PREVIOUS EXPLANATIONS FOR VARIATION

IN THE ELECTION OF WOMEN

In her model of candidate recruitment, Norris (1997)

identifies four types of explanations for why candidates

come to run for and win elective positions: the political

system; the recruitment processes of the political parties; the

supply of candidates; and the demands of gatekeepers. We

organize the research regarding women in state legislatures

around her four categories. The political system explanation

focuses on the configuration of legal, electoral, and party

structures within a state. These are “rules of the game”

(Norris 1993: 312), but the rules do not apply neutrally to

all candidates. Some facilitate while others hinder the elec-

tion of women. For example, women are more likely to

serve in state legislatures with multimember districts (Mon-

crief and Thompson 1992; Norrander and Wilcox 1998;

Rule 1990) or with smaller district populations (Nechemias

1987; Rule 1990). Women have been more likely to serve in

positions with low prestige and fewer professional benefits

(Diamond 1977; Nechemias 1987; Sanbonmatsu 2002; but

see Norrander and Wilcox 1998). Recent research has also

examined whether term limits h ave provided w omen

greater access to public office; to date, the findings are

113

Women in Executive Office:

Variation Across American States

ZOE M. OXLEY and RICHARD L. FOX, UNION COLLEGE

The number of women serving in state-level executive office varies tremendously across the American states.

Drawing upon a comparative politics framework developed by Pippa Norris and findings from analyses of

women in U.S. state legislatures, we derive a set of hypotheses to explain this state variation. Our analysis of

elections held between 1979 and 1998 demonstrates that women are more likely to run for executive office in

states where more women are in the eligibility pool of candidates, and where the demands of gatekeepers and

recruiting practices of political parties favor women’s candidacies. Furthermore, the likelihood that women win

these elections is influenced by the supply of candidates, the demands of gatekeepers, and the characteristics

of a state’s political system. We also conclude that the predictors of women in executive office have changed

over time and that our explanations for state variation of women in these positions are more thorough for elec-

tions occurring before 1991.

1Throughout this article, we focus on only elective executive officials in

the American states, rather than all executive officials (some of whom are

appointed).

NOTE: An early version of this article was presented at the 2001 annual

meeting of the American Political Science Association. This

research was supported by a grant from the Dean of Arts and Sci-

ences, Union College. For assistance with data collection and

management, we thank Mark Anderson and Dale Miller. We are

also grateful to Kira Sanbonmatsu, Barbara Burrell and the anony-

mous reviewers for their useful comments on prior drafts.

Political Research Quarterly, Vol. 57, No. 1 (March 2004): pp. 113-120

mixed (Carey, Niemi, and Powell 2000; Carroll and Jenkins

2001; Sanbonmatsu 2002).

The election of women is also influenced by the recruit-

ment processes of the political parties. Direct examinations of

how internal party recruitment practices affect women in

the American states have been quite rare (Burrell 1993;

Niven 1998) and quantitative analyses have tended to use

indirect measures of recruitment processes, such as the level

of state party competition. Democratic-dominant states con-

tain fewer women in their legislatures than do states where

the Republican Party is dominant (Nechemias 1987; Nor-

rander and Wilcox 1998; Rule 1999), likely because of the

lack of effort among some Democratic state parties to recruit

women candidates (Deber 1982; Fowlkes, Perkins, and

Tolleson Rinehart 1979). Furthermore, if one party domi-

nates elections in a state, regardless of which party it is,

women will be less likely to serve in the legislature (Nelson

1991). Finally, there is some evidence that in states where

the parties have more control over candidate nominations,

fewer women serve in the legislature (Nelson 1991; but see

Norrander and Wilcox 1998 and Sanbonmatsu 2002 for

mixed results).

The final two components (supply of candidates and

demands of gatekeepers) highlight a set of social and cultural

explanations regarding the inclusion of women in politics.

