Women are underrepresented at all levels of government (Center for American Women and Politics 2023), which affects representation (Gerrity et al. Reference Gerrity, Osborn and Mendez2007), the behaviour of public officials (Dodson Reference Dodson2006), and citizens’ responses to government (Campbell and Wolbrecht Reference Campbell and Wolbrecht2006; Stauffer Reference Stauffer2021). While other barriers for women to be elected office exist (for example, voter biases [Karpowitz et al. Reference Karpowitz, Monson, Preece and Aldridge2024] and elite support [Barber, Butler, and Preece Reference Barber, Butler and Preece2016]), one key barrier to representational equality is gender differences in ambition (Fox and Lawless Reference Fox and Lawless2014; Fulton et al. Reference Fulton, Maestas, Maisel and Stone2006; Preece Reference Preece2016).
Recent studies have highlighted elections as reducing ambition among women (for example, Kanthak and Woon Reference Kanthak and Woon2015; Koltveit Reference Koltveit2022; Preece and Stoddard Reference Preece and Stoddard2015), suggesting the gender distribution of office seekers might be fundamentally different without elections. However, studies asserting elections’ critical role in shaping ambition do not distinguish whether elections change or merely reinforce gender differences in ambition.Footnote 1 Indeed, there are reasons to doubt elections are unique in deterring women. Women are underrepresented in non-elected political positions such as the bureaucracy (Bishu and Headley Reference Bishu and Headley2020) and the judiciary (American Bar Association 2023).Footnote 2 Yet, with few exceptions, scholarship on gendered ambition has focused on candidate emergence and ambition for elected office (for example, Crowder-Meyer Reference Crowder-Meyer2020; Fox and Lawless Reference Fox and Lawless2005), largely ignoring ambition for other forms of public service (see Ammassari et al. [Reference Ammassari, McDonnell and Valbruzzi2022], Badas and Stauffer [Reference Badas and Stauffer2023], Bauer and Darkwah [Reference Bauer and Darkwah2019], and Koltveit [Reference Koltveit2022] for exceptions).
In this work, we test whether the presence or absence of elections changes women’s ambition for public service by examining the relationship between gender and interest in high-profile elected and non-elected positions in government. Using samples of federal employees, law students, and the general public, we find gender differences in ambition for elected office are similar to those for influential unelected offices.
However, because these positions in public service vary beyond merely the presence or absence of elections, we also separate public service into its various components and examine whether gender predicts aversion to the electoral process and other aspects of public service. We find gender has a similar relationship with the (un)attractiveness of elections as it does with the (un)attractiveness of other components of public service. Consistent with work on gendered political socialization (Bos et al. Reference Bos, Greenlee, Holman, Oxley and Celeste Lay2022; Fox and Lawless Reference Fox and Lawless2014, Preece Reference Preece2016), our work indicates women are deterred by multiple components of public service. Rather than unique, elections are duplicative in dissuading women from public service.
Political Ambition and Electoral Politics
Even if elections deter women from running, adding or removing elections may have no effect on who is interested in public service because those deterred by electoral procedures might also be deterred by other components of public service. Yet previous work does not provide a definitive answer. The two previous attempts to answer this question, of which we are aware, examine gender differences in ambition for elected and appointed positions among young party activists in Norway (Koltveit Reference Koltveit2022) and in Australia, Italy, and Spain (Ammassari et al. Reference Ammassari, McDonnell and Valbruzzi2022). Yet, because the appointed positions studied (party workers or advisors) are lower profile and operate behind the scenes, differences in ambition could be the result of elections or position salience. Indeed, Koltveit (Reference Koltveit2022) finds significant negative effects of gender on ambition for the most salient appointed position (state secretary) while finding no gender effects for less salient appointed positions (advisors and hired party staff). Moreover, there are gender gaps in ambition for clerkships (unelected positions) within the US court system (Badas and Stauffer Reference Badas and Stauffer2023).Footnote 3
Moreover, previous work has not examined whether aspects deterring women from elected office might also discourage non-elected public service. From a young age, women are socialized to see political leadership positions as a ‘man’s world’ (Bos et al. Reference Bos, Greenlee, Holman, Oxley and Celeste Lay2022). Generally, agentic values of being ‘competitive’ or ‘outgoing’ (generally viewed as more masculine traits) are more congruent with the ethos of politics (Conroy and Green Reference Conroy and Green2020). In this vein, women generally are disinclined towards public speaking (De Paola et al. Reference De Paola, Lombardo, Pupo and Scoppa2021) and less inclined towards leadership roles (Alan et al. Reference Alan, Ertac, Kubilay and Loranth2020; Ertac and Gurdal Reference Ertac and Gurdal2012), which are key components of public service.
