In 2018, Delaware State Representative Helene Keeley told PBS that “when she was first elected … the major obstacle was being taken seriously by men.”Footnote 1 Describing how this phenomenon evolved over time, she explains that “in some cases, there is a questioning of your knowledge … I would say within the last five to 10 years, that was more of a hurdle, as opposed to strictly just having the respect you won the office and have the right to be there.”Footnote 2 While straightforward descriptive statistics paint a portrait of gender equality and women’s success in governing, a deeper look suggests that these equal outcomes mask substantial inequities in effort and talent. The idea that women feel they need to prove their competence and overcome marginalization in governing institutions is widely supported by both firsthand accountsFootnote 3 and an impressive array of work in political science (e.g., Michelle Heath, Schwindt-Bayer, and Taylor-Robinson Reference Michelle Heath, Schwindt-Bayer and Taylor-Robinson2005; Lazarus and Steigerwalt Reference Lazarus and Steigerwalt2018). Although women win congressional elections at rates similar to men, they have to be more qualified to do so (Milyo and Schosberg Reference Milyo and Schosberg2000). Once in office, they bring home more federal funds to the district (Lazarus and Steigerwalt Reference Lazarus and Steigerwalt2018), are more successful at sponsoring bills when in the minority party (Volden, Wiseman, and Wittmer Reference Volden, Wiseman and Wittmer2013), and receive more issue requests from their constituents (Butler, Naurin, and Öhberg Reference Butler, Naurin and Öhberg2022).
In this article, we explore a potential explanation for gender inequity in legislatures: that women in office may be held to a higher standard by fellow legislative insiders.Footnote 4 Legislatures are social environments that, in the United States, have overwhelmingly been comprised of men. Interviews with women legislators and their staffers suggest that women feel a pressure to overachieve in order to demonstrate that they are capable (Swers Reference Swers2013; Lazarus and Steigerwalt Reference Lazarus and Steigerwalt2018), and work in comparative politics suggests that men politicians keep women on the sidelines in legislatures (Michelle Heath, Schwindt-Bayer, and Taylor-Robinson Reference Michelle Heath, Schwindt-Bayer and Taylor-Robinson2005). Thus, it could be that legislative insiders undervalue women’s accomplishments, presenting numerous obstacles to women’s influence inside the chamber. Alternatively, legislators and those who work closely with them may be uniquely aware of the effort that goes into representational activities and can observe firsthand the work that their women colleagues perform. It may be the case that insiders are uniquely well situated to appropriately value women’s work.
In order to examine whether women’s accomplishments are noticed and appreciated by those working in the legislature, we explore how legislative and electoral accomplishments translate into elite evaluations, and whether that translation differs between men and women. Following Haynie’s (Reference Haynie2002) exploration of the relationship between race and perceived legislative effectiveness, we leverage data from the North Carolina Center for Public Policy Research on effectiveness rankings of North Carolina state legislators from 1993 to 2015 (Kaslovsky et al. Reference Kaslovsky, Koch and Olson2025). These scores are valuable in that they are created by having current state legislators and other legislative insiders rank the effectiveness of sitting legislators. We combine this subjective measure of effectiveness with a variety of measures that ought to capture in a more objective way whether a legislator is influential; these measures include bill sponsorship, committee assignments, and leadership positions (Bucchianeri, Volden, and Wiseman Reference Bucchianeri, Volden and Wiseman2024). These data allow us to explore whether men’s and women’s legislative and electoral accomplishments translate similarly into perceived effectiveness and influence among legislative insiders. If men and women who perform in comparable ways are “rewarded” differently by legislative insiders, this would provide suggestive evidence of a gender gap in expectations within the legislature. If, however, they are rewarded similarly, it would provide some evidence that legislative insiders are accurately perceiving and rewarding women’s work.
After developing our theoretical expectations and introducing our data, we first descriptively explore the evolution of subjective effectiveness over men’s and women’s legislative careers. We ask whether men’s and women’s perceived effectiveness evolve similarly as they gain seniority, and whether their paths to valuable institutional positions such as committee chairs and chamber leadership are similar. Following this, we use a regression-based approach to examine whether within-legislator changes in accomplishments—becoming a committee chair, for example, or accruing seniority, or being in the majority party—lead to a different “bump” in subjective evaluations for men and women. We conduct a within-legislator comparison, exploring whether perceptions change differently for men and women with comparable accomplishments. Notably, this analysis focuses on how perceptions adapt to achievements, and not on whether similarly situated men and women are evaluated equally. This question is critical for understanding bias within the chamber, as women may not be similarly situated throughout the legislature. Further, this approach allows us to hold unobservable characteristics—such as talent or ambition—fixed, which we could not do if we simply controlled for legislator traits and actions.
Across our analyses, we find little evidence that women are discriminated against by legislative insiders. First term men and women legislators appear to enter the chamber on equal footing and have relatively similar career arcs. And when men and women ascend to institutional positions or have similar success in the electoral arena, insiders adjust their evaluations accordingly. We emphasize, however, that biases at earlier stages of women’s political careers—selection into political careers (e.g., Fox and Lawless Reference Fox and Lawless2005), candidate recruitment by parties and interest groups (e.g,. Fox and Lawless Reference Fox and Lawless2010), and vote choice at the ballot box (e.g., Ono and Burden Reference Ono and Burden2019)—indelibly color our findings. Because a woman may need to be more capable than a counterfactual man in order to overcome these barriers, our null findings may in fact provide indirect evidence that bias is occurring (Knox, Lowe, and Mummolo Reference Knox, Lowe and Mummolo2020). Further, if women feel such pressures, even if legislative insiders are not applying them in their evaluations, they will continue to feel marginalized and have negative lived experiences within the chamber. We hope that our study provides a valuable contribution to a broader conversation on the barriers that women may face in achieving influence in legislative politics.
