The “social eligibility pool” stands as one of the most common, and most powerful, explanations for women's underrepresentation in elective office.1
For the purposes of this essay, I limit my interest in the pool to occupational background rather than previous officeholding experience. The term “eligibility pool” may include lower-level offices, particularly if the office has a well-defined opportunity structure or ladder (Schlesinger 1966). This is known as the “pipeline” problem facing women in politics. To the extent that previous officeholding is a valued credential for a given office (e.g., president, senator, member of Congress, or governor), women confront a serious structural problem resulting from their relatively recent entry into the pool of lower-level offices (Duerst-Lahti 1998). Whether women run for the legislature has implications for state policymaking, but women's candidacies have additional consequences: State legislative office is a stepping-stone to congressional and statewide office, making women state legislators a natural pool of women poised to run for higher office.
The eligibility pool varies somewhat by the type of office and by state, and some women officeholders were homemakers or had careers in female-dominated professions (Darcy, Welch, and Clark 1994). Nevertheless, gender differences in occupational background are believed to pose a substantial obstacle to increasing the presence of women in public office. R. Darcy, Susan Welch, and Janet Clark (1994, 112) explain that “it is safe to say that women's occupations and activities have not provided the same sort of gateway to political office as prestigious male occupations.” This account predicts that the level of women's representation will inevitably increase in tandem with women's gains in the professions and improvements in women's general socioeconomic position (Darcy, Welch, and Clark 1994; Seltzer, Newman, and Leighton 1997). Indeed, the percentage of women in the labor force and the percentage of women lawyers are usually positively correlated with the presence of women state legislators (Norrander and Wilcox 1998; Williams 1990).
By all indicators, women continue to make gains in the eligibility pool (Costello, Wight, and Stone 2003). Women are increasingly educated and are moving into the professions in record numbers. Women's presence in elite politics has increased as well: Women comprised fewer than 5% of state legislators in 1971, but today they are over one-fifth of legislators (CAWP 2005b). More women have also won statewide office: Women were fewer than 10% of statewide elective executive officeholders in the early 1970s but are now over one-quarter of these officeholders (CAWP 2004, 2005a).
Yet a close inspection of the 2004 election results reveals that women's progress in achieving elective office in the states has slowed. The 2004 elections continued the plateau in women's representation that began in the mid-to-late 1990s. Incumbency remains an obstacle to women's candidacies (Burrell 1994; Darcy, Welch, and Clark 1994). Yet there are many more open seats in state legislative races than congressional ones. After the 2004 elections, women comprised 22.5% of state legislators, in keeping with the level of women's representation since 1999 (Carroll 2005). Thus, although more than one-quarter of state legislative seats turned over in the 2004 elections, the presence of female legislators did not increase.2
The 2002 turnover data were provided to me by the National Conference of State Legislatures.
The plateau in women in state legislative office—which is usually an entry level office—was not predicted by the social eligibility pool explanation.3
Most state legislators have no prior elective officeholding experience (Pew Center on the States 2003).
I borrow the idea of a “women's path” from the “widow's path” that women have frequently taken to Congress (Gertzog 1995).
Taking Gender Differences in the Path to Office Seriously
Scholars have long observed that many women take a different path to office compared to men (Burrell 1994; Carroll and Strimling 1983; Diamond 1977; Dolan and Ford 1997; Kirkpatrick 1974; Thomas 1994, 2002). Today's women legislators come from a more diverse set of backgrounds and more professional backgrounds than the women who served in the 1970s and earlier decades (Dolan and Ford 1997; Thomas 1994), but gender differences in the path to office are hardly limited to the past (Burrell 1994).
Indeed, more than a few women officeholders have backgrounds that differ from their male counterparts. A Center for American Women and Politics (CAWP 2001) survey found that the lawyer's path to office is much more likely to be taken by men than by women: 19% of men state legislators but only 10% of women are attorneys. Meanwhile, 22% of women but only 7% of men are school teachers or administrators, and 8% of women but only 1% of men are nurses or health-care professionals (CAWP 2001). Thus, twice as many women legislators hail from education as from legal fields. Gender differences are evident in the biographies of members of Congress as well, with men much more likely to be lawyers than women, and women more likely to be educators than men (Burrell 1994). On the whole, state legislators come to office from a wide range of occupational backgrounds (Hirsch 1996). The percentage of legislators who are attorneys has declined, though they remain the single largest occupational group at 16% of legislators (ibid. 1996).
