Pushing for Gender Mainstreaming:Women's Presence and Alliances at the Local Level in Spain
A most important debate in gender representation studies revolves around whether women's presence in politics correlates with the advancement of gender-equal policies, given the mixed results observed. In this article, the authors aim to contribute both theoretically and empirically to this debate by investigating whether distinct types of women's presence/alliances have any differential impact on the promotion of gender mainstreaming at the Spanish local level. Our study shows a descriptive representation loop wherein the presence of women seems to attract more women into political positions. Importantly, the presence of women in local politics and the potential for them to form alliances between themselves has the ability to promote gender mainstreaming at the local level, albeit not all types of presence are equally effective. Our findings offer significant insights for reevaluating the connection between women's political involvement and gender equality.
local gender policies, local government, gender mainstreaming, women's presence and alliances, Spain
Introduction
An important debate in gender representation studies revolves around whether women's presence in politics correlates with the development of gender-equal policies. Some argue that the limited presence of women, presumed to be key advocates for gender equality, in political institutions hampers the adoption of feminist policies (Catalano Weeks 2022a). Others [End Page 120] contend that women's presence as representatives alone is not sufficient to overcome staunch resistance to implementing a gendered political agenda (Benschop and Verloo 2006; Freidenvall and Ramberg 2021; Mergaert and Lombardo 2014; Ylöstalo 2016), particularly in the face of radical right populist opposition to gender equality (Kantola and Lombardo 2020).
Existing evidence mirrors this debate; there is no systematic support for any of these arguments. Some studies find a positive relationship between the presence of women in power and the development of gender-related policies (Catalano Weeks 2022b; Kittilson 2008; Lippmann 2022; Schwindt-Bayer and Mishler 2005; Thomas 1991; Yamada 2024). However, other scholars conclude that there is no relationship between the presence of women in government or parliamentary positions and the existence of well-established gender policies (Bagues and Campa 2021; Ferreira and Gyourko 2011; Fokum, Fonjong, and Adams 2020; Geys and Sørensen 2019). There may be two primary reasons for this discrepancy in findings: variation in the issues or areas studied and different methodologies used to measure women's political presence. As for the first, the studies' focus varies across different (women-related) issues or areas (see, e.g., Mazur and Pollack 2009). Yet women's capacity to influence policy design might vary across different types of issues. Second, the presence of women in political positions is accounted for differently by these studies (number of women in parliament, number of women in government, presence of women in the overall organization, etc.), which might lead to different conclusions on the link between women's presence and women's policy responsiveness.
In this article, we attempt to contribute both theoretically and empirically to this debate by circumventing these potential biases. Instead of focusing on specific gender policy subsectors—family law, reproductive rights, violence, and so on (Mazur 2002; McBride and Mazur 2010)—we concentrate on gender mainstreaming. Gender mainstreaming permeates all policies, as "gender considerations are routinely included in the assessment of policy issues, options and impacts, along with other considerations such as socio-economic dimensions" (United Nations 2002, 13). By focusing on gender mainstreaming, we mitigate potential biases related to specific issues and bring novel evidence to a central dimension of gender politics (gender mainstreaming). Moreover, while primarily quantitative in approach, this article expands the scope of research to include various types of women's presence/alliances beyond numerical representation. Both theoretically and empirically, we incorporate additional layers of the political presence of women and help uncover what type of women's presence matters when it comes to advancing gender mainstreaming.
To address this question, we examine gender equality policies at the local level in Spain. Using a case study involving unique survey data collected for ninety-eight municipalities in the Autonomous Community of Galicia in 2010 and replicated in 2019, we test for a relationship between different types of [End Page 121] women's presence and the development of gender mainstreaming. The focus on the local level is particularly relevant from a gender perspective because municipalities are the level of government closest to citizen demands and needs, and, in the Spanish context, are where gender equality can be put into practice most effectively (Alfama and Paleo 2022; Weldon 2002). This level of government might also offer possibilities for women to participate because local legislatures are reputedly lower in status and geographically closer to women's familial responsibilities (Ortbals, Rincker, and Montoya 2012, 84). As argued in the section dedicated to it, the Spanish case is especially well-suited for local-level studies. Spanish municipalities have gradually acquired greater competence in gender equality, which has created a new opportunity structure for women in government to develop gender policies (Haussman, Sawer, and Vickers 2010). Spain was also hit hard by the global financial crisis that began in 2008, which entailed major budgetary cuts to soft policies. This may have limited women's agency and the implementation of gender policies.