Gendered Expectations Outside and Inside the Legislature
We build our theoretical expectations for the relationship between legislator gender and elite evaluations of their effectiveness by drawing on the rich existing scholarship on women in electoral and legislative politics. This literature can be roughly thought of as divisible into studies that focus on the impact of gender outside and inside the legislature. We discuss these literatures in turn before articulating our specific hypothesis in the context of the literature on elite evaluations.
Outside the Legislature
Research in gender and politics establishes that women and men candidates win elected office at similar rates (Seltzer, Newman, and Leighton Reference Seltzer, Newman and Leighton1997; Sanbonmatsu Reference Sanbonmatsu2006). However, this does not mean that the electoral playing field is equal for all candidates. Women pursuing a career in politics still face significant gender-related barriers on the campaign trail. There is substantial evidence that women candidates tend to be more qualified than men running for the same position (Milyo and Schosberg Reference Milyo and Schosberg2000; Fulton Reference Fulton2012). Democratic and Republican women alike face high barriers to entry (Pearson and McGhee Reference Pearson and McGhee2013), and these gendered experiences extend into the sphere of campaign fundraising (Jenkins Reference Jenkins2007; cf. Burrell Reference Burrell2014). In particular, recent work shows that Republican women (Thomsen and Swers Reference Thomsen and Swers2017) and women of color (Sorensen and Chen Reference Sorensen and Chen2022) have a more difficult time securing financial support than men.
These electoral barriers have important consequences for legislative behavior and effectiveness inside the chamber (Lazarus and Steigerwalt Reference Lazarus and Steigerwalt2018). Anzia and Berry (Reference Anzia and Berry2011) argue that the gender disparity in candidate qualification standards allows only the most capable and hard-working women to win their elections. As a result, they hypothesize that this over-qualified group of women should out-work the men they serve with once in office. The extant work supports this hypothesis, finding, for example, that minority party women are more successful in the legislative process than minority party men (Volden, Wiseman, and Wittmer Reference Volden, Wiseman and Wittmer2013). Further, it also appears that women have a higher incentive to demonstrate competency on a wider variety of issue areas than men: Atkinson and Windett (Reference Atkinson and Windett2019) demonstrate that women members must introduce twice as much legislation in order to face the same number of challengers as men. Heightened levels of electoral threat in combination with gendered differences in voter qualification standards appear to motivate women representatives to have a high legislative output.
Beyond legislation, the pressures to outperform also seem to influence women’s district-oriented behaviors. For example, it has been shown that women legislators are more successful at securing federal funding for their districts (Anzia and Berry Reference Anzia and Berry2011; Lazarus and Steigerwalt Reference Lazarus and Steigerwalt2018) and that women city council members spend more time serving their districts (Thomas Reference Thomas1992). State legislative scholars have shown a similar relationship when looking at casework performance (Richardson and Freeman Reference Richardson and Freeman1995; Thomsen and Sanders Reference Thomsen and Sanders2020). Overall, the gendered “expectations gap” appears to put additional pressure on women to go above and beyond both at the candidate emergence stage and after they enter political office.
Women and politics scholars have pointed to a variety of factors to explain what drives these gendered expectations. External pressures, such as voter biases and lack of support from party elites, are one potential source of the unequal expectations placed on women politicians. Empirical research shows that voters hold women candidates to a higher standard when evaluating their credentials and competency. For example, using a series of survey experiments, Bauer (Reference Bauer2020) finds that voters use more stringent criteria to assess the political readiness of women. Other work demonstrates that citizens look at more competence-related information about women running for office than they do for men (Ditonto, Hamilton, and Redlawsk Reference Ditonto, Hamilton and Redlawsk2014), that constituents ask women to perform more casework (Richardson and Freeman Reference Richardson and Freeman1995), and that women are expected to cover more issue areas (Butler, Naurin, and Öhberg Reference Butler, Naurin and Öhberg2022). Therefore, it is possible that anticipation of gender bias from voters has led women to think that they must work harder to credibly demonstrate competence and satisfy constituents’ needs.
Women may also be perceived, both by challengers and party leaders, as being more electorally vulnerable than men. Numerous studies have found that when women run for office they have a significantly greater probability of facing a challenger both during the primary process (Lawless and Pearson Reference Lawless and Pearson2008) and in the general election (Fulton Reference Fulton2012). Their political opponents are also more likely to be higher quality (Milyo and Schosberg Reference Milyo and Schosberg2000). The increased likelihood of facing a credible challenger could certainly be an additional factor causing the gender disparity in candidate experience, insofar as women may think they have to be more qualified in order to fend off quality competitors. Women may also need better credentials than men in order to be recruited as candidates. Women are not as likely to be asked to run by local party networks and recruiters (Carroll Reference Carroll1994; Fox and Lawless Reference Fox and Lawless2010) and party leaders tend to be more uncertain about women’s ability to achieve electoral success (Niven Reference Niven1998; Sanbonmatsu Reference Sanbonmatsu2006). As such, women candidates and legislators appear to face additional pressures to excel from party elites.
Internal factors, such as women candidates’ perceptions of their own abilities, may also be driving the gender gap in candidate caliber and legislative effectiveness. In a survey comparing women and men in “pipeline” professions for politics (i.e., business, law), Fox and Lawless (Reference Fox and Lawless2005) demonstrate that there is a gender difference in political ambition that, in part, comes from gaps in self-perceived readiness for politics. They find that women are significantly less likely to think that they are qualified to campaign for public office than are men with similar professional and personal backgrounds. It is then unsurprising that women tend to be more hesitant to take the leap and run for office and are more concerned about their electoral viability than men (Lawless and Fox Reference Lawless and Fox2010). Women may also doubt that their party will support them financially during their campaign (Fowler and McClure Reference Fowler and McClure1989, 114). As a result, only the most ambitious, experienced, and well-positioned women end up emerging as candidates and winning political races.