If we revisit the qualifications question from the perspective of state legislators' actual backgrounds, the pool of potential female state legislative candidates increases dramatically. Despite the movement of women into nontraditional fields, occupations remain highly segregated by sex (Reskin and Roos 1990). According to data from the U.S. Bureau of Labor Statistics, for example, women were 27.6% of lawyers in 2003 (U.S. Bureau of the Census 2004), yet women were 73.8% of workers employed in education-related occupations, including 81.7% of elementary and middle school teachers, 55.2% of secondary school teachers, and 44.9% of postsecondary teachers (U.S. Bureau of the Census 2004). Women were only 29.9% of physicians and surgeons, but 92.1% of registered nurses. The total number of those employed in education and health-care occupations also greatly exceeds the number of lawyers. Meanwhile, women in government jobs provide another source of candidates: women were 56.8% of government workers in 2000 (Costello, Wight, and Stone 2003).
Moreover, the time demands of legislative service and, frequently, relatively low legislative compensation mean that the opportunity costs of serving in the legislature may discourage officeholding. The fact that women's earnings are, on average, lower than those of men, however, may make legislative service of disproportionate interest to women. Precisely because the opportunity costs of service are likely to be lower for women, women may be more interested in pursuing state legislative office than men (Maddox 2004). In short, the states provide numerous opportunities for women candidates.
Studying Change in Informal Qualifications
Rethinking the informal qualifications for political leadership is as large a project as rethinking gender itself. It is important to acknowledge the hurdles of launching a political career from a female-dominated occupation. Susan Carroll and Wendy Strimling (1983, 5) observed that the “standards by which we evaluate qualifications for public office-holding are defined by men's experiences.” The standards of voters, political parties, interest groups, political action committees (PACs), and donors may be difficult to change. Carroll (1993, 204) acknowledged that women from female-dominated occupations “may have to work harder to prove that they are ‘qualified.’ ” Female-dominated occupations tend to be lower in prestige and lower paid than male-dominated ones (Reskin and Roos 1990). Individuals in these careers may also lack flexible work schedules, which may make it harder to combine outside employment with officeholding (Carroll and Strimling 1983). The personality traits deemed most desirable in politicians are also linked to gender and to gender roles, which poses an additional challenge for women candidates (Deaux and Lewis 1984); traits associated with men are typically seen as more important to officeholding than those traits associated with women (Huddy and Terkildsen 1993; Rosenwasser and Seale 1988).
Because women are unequal in society and have fewer socioeconomic resources, they may be less attractive candidates in comparison with men (Chapman 1993). The idea that women have a lesser role in politics because of gender inequalities in social status is a parsimonious and persuasive account. Political elites do not typically represent a broad range of social groups (Aberbach, Putnam, and Rockman 1981; Eldersveld 1989; Matthews 1983; Prewitt 1970; Putnam 1976; Seligman et al. 1974).
But ideas about politicians do change. This suggests that our theories need to specify the conditions under which the image of the “typical politician” changes. What factors can fuel an increase in the pool of women who are considered “qualified” for public office? An obvious answer, and the one proposed by the social eligibility pool account, is advances in women's professional standing and educational attainment. This route does not necessitate a change in beliefs about qualifications but, instead, an increase in the number of qualified women.
Yet there are other routes by which the eligibility pool can become more gender inclusive. For example, one such mechanism is changes in gender-role attitudes and greater acceptance of women in politics (Dolan 2004; Ferree 1974). The survey question often used as a benchmark of voter bias against women candidates is a hypothetical one about whether voters would support a qualified woman presidential candidate. Gallup asked voters in 1937: “Would you vote for a woman for President if she were qualified in every other respect?” (Falk and Jamieson 2003, 46). Times have changed, and women are no longer thought to be automatically disqualified for office by virtue of gender; indeed, the updated survey question recently used by Roper asks about voting for a woman for president if nominated by your party “if she were qualified for the job” (Falk and Jamieson 2003, 46). As the electorate has become more accepting of women, women should be treated equally by voters. Indeed, more liberal gender-role attitudes and ideology are typically positively related to the pattern of women's officeholding in the states (Brace et al. 2002; Norrander and Wilcox 1998).
Meanwhile, some electoral contexts present opportunities for female candidates. One such example is the so-called 1992 Year of the Woman, in which women benefited from the Anita Hill–Clarence Thomas hearings as well as from anti-incumbency sentiment (Carroll 1994; Cook, Thomas, and Wilcox 1994; Duerst-Lahti and Verstegen 1995; Paolino 1995; Witt, Paget, and Matthews 1994). The idea of balancing the gubernatorial ticket by including a female candidate for lieutenant governor has become popular in recent years (Fox and Oxley 2004). Today, only eight women serve as governors; yet, 15 women serve as lieutenant governors, making it the most common statewide elective executive office held by women (CAWP 2005a). Under some conditions, then, being a woman may itself be a desirable candidate qualification.
