POLITICAL PARTIES, IDEOLOGY, AND WOMEN'S REPRESENTATION

Political parties are the central political actors in established democracies, determining both who comes into elected office and which policies reach the legislative agenda. Parties control the recruitment of candidates, for example, dictating how many (and which) women stand for office. Once elected, parliamentarians influence policy primarily by acting within the party group to create support for their cause (Mattson Reference Mattson and Döring1995). Parties, in turn, work to implement their electoral programs, and winning parties' policy platforms typically serve as the basis for the legislative agenda (Thomson et al. Reference Thomson, Royed, Naurin, Artés, Costello, Ennser-Jedenastik, Ferguson, Kostadinova, Moury and Pétry2017). Government agenda control and high levels of party cohesion together ensure that governing parties' bills are typically adopted, whereas legislation introduced by individual legislators mostly fails.

As parties determine both candidate selection and policy formation, any account of women's descriptive or substantive representation must consider their influence. Indeed, many studies examine women's representation within one or a small number of parties (Childs and Webb Reference Childs and Webb2012; Murray Reference Murray2010; Wiliarty Reference Wiliarty2010), and research on (female) legislators' behavior always accounts for partisanship. At the same time—and in contrast to the many studies of women's representation at the national level—little cross-national work examines party-level trends in women's descriptive representation (but see Caul Reference Caul2001; Kittilson Reference Kittilson2006; Kunovich and Paxton Reference Kunovich and Paxton2005), and virtually no research considers parties' efforts to substantively represent women over place and time.

Though few works assess women's representation within parties across states, existing scholarship highlights several factors that likely influence women's inclusion in parties' parliamentary delegations and on their policy platforms. Among these features, researchers return time and again to the role played by party ideology. In established democracies, the traditional left-right divide was structured by class, religious, and urban-rural cleavages (Lipset and Rokkan Reference Lipset and Rokkan1967). Beginning in the 1970s, a new set of divisions emerged with respect to postmaterialist issues such as environmental protection and acceptance of alternative lifestyles (Inglehart Reference Inglehart1997). While the meanings of “left” and “right” vary across time and place, leftist ideology typically advances the interests of working-class citizens, and it is often linked to secularism and a cosmopolitan world outlook. A right-leaning orientation is traditionally associated with upper-class voters. It is often more concerned with preserving traditions and frequently prioritizes nationalism and/or religion.

Despite political developments, the notions of left and right remain the “major organizing principles” of—and the “foundation for contemporary patterns of political competition” in—established democracies (Mair and Smith Reference Mair and Smith1990, 175). With respect to women's political representation, many studies emphasize the left-right distinction. Left parties have long traditions of publicly embracing gender-egalitarian ideologies and historical ties with progressive women's movements (Beckwith Reference Beckwith2000). These parties have also been more open to selecting female candidates and cabinet ministers (Caul Reference Caul2001). Progressive opinions on gender issues likewise often correspond to left party membership (Tremblay and Pelletier Reference Tremblay and Pelletier2000), and these organizations have been more likely to integrate feminist demands into their policy programs (Young Reference Young2000).

Just as left parties have been more receptive to women's representation, right parties in established democracies have traditionally been viewed as failing on this front. Preferences for traditional gender roles are strongly related to political conservatism (Cassese and Holman Reference Cassese and Holman2017). Unsurprisingly, in addition to nominating fewer female candidates, these parties have been more apt to advance nonfeminist and antifeminist claims (Lovenduski and Norris Reference Lovenduski and Norris1993; Wängnerud Reference Wängnerud2000). They are also less committed to feminist policy change and less supportive of the activities of women's policy agencies (Banaszak, Beckwith, and Rucht Reference Banaszak, Beckwith, Rucht, Banaszak, Beckwith and Rucht2003; Lovenduski Reference Lovenduski2005). Though right-leaning female voters may hold more moderate views than their male counterparts (Barnes and Cassese Reference Barnes and Cassese2017), they also express higher levels of sexism and racism—and show lower levels of support for fair pay policies—than women on the left (Cassese, Barnes and Branton Reference Cassese, Barnes and Branton2015).