Inside the Legislature
Much scholarship from American politics emphasizes the electoral barriers that aspiring women legislators face. Yet a variety of other work also suggests that these barriers persist inside the walls of the legislative chamber. Existing qualitative research, for example, supports the expectation that legislative insiders will hold biases similar to those of voters and other party elites. Political scientists’ current knowledge of the intra-institutional pressures felt by women members comes primarily from in-depth interviews with women officeholders. These studies have found that women representatives regularly express feeling that their talents are undervalued and underestimated by the men in the legislature. Early interviews with members serving in state legislatures indicate that women are more likely to report difficulties in developing collegial bonds, particularly in settings with low levels of women’s representation (Blair and Stanley Reference Blair and Stanley1991). Women representatives frequently cite feeling hyper-aware of their surroundings and are more cognizant of the impression that their actions could be sending (Lazarus and Steigerwalt Reference Lazarus and Steigerwalt2018).
Women legislators have also reported feeling that they must work harder to prove that they belong in men-dominated spaces (Lazarus and Steigerwalt Reference Lazarus and Steigerwalt2018; Dittmar, Sanbonmatsu, and Carroll Reference Dittmar, Sanbonmatsu and Carroll2018). In one interview, former Senator Barbara Mikulski (D-MD) explains that the “whole architecture” of the Senate was designed with men in mind (Dittmar, Sanbonmatsu, and Carroll Reference Dittmar, Sanbonmatsu and Carroll2018, 69). As a result, women who make it into public office feel they must actively work to combat stereotypes that they are not suited for political life. Swers (Reference Swers2013) finds that staffers for women members felt that Democratic women senators had to do more than similarly situated men and believed that they were taken less seriously by Pentagon officials. Notably, in her analysis of state legislators’ self-reported strategies for attaining power, Reingold (Reference Reingold1996, 475) finds that “Arizona female officials (67%) were more likely than their male colleagues (44%) to mention the value of hard work and knowledge,” hinting “at the possibility that fewer Arizona women than men felt they had the latitude or ability to be successful without working extremely hard.” Analyses of legislative discussion dynamics also suggest that women have trouble being heard; numerous studies reveal patterns of women being more likely to be interrupted during policymaking (Kathlene Reference Kathlene1994; Vallejo Vera and Gómez Vidal Reference Sebastián and Vidal2022; Miller and Sutherland Reference Miller and Sutherland2023).
Theories of gendered institutions seek to explain these findings, arguing that men tend to be advantaged within the institutions that they created (Duerst-Lahti and Kelly Reference Duerst-Lahti and Kelly1995, 5). Specifically, the rules and norms of government bodies, such as committee assignments, rules of floor debates, and leadership structures, can be used to systematically “other” women in legislatures and reinforce gendered hierarchies (Hawkesworth Reference Hawkesworth2003, 531). For example, Michelle Heath, Schwindt-Bayer, and Taylor-Robinson (Reference Michelle Heath, Schwindt-Bayer and Taylor-Robinson2005) demonstrate that men legislators marginalize their women colleagues by disproportionately assigning them to committees that focus on women’s issues and social issues, and Franceschet and Piscopo (Reference Franceschet and Piscopo2008) show that women face more difficulty passing their legislation. Hawkesworth (Reference Hawkesworth2003) uses interview data to demonstrate that Congresswomen of color consistently report feelings of marginalization and are “systematically shut out of key decision-making arenas” (547), indicating that “institutional norms and practices may be raced and gendered, or that political institutions may play a critical role in producing, maintaining, and reproducing raced and gendered experiences within and through their organizational routines and practices” (530). As explained by Goodwin, Bates, and McKay (Reference Goodwin, Bates and McKay2021, 634), the consequence of such practices is gendered power imbalances in the legislative process that prevent women from influencing outcomes in the same way as men. Alternatively, Kerevel and Atkeson (Reference Kerevel and Atkeson2013) do not find significant evidence of women’s marginalization in the Mexican Chamber of Deputies, suggesting that previous findings of discrimination in legislatures may be tied to institutional structures that grant privileges on the basis of seniority.
Elite Evaluations and Theoretical Expectations
A robust literature on women and politics suggests that both voters and party elites set a high bar for women candidates and elected officials, and that many women doubt their own ability to run a successful political campaign. Other valuable research indicates that women report feeling “othered” in legislative settings by their men colleagues. Each of these factors are potential sources of women’s continuing underrepresentation in U.S. political institutions. However, it is our understanding that no research has quantitatively tested whether or not higher standards are placed on women by the legislative insiders working alongside them.Footnote 5 Our main hypothesis builds on these rich previous literatures.
We develop our hypothesis with respect to the distinctive data source that we use to test it: evaluations of legislators’ effectiveness by other legislative insiders. Some previous work in political science has used such elite surveys to get a quantitative measure of a legislator’s influence in the chamber. Using a nation-wide survey of Black state legislators in 1991, Hedge, Button, and Spear (Reference Hedge, Button and Spear1996) demonstrate that legislators with greater seniority report larger personal influence over party matters and are more likely to attain leadership positions. However, the survey also reveals that Black “women are more likely to report having experienced or observed discrimination in the legislature” (92). Meyer (Reference Meyer1980) developed and tested a theory of legislative influence using a measure based on North Carolina state legislators’ answers to the question, “Who are the five most influential members of the house?” Other researchers subsequently expanded on this work using a survey administered starting in 1979 by the North Carolina Center for Public Policy Research (NCCPPR) that asks legislators, journalists, and lobbyists to rank the effectiveness of each member of the House and Senate (Padró i Miquel and Snyder Reference Gerard and Snyder2006; Weissert Reference Weissert1991). In the research most similar to our own, Haynie (Reference Haynie2002) leverages these rankings to show that African American legislators are perceived to be less effective by their colleagues. In this article, we use this underutilized data set of legislative effectiveness rankings to assess whether or not the work of women representatives is undervalued by their fellow legislators and other legislative insiders.