Our study reveals the existence of a descriptive representation loop by which the presence of women in certain positions of power in local political institutions—for example, mayor's offices—increases the number of women in other institutional spaces, such as in the local assembly or in organized civil society. The existence of a descriptive representation loop underscores the relevance of increasing women's presence to foster the virtuous circle of gender representation. Most importantly, women's presence in local politics and the possibility of forming alliances between themselves has the potential to foster gender mainstreaming at the local level, although not all types of presence are equally effective. Our findings offer important insights to reevaluate the relationship between women's political presence and gender equality.
Women's presence and gender mainstreaming
Previous studies have generally concluded that the success of gender policies depends on a combination of factors, which can vary as a function of the specific policy issue being addressed (Htun and Weldon 2010; Mazur 2002; Mazur and Pollack 2009). Although research on the implementation of gender mainstreaming is less abundant, existing studies point to political and organizational will/resistance as key to the success/failure of gender mainstreaming (Alonso 2013, 2015; Benschop and Verloo 2006; Freidenvall and Ramberg 2021; Mergaert and Lombardo 2014; Ylöstalo 2016). Hence, the agency of certain actors—predominantly women—is typically considered a main push factor in the development of gender mainstreaming.
The presence of women in the political sphere is closely linked to the theoretical debates on the descriptive, symbolic, and substantive representation of women and on who represents women's interests, and how, in established democracies (Celis et al. 2014; Lombardo and Meier 2014; Lovenduski 2005). [End Page 122] Previous research has examined the relationship between descriptive representation—the number of women in public institutions—and substantive representation—the inclusion of women's interests and demands on the political agenda (Harder 2023; Mazur 2002; Young 2002). Following Mackay's idea that "the politics of presence should not be confused with notions of critical mass" (Mackay 2004, 101), we argue that the study of the relationship between women's presence and the development of gender mainstreaming should go beyond numbers and address what happens when women are (not) simultaneously present in various spaces, from which they can exert political influence. In this article, women's political presence includes women in executive leadership and in representative positions but also the presence of women in equality organisms or agencies and the presence of the women's movement. We review each of these in detail in the lines below.
The presence of women in executive positions (as prime ministers, mayors, etc.) has been fundamentally related to two types of outcomes: (1) the (gendered) composition and functioning of political institutions and (2) the type of policies that are designed and implemented. As for the first, some scholars have shown that women mayors are likely to increase the number of women in government (Alberti, Diaz-Rioseco, and Visconti 2022), somehow helping to escape the genderedness of organizations (Benschop and Verloo 2006). Escobar-Lemmon and Funk (2018) reach similar conclusions (see also Baskaran and Hessami 2018) but point out that women encounter more obstacles to entering executive positions than legislative ones, which is a major handicap to reaching gender equality in politics and advocating for gender policies. Regarding the type of policies, there is evidence that women mayors are more engaged in financial sustainability than their male peers (Balaguer-Coll and Ivanova-Toneva 2021; Cabaleiro-Casal and Buch-Gómez 2020; Cuadrado-Ballesteros, Guillamón, and Ríos 2021). The presence of women mayors is also positively and significantly linked to a decrease in the gender wage gap (Funk and Molina 2021) and higher spending on women-related issues such as education or healthcare (Funk and Philips 2019), particularly during crisis situations, such as the COVID-19 pandemic (Funk 2020). Moreover, women mayors were found to be more responsive than male mayors to violence against women and girls (Gains and Lowndes 2018). However, some research does not support the existence of a link between the presence of a woman mayor and the establishment of gender equal/balanced policies (Ferreira and Gyourko 2011; Paleo 2015, 2018; Weikart et al. 2007).