In contrast to these established patterns, a growing body of work suggests that rightist ideology may no longer be as important a predictor of women's representation. Kittilson (Reference Kittilson2006), for example, does not find ideology to be a significant determinant of women's presence in the parliamentary caucus. Celis and Childs (Reference Celis and Childs2014) similarly note that right parties are now electing more women. These parties are also actively participating in gendered debates and reforms (Kantola and Saari Reference Kantola, Saari, Celis and Childs2014; Wiliarty Reference Wiliarty2010), often in an effort to modernize and broaden their appeal (Campbell and Childs Reference Campbell, Childs, Seldon and Finn2015; Childs and Webb Reference Childs and Webb2012; Kittilson Reference Kittilson2006). Indeed, a number of case studies point to right parties' policy claims on behalf of women (Bryson and Heppell Reference Bryson and Heppell2010; Curtin Reference Jennifer, Celis and Childs2014; Xydias Reference Xydias2013).

Clearly, left parties no longer maintain a “monopoly on the substantive representation of women” (Murray and Sénac Reference Murray, Sénac, Celis and Childs2014, 246). Yet whether (and to what extent) parties on the right remain behind with respect to women's descriptive and substantive representation remains unknown. Existing work, moreover, has not identified which kinds of right parties are making inroads and which are failing to do so. The following section considers the diversity among right party families in order to shed light on this variation.

NUANCING THE LEFT-RIGHT DISTINCTION

Beyond the left-right binary, there is heterogeneity among right (and left) parties. It is especially useful to consider “party families”—sets of parties that share a resemblance based on their common origins, transnational links, and similar ideologies and policy aims (Mair and Mudde Reference Mair and Mudde1998). Breaking apart the category of “right parties” to consider these distinct families reveals two important features for women's representation in established democracies. First, while one party group dominates the center-left (social democrats), the center-right is divided between two party families: Christian democrats and conservatives. Second, center-right parties are distinct from their far-right counterparts.

Christian democrats and conservatives are rarely found in the same system; countries with well-established conservative parties tend to have no/weak Christian democratic parties (and vice versa). Yet these two families are not perfect substitutes. Christian democrats are the dominant right party in Austria, Belgium, Germany, and the Netherlands, among other states. These parties emerged in response to secular, left-leaning social democrats, and they often retain traditional values on issues such as euthanasia, gay marriage, and abortion. Stemming in part from their historical religious links, these parties are especially interested in policies related to education and family. They are also more inclined to advance centrist programs and to defend (and promote) the welfare state (Gallagher, Laver, and Mair Reference Gallagher, Laver and Mair2011; Kubicek Reference Kubicek2012). Indeed, Christian Democrats oversaw the development and expansion of strong welfare states in both Austria and Germany.

Conservative parties, in contrast, lack formal links to organized religion. These parties also tend to be more staunchly antisocialist and less committed to the welfare state than their Christian democratic counterparts. Though class-based voting has dissipated, conservatives traditionally received most of their support from the upper and middle classes, as well as the rural population. Beyond the economic dimension, conservatives are the “guarantors of the existing order” (Ball Reference Ball2013, 25). They stress the importance of law and order and national defense and uphold traditional values. In this way, they combine economic liberalism with social interventionism (Gallagher, Laver, and Mair Reference Gallagher, Laver and Mair2011; Kubicek Reference Kubicek2012). Conservatives are the dominant right parties in states including Australia, Canada, Denmark, Ireland, Norway, Spain, Sweden, and the United Kingdom.

Christian democratic and conservative parties dominate right-leaning politics. In recent years, a number of countries have also observed the growth of far-right parties. These parties, which are characterized by the promotion of populist and xenophobic political appeals, have grown in response to fears over immigration, globalization, and the growth of the European Union (Mudde Reference Mudde2007). In some states, they remain protest parties. In others, including Austria, Denmark, and Sweden, they have gained traction, particularly among young, working-class, and male voters (Arzheimer and Carter Reference Arzheimer and Carter2006). Because these parties are still peripheral competitors in most states, the primary focus of this article is the two center-right party families. At the same time, in the results section, I offer preliminary insights about women's presence in far-right organizations.

There is, of course, variation within party families across place and time. There are significant differences, for example, between Anglo and Scandinavian conservatives' commitment to neoliberal welfare state reform (Arter Reference Arter and Broughton1999). In recent decades, Christian democrats have, to varying degrees, softened their stance on moral issues as their electorates have secularized (Van Hecke and Gerard Reference Van Hecke and Gerard2004). At the same time, party family has long been meaningful for women's representation. The distinction between social democratic and new left parties is widely acknowledged (Caul Reference Caul2001). Though they have received less attention in the gender and politics literature, right parties also differ in important ways.