Why are such elite evaluations important? Put simply, “without the respect of colleagues a legislator has difficulty influencing public policy or advancing in status and power within the legislature” (Caldeira, Clark, and Patterson Reference Caldeira, Clark and Patterson1993, 5). Legislatures are social institutions in which perceptions of effectiveness, work ethic, and competence are critical currency. If legislators are strategic seekers of reelection, they should attempt to form collaborative relationships with other legislators and lobbyists who can best help them achieve their goals (Kirkland Reference Kirkland2011). Similarly, lobbyists likely want to work with, and therefore provide more resources to, the legislators who are most likely to successfully advance their clients’ agendas. Consequently, if women’s accomplishments are systematically discounted in comparison to men’s, legislative insiders will likely be less interested in forming coalitions with or elevating women to positions of leadership based on that work.
Combining insights from previous scholarship on gender and legislative politics with the context of elite evaluations leads us to our main hypothesis:
Hypothesis Women legislators’ accomplishments will translate into smaller gains in elite evaluations than comparable accomplishments by their men colleagues.
We acknowledge, however, that there are also theoretical reasons to expect that this hypothesis may not bear out in data. There is reason to believe that legislative insiders, who have direct experience working with women to develop policy and serve constituents, will evaluate women’s work accurately. Because they interact with their women colleagues on a day-to-day basis, insiders are well positioned to assess women’s contributions. If this is the case, legislators’ evaluations of the effectiveness of their women colleagues should meet or surpass what objective measures call for. It is also possible that the “Jackie Robinson Effect” (Anzia and Berry Reference Anzia and Berry2011) means that the women who serve in the chamber are generally more capable than men, possibly offsetting or masking bias within the chamber (Knox, Lowe, and Mummolo Reference Knox, Lowe and Mummolo2020). Finally, we emphasize that our study cannot establish whether women receive equal credit for equal work; rather, we can only show whether women receive equal credit for equal accomplishments.
Data on Legislative Influence
To measure subjective perceptions of legislators’ influence, we use an underutilized data set of elite legislative effectiveness rankings collected by the North Carolina Center for Public Policy Research (NCCPPR) from 1993 to 2015. The NCCPPR’s website provides the following description of their survey procedures:
The Center’s effectiveness rankings are based on surveys completed by the legislators themselves, by registered lobbyists who are based in North Carolina and who regularly work in the General Assembly, and by capital news reporters. These three groups are asked to rate each legislator’s effectiveness on the basis of participation in committee work, skill at guiding bills through committees and in floor debates, and general knowledge or expertise in specific fields. The survey respondents are also asked to consider the respect that legislators command from their peers, as well as his or her ethics, the political power they hold (by virtue of office, longevity, or personal skills), their ability to sway the opinions of fellow legislators, and their aptitude for the overall legislative process.Footnote 6
Each survey respondent was asked to rank the members of the North Carolina General Assembly on a scale from one to ten, though legislators were only asked to evaluate members of their own chamber. After the survey is completed, members in the House and Senate are ranked from least to most effective in their chamber. As a result, in the original data a lower score indicates that a member is more effective. However, for the sake of interpretability, we flip and re-scale the measure so that it ranges from 0 to 1, with 1 corresponding to the most effective, and 0 the least effective, member in a given chamber-session. These NCCPPR legislative rankings provide us with rare insight into how those most involved and invested in legislative politics, including legislators themselves, perceive legislators’ effectiveness.Footnote 7 While ideally we would be able to examine the raw survey data in order to examine relationships conditional on the profession (as in Haynie Reference Haynie2002) or gender of the respondents, we have been unable to access such data.Footnote 8 Fortunately, Padró i Miquel and Snyder (Reference Gerard and Snyder2006) report that for earlier years when average evaluations were released by respondent type, correlations across legislator, lobbyist, and reporter respondents were extremely high; the lowest such pairwise correlation was 0.89 (353).Footnote 9 Moreover, these evaluations correspond closely to legislators’ success at passing legislation in the chamber (Edwards Reference Edwards2018; Bucchianeri, Volden, and Wiseman Reference Bucchianeri, Volden and Wiseman2024), making them an ideal measure with which to evaluate how that relationship varies with legislator gender. While these facts assuage our concerns about pooling across respondent types, we nevertheless hope that future work will be able to parse our findings by respondent gender, in particular.
With this measure of subjective influence in hand, we also need measures of legislative and electoral accomplishments and activities that should be associated with influence. We primarily draw on a variety of data collected by Bucchianeri, Volden, and Wiseman (Reference Bucchianeri, Volden and Wiseman2024) that captures legislators’ institutional positions and their productivity in the legislative process. Generally speaking, prestigious legislative roles such as leadership positions or committee chairs both reflect and confer institutional prestige. As such, we include measures for whether legislators in our sample served as legislative leaders, served as the chair of a committee, or served on the Rules Committee in a given session.Footnote 10 Over the period we study, 44% of members in a session held committee chairmanships, 6% were party leaders, and 23% were on the Rules Committee.
Another way that legislators can signal effectiveness and influence is through legislative productivity (Volden and Wiseman Reference Volden and Wiseman2014). This was also an explicit criterion given to NCCPPR survey-takers, and Edwards (Reference Edwards2018) shows that these evaluations are strongly associated with legislators’ “hit rates”—the proportion of introduced bills ultimately passed. As such, we consider legislators’ involvement in the legislative process using the comprehensive measure of state legislative effectiveness developed by Bucchianeri, Volden, and Wiseman (Reference Bucchianeri, Volden and Wiseman2024).Footnote 11 Over the course of the entire time period, the average member in a given session sponsored nineteen pieces of legislation and shepherded four of these into law. Like Volden and Wiseman’s (Reference Volden and Wiseman2014) congressional legislative effectiveness scores, these state legislative effectiveness scores (SLES) are normalized to take a mean value of one within each chamber-session. We also include members’ seniorityFootnote 12 and majority/minority status, which are known to impact legislators’ influence in the chamber (Volden and Wiseman Reference Volden and Wiseman2014).