The supply of candidates explanation refers to the number

of individuals able, willing, and well situated to seek public

office. States with a “traditional” culture, which conforms to

the historic patterns of gender role socialization that pre-

cluded women from entering the public role required of

politics, have fewer women serving in office (Hill 1981;

Nechemias 1987; Rule 1990). Conversely, there are more

women in the legislatures of states characterized as having a

culture that emphasizes fairness and equity, or a moralistic

culture (Norrander and Wilcox 1998). A state’s supply of

women in the pipeline professions (such as lawyer or busi-

ness executive) that often lead to a career in politics is also

positively related to the presence of women in state legisla-

tures (Darcy, Welch, and Clark 1994; Norrander and Wilcox

1998; Rule 1990; Sanbonmatsu 2002).

The demands of those specifically charged with candidate

selection also affect whether women serve in elected posi-

tions. These gatekeepers are the voters, party members, finan-

cial supporters, and political leaders who select among the

pool of candidates. Bias against re c ruiting women candidates

for state legislatures has been found among some state and

local party leaders (Niven 1998). As for voters, in states where

the electorate is more conservative ideologically and/or less

s u p p o rtive of feminism, there is less support for women to

s e rve in the state legislatures (Norrander and Wilcox 1998)

and in the U.S. Congress (Fox and Smith 1998).

The findings cited above provide evidence supporting

each of Norris’ four explanations for why the percentage of

women state legislators varies across American states. In this

study, we examine whether these categories are also useful

for explaining state variation in women executive officials.

Our research also addresses other topics. First, we simulta-

neously include predictors for each category, and are thus

able to examine whether any of the four explanations is

more important than the others. Second, we assess whether

the predictors of women’s electoral success were consistent

between 1979 and 1998. Third, we separately analyze two

stages of the electoral process: the selection of candidates for

the ballot and the election among nominated candidates.

This approach departs from the state legislative studies,

which use only the percentage of women elected to legisla-

tures as their measure of women’s electoral success.

EXPECTATIONS REGARDING STATE EXECUTIVE ELECTIONS

On average, 16.2 percent of state executive off i c i a l s

between 1979 and 1998 were women. But, there is substan-

tial variation across the states (see Table 1). In six states,

women have never served in executive office, while in New

J e r s e y, Colorado, Indiana, Arizona, and Delaware, women

have held more than 35 percent of these offices. Furt h e r-

m o re, in 31 states the rates of women in legislative and exec-

utive offices are similarly high or similarly low.2For example,

c o m p a red to the national averages, women were more likely

to be legislators and executive officials in Arizona, Kansas,

Minnesota, and Connecticut, but less likely to serve in either

in Mississippi, South Carolina, New York, and Utah. There

a re exceptions to this trend. In nine states, the percentage of

women legislators is high while the percentage of women

executives is low. The opposite pattern—above average exec-

utives and below average legislators—occurs in ten states.

To explain these state disparities in the election of women

to executive office, we drew upon the state legislative

re s e a rch to identify a set of predictors for each of Norris’ four

b road categories. We would expect these predictors to be re l-

evant for executive office elections, because most states (31,

or 62 percent) have either above average rates of women

s e rving in executive and legislative offices or below average

rates for both. Specifically, we hypothesize that states with

the following characteristics should have higher numbers of

women executive officials: term limits;3Republican part y

dominance; moralistic political culture; liberal public atti-

tudes and public support for feminism. In contrast, women

executive officials should be rarer in states with more pow-

e rful executive offices, larger electorates (for these off i c e s ,

state populations), a single dominant political part y, part y

c o n t rol over nominations and a traditional culture .

Executive offices do have unique characteristics that dis-

tinguish them from other types of elected offices. To account

for these diff e rences, we incorporate a few other pre d i c t o r s

into our analyses. Under the category of political system, we

114 POLITICAL RESEARCH QUARTERLY

2Figures for the percentage of women legislators in each state were calcu-

lated from the Center for the American Woman and Politics Fact Sheets

(CAWP, various years).

3While the terms for governor have been limited in some states for many

years, term limits for other executive offices were rare until the 1990s.

Thus, we incorporate this predictor into our analyses only for elections

occurring after 1990.