Conservatives' and Christian democrats' distinct ideological commitments to liberalism, individualism, and the welfare state likely have ramifications for women's descriptive representation (including their openness to gender quotas) and substantive representation (such as their support for the social safety net on which many women rely). Indeed, since the advent of women's suffrage, these two party families have had distinct relationships with female voters. Morgan-Collins (Reference Morgan-Collins2016) shows that women's suffrage helped Christian democrats, as female voters were especially attracted to these parties' economic and social policies. Women have historically been important members of the Christian democratic coalition (Conway Reference Conway, Kselman and Buttigieg2003, 56–57). Conservative parties, in contrast, did not attract female supporters at the same rates. Yet whether—and to what extent—these differences carry into the modern era remains unknown. A comprehensive assessment of women's representation on the political right thus demands not only a broad reaching analysis that spans time and space but also one that accounts for the nuances among party families.

WOMEN'S DESCRIPTIVE REPRESENTATION

My study of right parties begins with an assessment of women's descriptive representation in parties' parliamentary delegations. I examine patterns in women's presence in office in 72 organizations in 12 democracies between 1980 and 2013.Footnote ¹ Following Kittilson (Reference Kittilson2011) and others, party family is identified using the Comparative Manifestos Project (CMP) (Volkens et al. Reference Volkens, Lacewell, Lehmann, Regel, Schultze and Werner2011). The CMP defines party family as a “tentative grouping of political parties and alliances” and classifies parties by “membership in international organizations, such as international party groups and factions in the European Parliament” (Volkens Reference Volkens2002, 158) (see the online appendix for more information). Like Adams et al. (Reference Adams, Clark, Ezrow and Glasgow2004), I identify Christian democrats, conservatives, and nationalists as right-wing parties.Footnote ² I compare these three party families with left-leaning and centrist organizations including green, socialist/communist, social democratic, liberal, and agrarian parties.

Model 1 is a linear model with a logistic transformation of the outcome variable that predicts the proportion of seats in the parliamentary caucus held by women. The results demonstrate that party family is a significant predictor of women's presence even after controlling for quota policies, female leaders, electoral system, party size and founding date, the passage of time, and country-level fixed effects (see the online appendix for more information on all models).

Table 1. Logistic regression models related to women's descriptive representation (with country-level fixed effects)

Notes: The outcome variable in Models 1, 3, and 5 is the logistic transformation of the percentage of seats in the parliamentary delegation held by women. The outcome variable in Model 2 is a dichotomous measure of the presence of a quota policy and in Model 4 of a female leader. For party families, the baseline category is Christian democratic organizations. AGR = agrarian parties; COM = socialist/communist parties; CON = conservative parties; ECO = ecological parties; LIB = liberal parties; NAT = nationalist parties; SOC = social democratic parties. Standard errors are presented in parentheses. See the online appendix for more information.

Conservative and Christian democratic parties have significantly fewer female members of parliament (MPs) on average than leftist parties (greens, communists, and social democrats). While there are examples of center-right parties with comparatively high levels of descriptive representation—such as New Zealand's National Party, the Dutch Christian Democrats, and the center-right Nordic parties—on average these organizations still perform more poorly than those on the left. It is especially interesting to compare the two center-right parties with their center-left counterpart. In 1985, women's average representation level was nearly identical across the three party types: women made up 13% of legislators in Christian democratic parties and 14% of conservative and social democratic representatives. Despite the increased attention to women's descriptive representation in recent decades, the gap between party groups has grown over time. In 2013, the mean percentage of seats held by women across all parliamentary parties was 28%. Christian democratic and conservative parties both fell below this average, with mean levels of women's representation at 23% and 24% of their parliamentary caucuses, respectively. In social democratic parties, on the other hand, the mean percentage of seats held by women was 37% (see the online appendix for plots).

Despite the claim that right parties are catching up to their left counterparts, the gap between the two is increasing as women's representation in center-left parties grows at a faster pace than in center-right organizations. Additionally, though there are only four nationalist parties in the sample, it is worth noting that these organizations perform especially poorly with respect to women's representation, lagging behind even center-right organizations. This is the case even though three of these parties are Scandinavian. Clearly, in established democracies, the variation in women's parliamentary representation is attributable both to the positive strides made by left parties and to the slower rate of growth among parties on the right.