Finally, we also consider members’ accomplishments outside the chamber. We include legislators’ previous general election vote share: higher vote shares may directly convey competence (at campaigning, for example), may reflect constituents’ appreciation of the legislator’s effectiveness (Padró i Miquel and Snyder Reference Gerard and Snyder2006), or may offer legislators leeway and political capital. In addition, members can demonstrate effectiveness through their fundraising efforts. It could be that legislators who contribute money to their party, a caucus committee, or other legislators are deemed more effective. In some analyses, we incorporate fundraising data collected by Kistner (Reference Kistner2022) that includes a variable for either the total amount or the proportion of such money contributed by each member from 2000 to 2012. The average member contributed about $39,000.Footnote 13 We combine these various measures with the NCCPPR legislative effectiveness rankings in order to determine whether or not objective influence translates into perceived influence in the chamber.
Table 1 shows the average value of each of the covariates by legislator gender.Footnote 14 Men tend to be slightly more represented on the powerful Rules Committee, with 24% of men serving on the Rules Committee in a given session compared to 19% of women. Women also have lower legislative effectiveness scores on average during the entire time period, and men in the legislature gave nearly twice as many contributions to peers as women legislators with about $45,000 given compared to $21,000. Women and men tended to have similar average levels of seniority, previous general election vote shares, and time spent as committee chairs and majority party members.
Mean of covariates by legislator gender
Before commencing our analyses, we first provide some context on the North Carolina General Assembly. While our study demonstrates a strength of using state legislatures to study general theories in the ability to draw on novel data sources, it is also potentially limited by our focus on a single—possibly distinctive—state legislature. In figure 1, we contextualize North Carolina relative to other state legislatures in four ways. Across the figures, we plot values for all (or nearly all) states during the period we study; we plot North Carolina in a darker shade.
The North Carolina General Assembly in Context
Note: The figure plots average values within state for four characteristics of state legislatures, with North Carolina emphasized in a darker shade. Legislature size is measured between 1992 and 2008 and is sourced from Dubin (Reference Dubin2007, as updated by James Snyder). Legislative professionalism averages Squire Index (Squire Reference Squire2017) values from 1996, 2003, 2009, and 2015. Average margin of victory is based on election data from Klarner (Reference Klarner2018), and limits data to single-winner general elections; it includes uncontested elections. The proportion of women legislators is based on the full period 1992 to 2014 and is based on data from Bucchianeri, Volden, and Wiseman (Reference Bucchianeri, Volden and Wiseman2024).
First, in figure 1a, we plot the size of all bicameral state legislatures, combining the two chambers. In figure 1b we plot each state’s legislative professionalism, using Squire’s (Reference Squire2017) familiar index. In figure 1c, we plot the average margin of victory in single-winner general elections. Finally, in figure 1d, we plot the proportion of state legislators who are women. Broadly speaking, Figure 1 suggests that the North Carolina General Assembly is exceedingly average. It is moderately sized, moderately professionalized, and its legislators experience a typical amount of electoral competition. Despite being a southern state, North Carolina is generally more professional and more competitive than most southern states, assuaging concerns that our results will disproportionately reflect the South’s uniqueness. While the legislature is disproportionately Democratic in the earliest years we consider and disproportionately Republican in the latest, the minority party never holds less than 30% of seats during our period of study. Perhaps most important for our purposes, North Carolina has a relatively typical proportion of women legislators. Nevertheless, recent history in the state suggests that its politics may be biased against women: in 2008 Bev Perdue became the state’s first elected woman governor—“no mean feat in a state with a legendary old boys’ political network”—but faced declining popularity throughout her term and ultimately chose not to pursue reelection.Footnote 15 This notwithstanding, in the aggregate North Carolina seems to be a reasonably average state in which to test our expectations.
Men’s and Women’s Legislative Careers
We begin our empirical exploration by descriptively examining how potential gender discrimination manifests at various stages of legislators’ careers in the North Carolina legislature. While our subsequent analysis attempts to address potential discrimination by controlling for legislator-specific attributes and examining whether subjective evaluations respond to legislator’s accomplishments differently for men and women, here we begin by simply establishing whether women and men progress through their legislative careers similarly or differently. To do so, we examine how the arc of perceived influence varies as legislators gain seniority. Existing scholarship emphasizes the importance of seniority for establishing the institutional expertise and gaining the institutional positions that afford legislators influence (Volden and Wiseman Reference Volden and Wiseman2014). We also explore whether women and men gain access to important institutional positions at similar points, on average, in their legislative careers.
Evolution of Influence over Careers
What does a legislator’s effectiveness trajectory look like? As an example, we can track the career of Deborah K. Ross. Ross was a member of the North Carolina House of Representatives from 2003–2013. During her first term in the chamber she was ranked as the forty-first most effective member of the chamber. This is impressive considering she was in the minority for her first term. In the following term, she became a member of the majority and received a committee chairmanship. Her legislative effectiveness score rose accordingly, and in that and the next two terms she was ranked as a top-fifteen most-effective legislator by her fellow legislative insiders. Subsequently she fell out of the majority and lost her status as a committee chair in her last two terms, and insiders’ perception of her effectiveness dropped accordingly, putting her as the fiftieth and fifty-sixth most effective legislator in those two terms.
We next consider more generally whether the evolution of men’s and women’s influence in the North Carolina General Assembly progresses similarly over the course of their careers. Seniority has historically been associated with increased influence in legislative politics (Padró i Miquel and Snyder Reference Gerard and Snyder2006). In part this is due to enhanced opportunities to hold valuable institutional positions, such as prestigious committee assignments, committee chairs, or leadership positions, but others have found that legislators are more effective later in their careers even controlling for such institutional positions (Volden and Wiseman Reference Volden and Wiseman2014). To explore whether men and women have similar career arcs, we simply plot the mean ranking by gender by seniority year, separately by chamber. For purposes of this analysis, this relationship that we document is explicitly unconditional—the patterns we document here may be due to differential access to valuable legislative positions, differential credit for similar accomplishments, or any other number of possible mechanisms. We explore the mechanisms in more detail later, but emphasize that the descriptive, unconditional pattern here is normatively important for descriptive representation (Mansbridge Reference Mansbridge1999).