Ambition for Public Service
To examine the effects of the presence or absence of elections, we rely on a general population sample and two samples of political elites.Footnote 4 In addition, we use two different methodological approaches. In the first approach, we examine ambition by gender for elected and unelected positions with similar leadership responsibilities and of similar public salience. In the second approach, we break public service into its various components to examine whether respondents have the same affinity or averseness towards both the electoral and the non-electoral components of public service. Below we first detail the different samples before analyzing the results from the two different approaches.
Sample #1: Law Students
The first sample is a survey of students enrolled at eight different law schools.Footnote 5 Links to the survey were distributed to students by administrators and faculty at each school, along with brief explanations. We intentionally identified a diverse set of public and private law schools and focused on law schools where we had personal connections to current students, former students, faculty, or administrators to increase the likelihood of being able to distribute the survey.
The sample of law students was recruited during the spring of 2022. While not representative of all law students, the law schools include both public and private law schools and are diverse in their ideological leanings and reputational rankings.Footnote 6 Though we do not know the total number of students invited to participate, we received 522 responses overall.Footnote 7
We survey law school students for two important reasons. First, law students are individuals for whom public service is realistically feasible as a law degree opens opportunities in the bureaucracy, the judiciary, and, potentially, elected office. Second, their training exposes them to the demands of these various positions.
Sample #2: Federal Bureaucrats
The second sample is of federal government employees. Like law students, bureaucrats are aware of the demands of public service. In particular, these individuals have extensive knowledge about bureaucratic public service, arguably the type of public service in our analysis least familiar to most people. We also use this sample because it is an elite sample with a wider age range as young adults exhibit particularly high levels of political ambition. Government employees also have the education and income to make public service a plausible endeavour (Fox and Lawless Reference Fox and Lawless2005).
The sample of bureaucrats comes from the second wave of the American Government Employee Survey (AGES), fielded in January/February of 2019. The primary purpose of the survey was to gauge the effect of the government shutdown on federal employees participating in a prior survey wave (n=3,053 in the initial wave). Recontacting original participants and inviting them to answer additional questions related to the federal shutdown resulted in 2,084 responses.Footnote 8
Sample #3: Nationally Representative Sample
Lastly, we also use a general population sample. Because law students and bureaucrats likely understand better the responsibilities of public service in non-elected positions, electoral components might be less influential in their considerations of public service. For this reason, we also use a nationally representative sample of US adults.
Survey Sampling International (SSI) recruited 1,777 subjects to participate in an online study in September 2019. The sample is comparable to other nationally representative samples and estimates from the American Community Survey (see the appendix for more information).
Measures
In all samples, the ambition instrument was the same with the exception of the question about ambition in the federal bureaucracy for the bureaucratic sample. We asked respondents, ‘How would you characterize your interest in _____ in the future?’ Respondents indicated their interest in serving in elected office, the judiciary, and the state and federal bureaucracy. Our judicial ambition measure used the words ‘becoming a judge or a justice’ in the blank, the measure of federal (state) bureaucracy ambition used the phrase ‘running or overseeing federal (state or local) agencies’, and the measure of electoral ambition used ‘holding elected office’. For each question, respondents could indicate they (1) had ‘absolutely no interest’; (2) were ‘open to the possibility’; or (3) were ‘actively working toward’.