Occupations and Qualifications
I propose that a missing piece of the puzzle in studies of gender and candidacy is attention to a particular group of potential women candidates: women from female-dominated occupations. In their discussion of the social eligibility pool, Darcy, Welch, and Clark (1994, 180) observed: “More research is needed on how women can enter political life using their existing backgrounds and experiences.” Few scholars have pursued such a line of inquiry. Because of the prevalence of gender differences in the path to office and continued occupational sex-segregation, however, this is a pressing research agenda for scholars interested in women's representation. The very definition of the social eligibility pool is endogenous to women's representation. If the social eligibility pool includes not just the traditionally male-dominated professions that have usually been stepping-stones to office but female-dominated occupations as well, then a shortage of “qualified” women candidates for entry-level public offices may not cause women's underrepresentation after all.
This suggests an investigation of the conditions under which women are able to take a “women's path” to office, by which I mean a background in female-dominated occupations. What explains changes in the desired qualifications for office? Under what conditions do conceptions of the social eligibility pool change? What mechanisms expand the social eligibility pool to be not only more inclusive of women but also more inclusive of women from education, nursing, and other predominantly female fields? If women, like men, are strategic (Fulton 2003; Pearson and McGhee 2004), then we would expect to see more women run for office from female-dominated occupations when the structure of political opportunities is more favorable.
I briefly outline three directions for research about candidacy and female-dominated occupations. Changes in the issues on the states' agendas, changes in recruitment patterns, and changes in the numbers of women in office may increase the likelihood that women can take a women's path to the state legislature.
New Issues and Changes in Candidate Criteria
We know that issue context can help women's election to office. Women candidates may benefit from voters' stereotypes about the personality traits and issue competencies of politicians (Dolan 2001; Herrnson, Lay, and Stokes 2003; Kahn 1996; Schaffner 2005). In a similar vein, Fox and Oxley (2003) suggest that some statewide elective executive offices are “feminine,” or in keeping with women's traditional areas of expertise, compared to “masculine” offices that are more consistent with stereotypes about men's abilities and issue competence. For example, in 2005, only five women serve as attorney general in the states, but 10 women serve as chief state education official (CAWP 2005a). Because of voter stereotypes, then, we can expect that women are more likely to be candidates in some years and for some offices than others, and that women may want to accentuate “female” campaign issues.
How beliefs and context shape women's opportunities has also been examined from the vantage point of political culture. Voters may be more comfortable with women in politics in settings where the political culture is more public-spirited than individualistic (Diamond 1977; Elazar 1984; Hill 1981). Eileen McDonagh (2002) also makes a cultural argument in her cross-national study, suggesting that where state policies recognize the significance of women's reproductive labor by providing social benefits for individuals engaged in care work, more women hold office.
I propose a related line of inquiry about how issues are related to women's opportunities for office. The mechanism identified by most of these studies is the fit between voter stereotypes and women candidates. An important mechanism that may connect the issue context to women's officeholding is the qualifications for office. Changes in issue context may expand the notion of the eligibility pool to include women from female-dominated professions.
For example, demographic changes such as suburbanization or federal policy changes could increase the salience of education as an issue in the states. In turn, an occupational background in education—a female field—may be a more attractive candidate qualification. Thus, an increase in the salience of policy debates related to the traditional expertise of women should lead to changes in the candidate pool.
Changes in Recruitment Patterns
The composition of the social eligibility pool may also be altered by the political actors engaged in recruitment. One might expect, for example, that changes in the involvement of interest groups or PACs tied to female-dominated occupations in electoral politics could have implications for the types of credentials that are sought in candidates. Similarly, were the gender and/or occupational background of party, interest group, or PAC leaders to change, ideas about which individuals are qualified might expand to include women from predominantly female occupations. Recruitment patterns and occupational background may also vary by party due to occupational differences between Democrats and Republicans (Sanbonmatsu 2002).
Meanwhile, party competition, party realignment, or candidate scarcity may lead to changes in the opportunity structure facing potential candidates (Canon 1990; Diamond 1977). Shifts in the normal recruitment patterns may lead to a search for new types of candidates, which may create more openings for women candidates with experience in female-dominated fields.