We document legislative influence over careers in figure 2. The patterns we identify are striking. First, we find little difference in effectiveness between first-term men and women, indicated by the leftmost points for each chamber. We view this as particularly important because first-term legislators have little access to institutional perquisites—such as leadership positions or committee assignments—that may skew perceptions of their influence. We show in appendix table B.1 that this parity in evaluations of first-term men and women also holds after controlling for a variety of first term-relevant covariates and year-specific effects.
Mean legislator influence by gender, chamber, and seniority
Note: The figure plots the mean ranking of legislators by seniority, separately by chamber. Seniority is chamber specific. Plotted numbers indicate the number of legislators included in the plotted mean. Seniority years over ten are omitted due to the small number of legislators serving that long.
This similarity between men and women persists over approximately the first six terms that a legislator serves. In the House, legislators of both genders enjoy a bump in perceived influence in their first few terms, particularly their second, and then appear to plateau after the third or fourth term served. In the Senate, legislators of both genders are evaluated consistently more positively through their fifth or sixth term. While we find that men remain similarly or more-positively evaluated beyond the sixth term in both chambers, women appear to fall off. This pattern, however, appears to be attributable to a combination of low numbers of women and minority party status; the single woman remaining after eight terms in the Senate was in the minority, and the six women who make it to their tenth term in the House are all in the minority, as well. Majority party status is worth about a thirty-ranking bump in the House and a fifteen-ranking bump in the Senate, on average, which suggests that this pattern likely represents the idiosyncrasies of the sample rather than a meaningful pattern.Footnote 16
Ascent to Institutional Positions
Second, we investigate whether and when legislators attain key institutional positions, focusing on the role of subjective influence in securing these positions. This analysis helps us to parse the patterns documented in figure 2 by considering whether legislators of different genders have different access to valuable institutional positions, and whether they must show greater subjective influence before being handed the keys to these positions. We summarize this information in table 2. In this table, we catalog the seniority and perceived influence of legislators in the year before they become chairs or leaders. Intuitively, if it takes women more terms of service, or they must have higher perceived effectiveness in order to be elevated to a valuable institutional position like a committee chair or leadership position, this could reflect biases in the legislature more broadly. As with our other analyses, we split our discussion by chamber.
Ascent to chair and leadership positions
Note: “Terms” refers to the number of terms served by a legislator before they first achieve the given role; “Ranking” refers to the legislator’s NCCPPR ranking in the session before they achieve the given role, with higher values corresponding to higher effectiveness.
In the top two rows, we explore committee chairs. In general, men and women appear to take a similar amount of time to become chairs (about three terms in the House and about 2.5 terms in the Senate). In both chambers, women are perceived as somewhat more effective than men before first becoming chairs, but the differences are not especially large. In the bottom two rows, we do the same exploration for leadership positions. On average, it takes legislators more terms of service and greater perceived effectiveness before they become leaders than chairs. In general, however, we find broad parity across men and women. The one exception to this is women in the House, who are considerably less effective on average before becoming leaders than men. While we caution against over-reacting to these descriptive values—especially given the potential effect of majority status on both effectiveness and the potential to assume these rolesFootnote 17—this nevertheless provides some prima facie evidence that women and men experience similar career trajectories in the North Carolina General Assembly.
Overall, we find relatively limited evidence of gender discrimination in the aggregate patterns of legislative careers in the North Carolina General Assembly. Men and women legislators start their careers on a relatively even footing, their careers evolve similarly, and they generally, but not always, appear to ascend to important institutional positions at comparable rates and speeds. Late-career women and leadership positions appear to be two important exceptions where we find potential evidence of discrimination. While these patterns are descriptively important, they may mask differences in women’s legislative effort—that is, women may be keeping up with men in their perceived influence, but may be working harder to earn those perceptions. To further explore whether reality matches perception in the same way for men and women legislators, we next explore whether men and women legislators receive similar subjective “credit” for their objective accomplishments.
Women’s and Men’s Accomplishments Are Perceived Similarly
We next undertake our main regression analyses. We deviate from many analyses in previous literature on discrimination in our approach; rather than exploring whether gender remains a relevant consideration after controlling for legislator characteristics and actions,Footnote 18 we ask whether within-legislator changes in those characteristics and actions correspond to similar changes in evaluations for men and women legislators. While this approach does not completely allow us to circumvent the issues generated by potentially biased selection procedures (Knox, Lowe, and Mummolo Reference Knox, Lowe and Mummolo2020), it does allow us to hold fixed legislator-specific characteristics—such as ambition, talent, or gender itself—that may have affected a legislator’s presence in the chamber to begin with. This design also holds constant legislator race, which as previously noted, extant work has found to impact perceived influence in legislatures (Haynie Reference Haynie2002).
To investigate whether objective influence translates into subjective influence similarly for men and women legislators, we use a model interacting gender with our measures of party leadership status, committee chairmanship, Rules Committee membership, seniority, majority party status, legislative effectiveness, and previous general election vote share.Footnote 19 We include legislator fixed effects, which control for time-invariant factors specific to each legislator, and year fixed effects, which account for year-specific trends. Standard errors are clustered on legislator. Recall that our outcome variable is oriented such that higher scores represent greater perceived effectiveness, with scores ranging from zero to one.