Because bureaucrats were already serving in the federal bureaucracy, we asked about progressive ambition – as opposed to nascent ambition – within federal agencies specifically asking their interest ‘in moving up the ranks in federal agencies’.
Figure 1 shows overall levels of ambition for each sample and area of public service. In line with previous surveys of the general public, nascent ambition for elected office among the public is low. We also find low levels of nascent ambition for other leadership positions in public service in the judiciary and the bureaucracy. Law students and federal bureaucrats, who have put themselves in positions where public service is possible, harbour greater ambitions for public service. Further discussion of these differences is available in the online appendix.
Figure 1. Proportion Expressing Electoral Ambition by Sample.
Note: Bars are the proportion of respondents in each survey who indicated ambition.
Sources: Law Student Survey 2022. American Government Employees Survey (2019). SSI 2019 National Survey.
Elections and the Types of Individuals Who Express Ambition for Public Service
While there is some variation across samples in the extent to which gender correlates with ambition (as might be expected given self-selection into becoming law students or federal bureaucrats), we are interested in whether there are gender differences in ambition across different areas of public service within samples. Figure 2 shows the effects of gender on political ambition for different types of public service for federal bureaucrats; Figure 3 shows analogous effects for law students; and Figure 4 shows effects for the general population. For ease of interpretation, we present results from Ordinary Least Squares (OLS) regression models with the dependent variable coded 1–3, but results are nearly identical when we use ordered logistic regression models. All models include controls for age, race, education, political ideology, party identification, political knowledge, and political interest, and, in models using the sample of law students, whether the law school was public or private.Footnote 9 Full results and alternative ordered logistic model results are in the appendix.
Figure 2. Effects of Gender on Ambition Among Bureaucratic Sample.
Figure 3. Effects of Gender on Ambition Among Law Student Sample.
Figure 4. Effects of Gender on Ambition Among General Population Sample.
The results from all three samples suggest elections do not fundamentally change the negative relationship between gender and public service ambition. In the bureaucratic sample (Figure 2), women are less likely to express ambition for elected office. They are also less likely to express ambition for judicial service or for the bureaucracy. For law students (Figure 3) women are less likely to express ambition for all types of public service as the coefficient on gender is consistently negative, although the effects for federal bureaucratic leadership are not significant. Although gender effects on ambition for state and federal agencies are smaller, the effect of gender on ambition for judicial office is not significantly different than that for elected office. For the general population (Figure 4) we find consistently sized gender gaps in ambition for all forms of public service. Because some states have judicial elections, we also reran models using the general population sample excluding individuals living in states where there are judicial elections, which are very similar to what we present in Figure 2. In short, across all three samples, women are consistently less likely to express ambition for all types of public service regardless of whether the means to entry includes elections.
Gender and Different Components of Public Service
While our previous analysis indicates gender does not differentially affect ambition for different positions in public service (both those with and without elections), there are other differences in these positions beyond the electoral components that could confound our analysis, potentially mitigating the effect of eliminating the electoral component. As such, we break public service into various components to examine whether electoral components of public service generate similar responses from women as other aspects of public service. This allows us to check whether the effects of elections duplicate the effects of other aspects of public service.
The logical yet untested explanation for the similarity in gender effects on ambition for elected and unelected positions is that women perceive the desirability of various elements of public service similarly.Footnote 10 For example, individuals who are willing to engage in the glad-handing essential to political campaigns might also be more attracted to legislative and policy negotiations. Thus, individuals attracted (averse) to electoral components of public service might also be attracted (averse) to other aspects of public service.