There are thus mixed results regarding the potential benefits of women in the executive when it comes to promoting gender equality. This might be related to the fact that the research focuses on gender-specific policies rather than gender mainstreaming. Most importantly, many of these studies indicate that the sex of the mayor does not matter until there is a critical mass of women in political institutions (Childs and Krook 2008). The concept of critical mass has been the most widely used to hypothesize the existence of a [End Page 123] relationship between descriptive and substantive representation, focusing mainly on the presence of women in parliaments. According to critical mass theory, when women's presence surpasses a certain percentage, institutional behavior and public policies begin to change (Kanter 2008).1 Again, the evidence about the relationship between a critical mass of women and substantive changes in gender policies is mixed. Some studies find a positive relationship between the existence of a critical mass and key aspects such as the political style at the institutional level (less conflictual and more friendly or cooperative) (Carey, Niemi, and Powell 1998; Childs 2004; Dolan and Ford 1998; Norris and Lovenduski 1995; Rosenthal 2000); attitudes toward gender equality within institutions (Dolan and Ford 1997; Lovenduski and Norris 2003; Mackay 2005); the survival of women in power (van de Wardt et al. 2021); or the legislative and social outcomes (Betz, Fortunato, and O'Brien 2023; Htun and Power 2006; Koop and Conrad 2021; Lovenduski and Guadagnini 2010; Park and Liang 2021; Swers 2005; Tremblay 2006). However, other scholars question critical mass theory. Some have found that institutions change even in contexts where the number of women does not surpass the critical threshold (Ayata and Tütüncü 2008; Celis 2009; Childs and Krook 2006). Other studies suggest that increasing the number of women has a negative impact on women's substantive representation because the increased presence of women in parliament reduces the intensity of the representation of female interests by male members, who perceive that women are already handling it (Höhmann 2020), or even increases hostility to feminist issues (Grey 2006). Still others find that the effectiveness of critical mass is contingent on the feminist conscience of individual women parliamentarians (Chaney 2006; Childs and Krook 2008; Tremblay and Pelletier 2000), the position they hold in their parties, and/or the ideology of the parties they belong to (Raaum 2005; Russell, Mackay, and McAllister 2002). According to this scholarship, it is not the critical mass that is relevant, but the critical actors (Childs and Krook 2006, 2009).
Critique of critical mass theory has enlarged the focus beyond the numbers of women in political institutions to include additional actors/conditions that are needed to ensure substantive representation. Some studies indicate a need to include the political context—the existence/functioning of particular institutions (Beckwith and Cowell-Meyers 2007; Childs and Krook 2008; Weldon 2002)—and other critical actors such as the women's movement and feminist activism (Mackay 2008; Weldon 2011; Weldon and Htun 2013). According to this line of argument, numbers alone are insufficient to account for the presence of women in politics, one that allows for effectively representing women's interests.
With regard to the political context, some research focuses on gender-equality organisms or agencies within the public institutions, which have been recognized as important actors for pushing gender-equality policies (Allen and Childs 2019; Stetson and Mazur 1995). These may play a relevant role in [End Page 124] identifying gender bias in public policy and proposing and promoting the development of gender mainstreaming (Alonso 2013, 2015; Gains and Lowndes 2018; Otero and Bouzas 2012). These are normally feminized institutions (Paleo 2015) and where the literature has shown that the so-called femocrats or "feminist insiders" (Mazur and McBride 2023)—that is, feminists who explicitly promote gender equality from the state—are likely to be located. Institutionalized bodies of this type cut across party identities and executive/legislative relations. As a consequence, they can promote the reconfiguration of the interests of women representatives, whose affinity and goals may resonate with women clerks and officials who share a "gendered reading of the institution" (Childs 2023, 509–10). Of course, efficiency and the capacity to influence vary according to the resources, the competences, and the situation of the gender-equality organism within government institutions (higher up in the hierarchy means more influence) (Curtin and Teghtsoonian 2010; Otero-Hermida and Bouzas 2020; Sawer 2014; Teghtsoonian and Chappell 2008). Conversely, the lack of a dedicated gender-equality organism in the institutional structure can negatively affect the legislation produced, and thus women's substantive representation, because bills are not preceded by a gender impact assessment (Erikson and Verge 2022, 10). Some authors go so far as to claim that organisms of this type constitute a form of feminist institution-building (Sawer 2020) or that they can operate as "movement institutions—that is, as bureaucratic instances routinely engrained with a protest dimension" (Bereni and Revillard 2018, 407).
Concerning feminist activism, extensive literature has demonstrated the essential role of the feminist or women's movement in defending and promoting women's rights and status (Beckwith 2013). Numerous studies have shown their notable impact on various spheres, such as the institutional structure—through the creation of equality bodies or agencies—in public policies—with the passing of regulations or the development of specific actions in favor of women—or in culture—through the change in values and political and social discourse (Ferree and Hess 1994). At the local level, there is also empirical evidence of the impact of local women's and feminist associations. For example, research on these policies in Spain concluded that in municipalities where there are women's and feminist associations, these policies are more highly developed than in municipalities where there are no such associations. Likewise, around 40 percent of municipalities state that these associations have a great or considerable influence on the development of local gender policies (Paleo 2015).