WOMEN'S SUBSTANTIVE REPRESENTATION

Descriptive representation is only one facet of women's representation. Just as significant are parties' efforts to provide women with substantive—or policy—representation by including women on their policy agendas. As with women's presence in elected office, moreover, there are strong reasons to expect that both the quantity and content of women's policy representation vary across party families. Right party politicians, for example, have been shown to be both less feminist (Tremblay and Pelletier Reference Tremblay and Pelletier2000) and less concerned with gender equality (Wängnerud Reference Wängnerud2000). Nonleft parties are also less likely to support legislation “liberalizing divorce, extending abortion rights, criminalizing violence against women, expanding employment opportunities, providing women's healthcare innovations, and advancing social welfare issues” (Beckwith and Cowell-Meyers Reference Beckwith and Cowell-Meyers2007, 557). Right-party references to women are more likely to fall under what Celis and Childs (Reference Celis and Childs2014, 11) define as Type II claims: those that “address women's concerns and perspectives in ways distinct from traditionally understood feminism.”

Women's party-level policy representation can be studied using the policy statements parties draft prior to elections, which outline the legislative priorities they hope to enact once in office. These statements communicate parties' major policy ideas, allowing voters to compare competing programs. Importantly, while these platforms are campaign documents designed to sway the electorate, parties in advanced industrialized democracies work to implement their manifestos. Single-party executives have especially high rates of pledge fulfillment, though coalition and minority governments also perform surprisingly well on this front (Thomson et al. Reference Thomson, Royed, Naurin, Artés, Costello, Ennser-Jedenastik, Ferguson, Kostadinova, Moury and Pétry2017). Inclusion on the manifesto thus simultaneously indicates the party's commitment to women and also represents an important step in the policy-making process. Indeed, a growing number of case studies examine claims related to women on parties' election manifestos (Campbell and Lovenduski Reference Campbell and Lovenduski2005; Childs, Webb, and Marthaler Reference Childs, Webb and Marthaler2010; Murray and Sénac Reference Murray, Sénac, Celis and Childs2014; Xydias Reference Xydias2013), and quantitative cross-national work explores women's influence on manifesto content more generally (Greene and Lühiste Reference Greene and Lühiste2017; Greene and O'Brien Reference Greene and O'Brien2016; Kittilson Reference Kittilson2011).

To assess parties' efforts with respect to women's policy representation, I built a data set composed of the manifestos of 56 parties in 12 countries between 1980 and 2008.Footnote ⁴ Each manifesto was translated into English and searched for instances of attention to women. The text analysis was based on a dictionary containing almost 100 words signaling feminine issue framing (see online appendix). This dictionary was drafted from close readings of out-of-sample party platforms. It was then expanded based on the Beijing Declaration and Platform for Action, a comprehensive agenda for women's empowerment authored after the Fourth World Conference on Women in 1995. Using this approach, I was able to generate a count of the number of references to women on parties' platforms. In the following subsections, I focus on conservative and Christian democratic parties' efforts to represent women on their agendas. As nationalist parties rarely participate in government, they are excluded from these analyses.

The Content of Women's Substantive Representation

Just as parties' willingness to discuss women on their platforms differs across party families, the content of this representation may also vary based on party ideology. Right-leaning organizations are thought to be more likely to address women's traditional gender roles, particularly those related to mothering, as well as the complementarity of the sexes and the need for equality of opportunity and equal treatment (Bryson and Heppell Reference Bryson and Heppell2010; Childs and Webb Reference Childs and Webb2012; Murray and Sénac Reference Murray, Sénac, Celis and Childs2014; Piccio Reference Piccio, Celis and Childs2014; Wiliarty Reference Wiliarty2010; Xydias Reference Xydias2013). Right parties are also expected to make more nonfeminist or even antifeminist gendered claims (Curtin Reference Jennifer, Celis and Childs2014). In addition to this left-right divide, there is reason to expect important differences among right-leaning parties. As Verloo (Reference Verloo2015) points out, an emphasis on the “polity dimension” of conservatism may lead conservatives, for instance, to focus on equal employment opportunities for men and women. A stronger commitment to the “moral dimension” may make Christian democrats more likely to valorize women's role in the family and to focus on balancing work and familial obligations.