To examine this possibility, we separate public service into various components and examine how gender affects the attractiveness of those various components. Utilizing data from 211 interviews about what aspects of public service are attractive or unappealing derived from a snowball sample of the personal contacts of undergraduate students at a Western US university, we generated a list of twelve items considered appealing or unappealing about public service.Footnote 11
We included these twelve items in the 2019 SSI survey and asked respondents to identify whether each aspect of public service would make them less likely, neither/neutral, or more likely to run for office. To group these items into larger components of public service, we conducted exploratory factor analysis with Promax rotation. Table 1 shows the items with their loadings.
Table 1. Attractiveness of Public Office Loadings
The results suggest interest in public office lines up along three dimensions. The first of those components relates to aspects of personal life and includes the lack of privacy, public scrutiny, difficulty on family/friends, and schedule/travelling. The second is directly related to electoral concerns and includes competition, fundraising, and voter persuasion. The third component relates to the job responsibilities of public service such as talking about solutions and decision making. These constructs are also internally consistent (Cronbach’s alpha of 0.84 for the personal life components, 0.81 for the electoral components, and 0.70 for the job responsibility components).
In the surveys of both the general population and of law students, respondents indicated whether each item would (1) make them less likely to run, (2) neither/neutral, or (3) more likely to run (3). Using these responses, we create an additive index consisting of items for each component. We added the rating of each of the items for each factor and divided it by the number of items. Each of these variables is scaled 1–3, with higher scores indicating a higher likelihood of seeking public office.
Figures 5 and 6 show the effect of gender on the attractiveness of the components of public service for the general population and the law student population, respectively.Footnote 12 As before, full models are available in the online appendix.Footnote 13
Figure 5. Gender’s Effect on the Attractiveness of Aspects of Public Office (General Population Sample).
Figure 6. Gender’s Effect on the Attractiveness of Aspects of Public Office (General Population Sample) (Law Student Sample).
The results in Figures 5 and 6 show the negative relationship between gender and the electoral components of public service is not unique. For both samples, gender is negatively correlated with the attractiveness of each component of public service. Women’s aversion to elections duplicates aversion to the personal life and job responsibility aspects of public service.
Conclusion
Although elections may seem like a prominent component that by themselves, if eliminated, might increase women’s ambition for public office, our results suggest this is not the case. Removing elections as a method for selecting public servants does not appear to fundamentally change women’s ambition for public service because elections appear to duplicate the negative effects of other aspects of public service on ambition for public service among women.
Our findings instead are consistent with existing research on gendered socialization into political leadership positions (Bos et al. Reference Bos, Greenlee, Holman, Oxley and Celeste Lay2022; Fox and Lawless Reference Fox and Lawless2014, Preece Reference Preece2016). Although women may be averse to competition, the same experiences and socialization that leads them to avoid competition also affect how they perceive themselves as fitting into the political realm and their comfort in engaging in the various responsibilities of public service. While elections are a deterrent to women’s political ambitions, so too are other components of public service in the public sphere. Rather than unique in their influence on ambition across genders, elections reinforce the attractiveness, or the lack thereof, of other components of public service to women. Increasing ambition for public service among women thus requires larger changes and cannot be achieved by merely eliminating elections as pathways to public service.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0007123424000449.
Data availability statement
Replication data for this article can be found in Harvard Dataverse at https://doi.org/10.7910/DVN/3BXMON.
Acknowledgements
The manuscript benefited from valuable feedback provided by colleagues during its development. We thank Meredith Conroy, Laura Dean, and Katey Stauffer for their insightful comments on the manuscript. We also thank Sarah Bertschy, Andrea Boyack, Jennifer Espinola, Corrie Evans, Haider Hamoudi, Elissa Karim, Kathryn King, Jake Linford, Brad Lingo, Parker Luke, Scheherazade Mason, and Dane Thorley for their assistance in distributing the law student survey. All errors remain our own.
Author contributions
Authors are listed in alphabetical order. The authors contributed equally to this manuscript. All authors share first authorship.
Financial support
None.
Competing interests
None.
Open access