Despite some variations, all these studies highlight that the presence of women in different spaces with political influence is relevant in the development of gender policies. In fact, different views on the existence of a relationship between descriptive and substantive representation may be an indication that none of these elements alone, but rather the interaction and the alliances that women establish among themselves from the different spaces they [End Page 125] occupy—or presences—is in fact the decisive element for promoting gender equality (Chaney 2012). Existing evidence points in this direction. For example, the effect of women's movements on gender policies is stronger when there is collaboration with institutionalized gender bodies (Andrew 2010; McBride and Mazur 2010) or when the presidency is held by a woman (Vijeyarasa 2021). Steklov, Gofen, and Reingewertz (2023) reported a relationship between the existence of a critical mass and substantive representation under women mayors. A critical mass of women was also found to increase receptiveness to feminist movement demands, thereby favoring the creation of feminist coalitions and the development of gender policies (Banaszak 2003; Ortbals 2008; Paleo 2015; Rodríguez-Garcia 2015). These findings resonate well with previously described notions, such as strategic partnerships (Halsaa 1998), triangles of empowerment (Vargas and Wieringa 1998), velvet triangles (Woodward 2004), feminist strategic partnerships (Mazur 2002), and women's cooperative constellations (Holli 2008). Although each of these concepts includes different sets of actors, the idea they share is that the presence of women promotes the creation of alliances, networks, and cooperative constellations that are crucial to advocating for and boosting gender equality. However, there is scant evidence to support this thesis or, more importantly, to support the relationship between different types of presences and alliances and the development of gender mainstreaming. This work departs from the previous theory to advance the following working hypothesis: As the presence of women increases (in executive leadership positions; in representative positions; as women's "organisms" or the women's movement), the potential to create alliances between these critical actors also increases and gender mainstreaming develops more.
Local gender policies in Spain
Since Spain's transition to democracy in 1978, the country has been structured into three political and territorial levels: state-wide, regional (autonomous communities), and local. The distribution of responsibilities among these levels has evolved through political negotiation. A significant law enacted in 1985 granted local governments open competences, leading to increased municipal activity, particularly in welfare and women's services (Navarro and Velasco 2016). However, funding was often inadequate. The economic crisis and subsequent legislation in 2013 reduced local expenditure capacity, transferring some responsibilities up to autonomous communities and limiting the competences of smaller municipalities (Navarro and Pano 2019).
Local governments, driven by activism and the women's movement, have since the 1980s developed various structures to promote gender equality, including specific departments, technical units, and initiatives within non-specific organisms (Otero-Hermida and Lorenzo 2012; Paleo 2015). Centers [End Page 126] for information on women's rights, launched in the 1980s and expanded in the 1990s, took on a core position in equality efforts (Lombardo 2002). Despite progress in the 2000s, the implementation of gender policies has been inconsistent due to resource deficits.
Municipalities initially addressed women's needs but gradually adopted broader gender strategies, requiring coordination across departments and contributing to gender mainstreaming. National and regional gender laws further obligated local governments to integrate gender perspectives into their policies, highlighted by the 2018 decree mandating local competences in gender equality and the prevention of violence (Alfama and Paleo 2022).
The Spanish local level therefore constitutes a relevant case study to test the relationship between the presence of women and the development of gender mainstreaming. Municipalities have had scarce economic and human resources—particularly after the economic crisis—for implementing gender policies, despite the great potential for action in a relatively underdeveloped policy area. Consequently, the development of gender mainstreaming has depended mainly on the political will and advocacy of local actors. It is in this particular context that women's presence might matter most.
Data and method
The data are based on a telephone survey conducted in 2010 and replicated in 2019 in Galician municipalities with a minimum population of 5,000 inhabitants (there are 113 municipalities in this category). The questionnaire included a large set of items related to the gender composition of the municipalities and the type of gender policies implemented at the local level by the government, including gender mainstreaming. The response rate was 87 percent, or 98 municipalities interviewed for both years of the study (for a total of 196). Of these, the eighty-eight municipalities in 2010 and ninety-six municipalities in 2019 that confirmed implementing local gender policies were used for this study.
Our analysis focused on two aspects of gender mainstreaming strategies at the local level, as identified in the relevant literature (Moser and Moser 2005). The first aspect is the embedding of gender equality within the institution. This means that the gender perspective is incorporated into all design and management processes of the municipality (administrative processes, budgeting, human resources, recruitment policies, mechanisms and indicators to evaluate the gender impact of all municipal policies, etc.). The second aspect is the incorporation of the gender perspective in the content of all public policies that are carried out at the local level and in all sectorial areas. Gender mainstreaming is thus highly cross-cutting, affecting all activities at the local level (McBride and Mazur 2010). With these two theoretically distinguishable dimensions in mind, we used two variables to measure the level of [End Page 127] development of gender mainstreaming at the local level. The first variable—institutionalization of gender equality in the municipal organization—is a composite indicator that considers the number of activities (0 to 6) carried out to promote gender equality within the municipal organization (see Tables A.1 and A.2 in the Online Appendix) and their frequency (never, 0; occasionally, 1; most of the time, 2). The variable took a value of 0 if none of the activities was carried out and a value of 12 if all six activities were carried out most of the time. The second variable aimed at capturing gender mainstreaming in the content of municipal public policies. It was operationalized as the number of gender-sensitive policies (out of a total of thirteen; see Tables A.1 and A.3 in the Online Appendix) being implemented in the municipality. It takes a value of 0 if none of the policies adopted a gender perspective and a value of 13 if all policies were gender mainstreamed.