Despite these expectations concerning the content of women's policy representation, no study to date has systematically analyzed whether—and to what extent—different party families make different claims on behalf of women over place and time (but see Celis and Erzeel Reference Celis and Erzeel2015 for related work on legislators' behavior). Table 2 presents the results from the first cross-national study of the content of women's policy representation on parties' policy statements. Models 7–10 once again use quasi-Poisson regression analyses to predict the count of words for women on parties' manifestos. In each instance, however, I focus on subsets of dictionary terms that capture distinct beliefs about women's role in the private and public spheres.

Table 2. Quasi-poisson and binomial logistic regression models of women's substantive representation (with country-level fixed effects)

Notes: Models 6–10 and 12 are quasi-Poisson models. The outcome variable in Models 6 and 12 is the total count of words related to women on parties' election manifestos. The outcome variables in Models 7–10 are subsets of words related to different claims for women. Model 11 is a binomial logistic regression model estimated using Firth's penalized likelihood approach. The outcome variable is a binary variable that indicates whether the word was used at least once on the party's manifesto. For party families, the baseline category is Christian democratic organizations. AGR = agrarian parties; COM = socialist/communist parties; CON = conservative parties; ECO = ecological parties; LIB = liberal parties; NAT = nationalist parties; SOC = social democratic parties. Standard errors are presented in parentheses. See the online appendix for more information.

Model 7 includes words related to women's traditional roles as mothers and housewives (see online appendix for complete list of terms for Models 7–11). The results show that Christian democrats are significantly more likely to include these references on their manifestos than communist, social democratic, agrarian, and liberal parties. There is no distinction between Christian democratic and conservative parties on this front, however. Different patterns emerge if we consider women's role as mothers but focus on terms related to working mothers and women's capacity to balance family and paid employment. Model 8 looks at terms such as “child care,” “flextime,” and “parental leave.” Here, Christian democrats and social democrats perform similarly to one another and outpace other parties. Conservatives, on the other hand, fall behind both their center-left and center-right counterparts. Center-right parties in general, and conservative parties in particular, are thus especially likely to discuss women in ways that reflect conventional gender norms.

Clearly, Christian democrats are especially concerned with women as mothers both inside and outside of the home. Indeed, the results from Model 9 suggest that they are more likely than almost any other party family to discuss issues related to pregnancy on their platforms. This finding is consistent with work that points both to these parties' historical interest in maintaining women's traditional domestic roles (Piccio Reference Piccio, Celis and Childs2014), as well as their more recent successes in advancing policies aimed at balancing work and family life (Wiliarty Reference Wiliarty2010). Conservatives, in contrast, are willing to reference women's position in the private sphere, but they seem to shy away from issues such as day care and parental leave. This is not surprising, given that these policies demand large public expenditures. Campbell and Childs (Reference Campbell, Childs, Seldon and Finn2015) highlight the tension between the British Conservatives' child care commitments and their austerity politics. Kantola and Saari (Reference Kantola, Saari, Celis and Childs2014) likewise note that concerns over costs constrain conservative parties' efforts with respect to gender equality.

Just as there are policy areas in which right parties (particularly Christian democrats) outpace others, there are also arenas that remain out of bounds for these party families. Consider, for example, overtly feminist issues such as lesbianism, feminism, and sexism. The data suggest that center-right parties simply do not use these terms. Lesbians, for example, are never referenced by parties on the right. Feminism, moreover, is only explicitly addressed by green and communist organizations. Table 2 presents the results from a logistic regression analysis using Firth's penalized likelihood approach that models the presence of at least one of these terms on the manifesto (Model 11). Christian democrats and conservatives—as well as liberals and social democrats—are each significantly less likely to address these issues than communist and green parties.Footnote ⁵ Clearly, neither right parties nor center-left organizations are likely to address these more “radical” versions of women's representation on their agendas. Only far-left organizations aim to fundamentally transform the social systems that reinforce and promote gender-based discrimination.

WOMEN'S DESCRIPTIVE AND SUBSTANTIVE REPRESENTATION

Looking across place and time, both women's descriptive and substantive representation appear to be shaped by party family. Center-right parties elect fewer female parliamentarians than social democrats, and conservatives (though not Christian democrats) use fewer words related to women in their policy statements. Though the previous sections consider these forms of representation separately, a large body of literature suggests that the two are fundamentally linked. The final empirical analysis connects these two pieces of the puzzle and examines the relationship between numeric and policy representation across party families.