Changes in legislative institutions may also alter the typical qualifications sought in officeholders. For example, the state legislatures have become more “professionalized” and more similar to the full-time institution of the U.S. Congress (Rosenthal 1989; Squire 1988, 1992, 1997; Thompson and Moncrief 1992). As legislative service becomes more of a full-time job and less of a part-time job, the occupational backgrounds of legislators may become more diverse; it may no longer be necessary to combine legislative service with additional employment, such as a law practice (Squire 1992). Other institutional changes such as term limits may also have implications for occupational backgrounds.
The “Critical Mass Hypothesis” with Respect to Candidacy
One mechanism that facilitates women's election to office is the precedent of female officeholders. An historical legacy of women's officeholding can undermine voter stereotypes and reduce bias against women, reshaping the public's expectations about what a politician should look like (Diamond 1977; Dolan 2004; MacManus 1981; Mandel 1981). In addition, Kanter's (1977) “critical mass” theory about sex ratios within organizations has led to a growing body of work that investigates whether a critical mass or certain threshold of women within legislatures is needed in order for gender differences in behavior to emerge (Beckwith 2002). According to this logic, when women are vastly outnumbered within legislatures, there is more pressure to conform to dominant norms, reducing the likelihood that women will act “as women” on behalf of their women constituents. Thus, increasing the presence of women within the institution may enable women to pursue a wider range of activities.
Extending this logic to the realm of candidacy, we might expect that greater numerical representation of women provides more freedom for women to pursue office from a wider range of backgrounds. In those settings with a stronger track record of women in public office, it may be easier for women to run for the legislature from female-dominated occupations. Thus, numbers may be related to qualifications. If this dynamic is correct, an increase in female legislators could subsequently lead to more heterogeneity among the types of women who run for office. Indeed, the idea of critical mass has received some criticism for assuming a link between descriptive and substantive representation; after all, increasing the presence of women legislators may lead to more diversity among women legislators and less cohesive behavior by women as a group (Reingold 2000). In sum, the level of women's representation may have implications for the types of backgrounds from which women can legitimately seek office.
Research on these questions would require data on the occupational backgrounds of potential candidates, candidates (both winners and losers), and legislators. In general, we know very little about how voters use candidates' occupations to make inferences (McDermott 2005). Are women successful when they run for the legislature from a female-dominated occupation? Does the likelihood of candidacy from female professions depend on the track record of female candidates with those biographies? Aggregate election results cannot reveal how occupational background affects candidates' chances or how these relationships have changed over time. Instead, we would want to examine individual-level data. We would want to consider how gender and occupational background shape electoral results by analyzing who does and does not run, as well as how various candidate backgrounds affect electoral success (Black and Erickson 2000).
It is worth noting, however, that we cannot assume that the pool leads directly to officeholding. Jennifer Lawless and Richard Fox (2005) find that women in the eligibility pool (i.e., lawyers, educators, business people, and political activists) are much less likely to consider running for office than similarly situated men. Meanwhile, studies of political parties reveal that the dearth of women party leaders, the existence of gender bias, and gender differences in social networks reduce the likelihood that the parties will recruit women candidates (Niven 1998; Sanbonmatsu 2006b). Thus, the relationship of the eligibility pool to officeholding is contingent rather than automatic.
Conclusion
Prior to winning suffrage, women used their location in the private sphere as a justification for their political involvement (Baker 1984; Cott 1987; Skocpol 1992). Women's family responsibilities and housekeeping role translated into concern for social provision in the public sphere. These social reform and housekeeping issues often served as the basis for the candidacies of the few women who ran for office in the early twentieth century (Andersen 1996). For much of the twentieth century, the accepted pathways for women's officeholding were uniquely female; women achieved office as housewives, widows, and from women's occupations (Carroll and Strimling 1983; Gertzog 1995; Kirkpatrick 1974).
The bases from which women can launch political careers have expanded: Women now have more paths available to them with the enactment of equal educational and employment laws. But inequities in the economy persist, with important consequences for politics. The dearth of women in the social eligibility pool is a well-recognized structural barrier to women's representation. However, if women in female-dominated occupations are also eligible, then we have discovered a large supply of potential candidates.
Instead of treating the contours of the eligibility pool as given, we need to revisit the very notion of the pool and identify the conditions that lead to changes in how the pool is defined. What types of electoral conditions and institutional settings have been conducive to women's pursuit of political careers from female-dominated occupations?
As we look ahead to the 2008 elections in the states, we should reconsider whether women are truly as underrepresented in the pool as we too often assume. Waiting another generation for women's share of the eligibility pool to expand seems misguided when a pool of qualified women is already waiting in the wings. After all, as Jeane J. Kirkpatrick (1974, 60) observed: “To become a legislator it is only necessary to win an election.”