The results from this analysis are presented in figure 3. We present the marginal effect of each covariate for men and for women and the p-values of the interaction estimate; a more-positive coefficient estimate for women, relative to men, indicates that women get “more credit” for that particular covariate, while a less-positive coefficient estimate for women than men suggests that women get less credit for that achievement. The top panel of figure 3 displays findings for the House and the bottom panel displays findings for the Senate. We present full regression estimates for the models that these estimates are based on in table F.1 in the appendix. The table includes point estimates and standard errors for the interaction between legislator gender and these characteristics, the p-value for which is included in figure 3.
Perceptions of legislative influence, within-legislator models
Note: “Men” and “Women” estimates are based on a single model for each chamber that interacts the included covariates with legislator gender; the marginal effects by gender are presented. The model includes legislator and election year fixed effects. The House model has 1,363 observations; the Senate model has 562 observations; 95% confidence intervals are based on legislator-clustered standard errors. Continuous variables are standardized for ease of comparison.
We note first that the basic pattern of our results has a great deal of face validity. We find that becoming a leader or committee chair, serving in the majority party, or having greater legislative effectiveness are consistently associated with higher perceived effectiveness. Less clear but still potentially positively associated are serving in the Rules Committee, gaining a higher vote share, and having more seniority. Overall, our independent variables have the strong positive association with perceived effectiveness that one would expect theoretically.
Are there any findings that support the claim that women are perceived as less effective than men? We find two instances in which women appear to be given less credit for work. Women in the House are ranked as significantly less effective than men when they accumulate additional years of seniority. In the Senate, women on the Rules Committee are ranked as less effective than their men peers.
What findings push back on the idea that women are discriminated against by legislative insiders? Overall, there appear to be very few differences between the rankings of men and women legislators; most of the marginal effects for gender are not significantly different from each other. Leadership status, committee chairmanship, and effectiveness all appear to work similarly for men and women across the two chambers. Focusing on the House, women are ranked as significantly more effective when they receive comparatively higher vote shares than men. The interaction on Woman x Vote Share, which captures the difference in the marginal effects between men and women, takes on a value of 0.031, indicating that women gain about four additional ranking positions (in the 120 person chamber) when Vote Share increases by one standard deviation than men do.
Taken together, we believe these analyses largely suggest that women are ranked similarly to men by legislative insiders. While certain coefficients suggest differences, they point in opposing directions and do not indicate a clear pattern at work.
Importantly, the variables we consider vary in the degree of discretion or effort that underlie them. While we cannot rule out that women must work harder for high-quality committee assignments or leadership positions, for example, the fact that women are rewarded similarly to men for being in the majority—something relatively outside of their control—is heartening. Similarly, while we cannot rule out that women are producing higher-quality bills, the fact that the marginal effect of legislative effectiveness is similar for men and women suggests that a simple version of effort—sponsoring and passing bills—is similarly rewarded. Finally, we note that even if we cannot rule out that women have to work harder to achieve the same objective accomplishments, those accomplishments seem not to be devalued by insiders—rather, women and men appear to receive generally similar “bumps” in effectiveness perceptions as a function of objective accomplishments. In short, while we cannot discount the possibility that women must work harder to get into leadership or onto the Rules Committee, we can rule out that they have to achieve more such feats in order to get the same bump in evaluations as men.
Additional Analyses and Robustness Checks
It may be the case that these results are masking heterogeneity by party. While we are unable to know the traits of individual respondents, the gender breakdown of the parties suggests that Republican and Democratic women may face different standards for advancing to important positions within the chamber. To explore this possibility, we break down the analyses by party and present the results in figures G.1 and G.2 in the appendix. Generally, the results appear to be similar by party. For example, across both parties, women in the Senate seem to get less credit for Rules Committee membership. There are a few differences across the parties, however. First, the results reveal that the negative impact of seniority appears to be largely driven by Democrats. Women Republicans in the Senate also appear to receive less credit for legislative effectiveness. Turning to areas in which women receive more credit, women Republicans in the House and Senate appear to receive more credit than men for previous general election vote share, and women Republicans in the House receive more credit for committee chairmanships.
There may also be heterogeneity by time period. Gendered expectations of women have evolved over time, and biases may have abated in the more modern period. To test this idea, we present results broken down by decade in figures G.3 and G.4. Results for election years 1992 to 2002 are presented in the top panels, and results for election years 2004 to 2014 are presented in the bottom panels of these figures. When parsed this way, the results still largely suggest that there are not significant differences between the effectiveness evaluations of men and women legislators for either period. For the period from 1992–2002, there are no significant differences for women in the House and women in the Senate appear to be given more credit for leadership positions and achieving additional years of seniority. For the period from 2004–2014, women in the House appear to be given more credit when they are a member of the majority party and receive a higher previous vote share, and less when they accumulate additional years of seniority. Women in the Senate appear to be given less credit for Rules Committee membership. Overall, it does not appear to be the case that gendered expectations have changed substantially over time in the North Carolina General Assembly.
Importantly, we note that our main analyses do not account for the intersectional nature of race and gender. Existing work demonstrates the unique challenges that women of color face in state legislatures (Brown Reference Brown2014; Hawkesworth Reference Hawkesworth2003). While the limited number of Black women in our dataset (85) precludes us from including an additional interaction for race in our regressions, we can re-run our analyses comparing white women to (all) men legislators and comparing Black women to (all) men legislators. The results from these analyses are presented in figures G.5 and G.6. When subset in this way, the marginal effects remain largely statistically indistinguishable. One notable difference, however, is that Black women in the Senate appear to receive more credit than men for accumulating Rules Committee service (p = 0.15), while white women appear to receive less credit than men for the same. Alternatively, white women in the Senate appear to receive relatively more credit than Black women for receiving higher vote shares, relative to men. We note two important limitations of these analyses: 1) race is likely to be substantially correlated with partisanship, and 2) our sample, particularly for Black women, is quite small. These results should be interpreted with these limitations in mind.