Parties' efforts to include women on their policy statements may be shaped by women's presence (or absence) from their parliamentary delegations. Increasing the percentage of seats held by female MPs both diversifies parties' platforms and pushes them leftward (Greene and O'Brien Reference Greene and O'Brien2016). This is partly because female parliamentarians are more likely to prioritize women's equality and family issues (Esaiasson Reference Esaiasson, Esaiasson and Heidar2000; Wängnerud Reference Wängnerud, Ballington and Karam2005). Likewise, in a number of institutions and across a variety of legislative activities, female legislators aim to represent women's interests (Barnes Reference Barnes2016; Tremblay Reference Tremblay1998; Xydias Reference Xydias2007). Unsurprisingly, the proportion of legislative seats held by women is often correlated with female-friendly policy outcomes (Bratton and Ray Reference Bratton and Ray2002; Kittilson Reference Kittilson2008).

While descriptive and substantive representation are often linked, women's presence alone does not guarantee attention to women on the policy agenda. To the contrary, this relationship is context dependent (Beckwith and Cowell-Meyers Reference Beckwith and Cowell-Meyers2007). In strong party systems like those studied here, it is especially difficult for female politicians to represent a distinct set of perspectives (Lovenduski Reference Lovenduski2005; Lovenduski and Norris Reference Lovenduski and Norris2003). Rather, the party leadership is responsible for policy decisions, and rank-and-file legislators follow the party line (Celis Reference Celis2008; Studlar and McAllister Reference Studlar and McAllister2002). The extent to which women's descriptive representation shapes policy representation thus likely varies both from party to party and among party types. That is, the ideological and organizational features of different party families create environments that make women's descriptive representation a stronger or weaker predictor of attention to women on the policy platform.

Extending Model 6, Model 12 reports the results of a second quasi-Poisson regression analysis of the count of words for women on parties' manifestos. This model, however, includes an interaction term between party family and the proportion of seats held by female MPs. This interaction effect shows how variation in women's presence affects the total references to women across different party types. The results are striking and differ from traditional expectations on this front. Controlling for a myriad of other factors, the proportion of seats held by women significantly affects the number of women-related terms on Christian democratic (as well as green and communist) parties' platforms. Parties with more female MPs use significantly more gynocentric words than those with fewer women. This effect size is largest, moreover, for Christian democratic parties: holding all other variables constant, moving from the first to the third quartile of female MPs (from 10% to 37%) increases the predicted count of the number of words for women by 12.

While women's descriptive and substantive representation are highly correlated in Christian democratic parties, this relationship does not hold for other major party families. Conservatives lag behind their counterparts, irrespective of women's presence in (or absence from) elected office. Among social democrats, women's policy representation appears to be well established at comparatively high levels and does not vary based on the proportion of seats held by female MPs.

The difference between Christian democratic and conservative parties is particularly striking. The link between descriptive and substantive representation within Christian democratic organizations may reflect the party family's unique ideological commitments and distinct history with female voters. It may also be a function of party structure. Christian democratic organizations are typically corporatist catchall parties. Consequently, “policies on women's issues are largely the product of internal interest group politics” (Wiliarty Reference Wiliarty2010, 47). In Germany, for example, the CDU's women's policies depend on the “success of the party's women's auxiliary organization in mobilizing and forming alliances” (47). Strong intraparty women's organizations may jointly promote women's descriptive and substantive representation, thus explaining the correlation between the two. A strong women's group may also empower female legislators to act on behalf of women.

In contrast to Christian democrats, conservative parties have more hierarchical internal organizations and lack voluntary quota policies. These two structural features likely depress the descriptive representation-substantive representation link. When power is concentrated among the party leadership, it is difficult for members of the parliamentary party to shape the policy agenda. In the absence of quota generated “mandate effects” (Franceschet and Piscopo Reference Franceschet and Piscopo2008), moreover, female legislators may not feel an obligation to act on behalf of women. Even if they would like to do so, conservative parties' neoliberal ideological views may make it harder to successfully advocate for female-friendly policies (particularly those that demand a strong welfare state).