Finally, we run two additional analyses probing the strength of our results. First, we replace our within-legislator model with a lagged dependent variable specification. The results, presented in figure H.1, are substantively similar to those presented in the text. While we do find some large positive and negative coefficients in the Senate, we expect that this is due to overfitting as a result of our substantially reduced number of observations. Second, we re-run our main regressions while also incorporating campaign contributions to the party from Kistner (Reference Kistner2022), as described earlier, as an additional independent variable; this results in a substantially reduced sample size due to the shorter time frame for which we have campaign contribution data and missing data. The results are presented in figures H.2 and H.3. There do not appear to be any significant differences between the effectiveness rankings of men and women based on total contributions to the party; however, women in the Senate appear to receive a larger boost in ratings in comparison to men when they contribute a greater proportion of their total contributions.
Taken together, these findings provide suggestive evidence that objective influence translates into subjective influence similarly for women and men. Though there are some results that indicate there are differences between the evaluations of men and women legislators, these results trend in conflicting directions and are not enough to demonstrate a consistent pattern.
Conclusion
Extant work indicates that women legislators face a unique experience in office, going above and beyond the men they serve with to work against stereotypes and prove their competence to voters (Swers Reference Swers2013). This feeling of “gendered vulnerability” has important consequences for legislator behavior, leading women to sponsor more legislation and allocate more resources towards the district (Lazarus and Steigerwalt Reference Lazarus and Steigerwalt2018). Although much existing work in American politics has examined the presence and consequences of gender biases in the electorate (Bauer Reference Bauer2020; Lawless and Pearson Reference Lawless and Pearson2008) and among party elites in the recruitment process (Niven Reference Niven1998; Sanbonmatsu Reference Sanbonmatsu2006), surprisingly little is known about the biases women legislators face among their colleagues inside the chamber. This question has important implications for women’s success in government, as legislatures are ultimately social institutions; making public policy and gaining power inside the legislature are dependent on one’s ability to work with other legislators and receive credit for one’s work.
We take a first step at exploring gender biases among actual legislators by leveraging a unique dataset of legislative effectiveness rankings. Covering more than twenty years, we test how North Carolina elites rank their colleagues. In summary, we observe limited indication of gender bias within the North Carolina state legislature. Both men and women lawmakers experience similar career progressions (at least until the later stages), attain many key roles at comparable rates, and seem to receive comparable amounts of credit for objective measures of influence inside the chamber. Although some coefficients in our within-legislator analyses do reach traditional levels of significance, the models do not tell a cohesive story; coefficients frequently point in opposing directions and do not indicate a clear pattern.
The data do suggest, however, that ageism and sexism may form an important intersectional experience of bias for women. Women in the North Carolina House appear to be considered less effective when they accumulate additional years of seniority. While this may be an idiosyncratic result due to the low number of women reaching that stage of their careers in our sample, future studies should investigate how these traits impact power within the chamber and the political pipeline. If women’s lived experiences within the chamber are different than men’s at higher levels of seniority, this may impact their decision to stay in the political arena or seek higher office. The findings also indicate that exposing the public to women’s work ethic behind the scenes may be a potential avenue for alleviating biases about women’s performance in office. For example, sharing information with the public related to effort, rather than outcomes, may alter perceptions of competence and quality of representation.
While our analyses offer valuable insights, they certainly do not offer dispositive evidence that women do not face bias and discrimination in legislative politics. This study has important limitations that should be addressed in future work. First, while our study covers a relatively long time frame, it is notable that data stops right before the 2016 presidential election, which not only put issues of gender and sexism front and center but pushed more women to run in 2018 (Boatright and Sperling Reference Boatright and Sperling2019). It is possible that since this study’s time frame things may have gotten worse for evaluations of women; alternatively, it is possible that the rise of women in legislatures has continued to improve women’s experiences. Second, the legislative effectiveness measure we focus on is limited in scope. Although we do not find strong evidence of bias in our analyses, measures capturing perceptions of whether a legislator is a “team player,” “serves their constituents,” or is a “strong leader” may lead to results that more closely reflect the interviews given by women discussing their experiences in legislatures. Next, we focus on a single legislature, limiting our ability to generalize to settings beyond the North Carolina General Assembly. Exploring similar data for other legislatures (Haynie Reference Haynie2002, 299) would allow for both greater generalizability and an enhanced ability to explore legislators’ intersectional experiences, such as the intersection of gender with race and class. Further, in this article we are unable to separate out evaluations by gender. As explained by Wineinger (Reference Wineinger2022), Democratic and Republican women face different challenges in forming perceptions of competence, so the gender balance of the pool of evaluators is potentially consequential. Thus, future work could examine if women legislators rate women legislators differently than men. Scholars could also incorporate alternative measures of influence, such as media presence and outside connections.
Finally, and perhaps most importantly, our study does not account for the fact that, because of bias in the process that leads women to the legislature, comparing women and men legislators is not necessarily comparing like to like. As emphasized by Knox, Lowe, and Mummolo (Reference Knox, Lowe and Mummolo2020), bias affecting selection into a sample can produce situations where bias within that sample is underestimated. While the complex set of steps at which bias may stand in a woman’s way before reaching the legislature would make it difficult to accurately account for those biases when considering intra-legislature bias, such an endeavor would be valuable in the future to more completely capture the barriers that women face to making an impact in legislative politics.
Supplementary Material
To view supplementary material for this article, please visit http://doi.org/10.1017/S1537592725000477.
Data Availability Statement
Data replication sets are available in Harvard Dataverse at: https://doi.org/10.7910/DVN/IB9HR6
Acknowledgments
A previous version of this paper was presented at the 2024 Meeting of the Southern Political Science Association, the 2024 State Politics and Policy Conference, and the 2024 Center for Effective Lawmaking Conference. The authors thank Michael Kistner, Peter Bucchianeri, Craig Volden, and Alan Wiseman for sharing data, and Diana O’Brien, Leslie Schwindt-Bayer, Craig Volden, and audience members at the Rooney Center Seminar Series at Notre Dame University for valuable feedback. Kaslovsky thanks the Rice University Department of Political Science for research support.
Open access