Table 1 shows variation in the development of these two dimensions of gender mainstreaming. On average, Galician municipalities have applied a gender mainstreaming perspective for nearly eleven of the thirteen policies, indicating that gender mainstreaming is relatively well-developed in the implementation of public policies at the local level. In contrast, the institutionalization of gender mainstreaming within the municipal organization is weaker, as on average (3.2) Galician municipalities only occasionally apply the gender perspective to three of the activities included in the questionnaire (or most of the time to only one activity).
To create the main independent variable, we combined four types of indicators: (1) the (non)existence of a critical mass of women in the local assembly, understood here as 40 percent or greater (following Kanter 2008); (2) the (non)existence of an institutionalized gender equality body (with exclusive or shared competences in gender equality); (3) the sex of the mayor;2 and (4) the (non)presence of registered feminist organizations3 in the municipality. As alluded to in the theory section, the logic behind this variable is cumulative and based on the idea that increasing women's presence and potential alliances between women will push gender mainstreaming forward. As such, the main independent variable accounts for the number of critical actors that are present at the local level. The existence of a critical mass of women councilors and the existence of women mayors reflect the presence of women in institutional
Development of gender mainstreaming in Galician municipalities.
[End Page 128] positions of a representative nature in local governments. In relation to the presence of gender equality organisms or agencies in the municipalities, we assume that there will be women—perhaps femocrats or "feminist insiders," although not necessarily—leading and managing these institutions and that they are susceptible to establishing and generating alliances with other women to promote gender equality. Finally, regarding the existence of feminist organizations in the municipalities, their presence may play an important role in channeling feminist and/or women's demands before the local institution, and concrete alliances may be established with women located within the institution to advance common objectives in gender equality.
The resulting variable assumed values from 0, indicating no presence of women in the local government, to 4, indicating a very strong presence of women. Table 2 provides a detailed description of the categories and distribution. It reveals moderate presence/alliances of women in 34.2 percent of the municipalities, the biggest group. Remarkably, the percentage of municipalities with a very strong presence/alliances of women was almost residual (2.2 percent), and women were not present in 14.7 percent of municipalities.
In the following sections, we start by describing whether and how the different types of women's presence—the components of the main independent variable—relate to each other and to other characteristics of the municipality. This allows us to introduce the notion of a descriptive representation loop while also substantiating our choice to create a categorical variable for women's presence/alliances rather than considering each of the components (critical mass of women in the local assembly, institutionalized gender equality body, woman mayor, and feminist organizations) as separate independent variables. The next section also shows that the empirical strategy matched the theoretical expectation that presence and alliances foster more presence and alliances.
To test our main hypothesis, we ran linear regression models with errors clustered by municipality for each of the dependent variables (institutionalization of gender equality in the municipal organization and gender mainstreaming in the content of municipal gender policies).4 In addition to the main independent variable, the regression models controlled for other factors that are considered important predictors of gender policy development: the size of the municipality, the party in government, and the year in which the survey was conducted. The first, municipal size, has often been correlated to gender policy development in the Spanish context (Gelambí 2005; Roldán 2004; Valiente 1999). The bigger the municipality, the greater the probabilities of gender policies because larger municipalities tend to have more gender-related competences and more people overseeing gender policies than small ones (Paleo 2015). As for the second, the party in government, research has shown that women's movements and women's demands have been more successful when allied with left-wing parties—socialists, communists, social democrats, labor—than when acting without party support or supported by other parties [End Page 129]
Women's presence and alliances at the local level.
(Chappel 2000; Della Porta 2003). Gender policies also tend to be more evolved in contexts where a left-wing party is in power (Paleo 2015). In contrast, gender policy backlash is generally associated with governments run by conservative or right-wing parties (Hankivski 2008; Paleo and Alonso 2014; see also Celis and Childs 2012). Finally, we controlled for the year the survey took place, as the interval covers the period of the global financial crisis and the ensuing regression in gender policies across Europe (Durbin, Page, and Walby 2017; [End Page 130] Kantola and Lombardo 2017; Paleo and Alonso 2014). In Spain, austerity measures following the 2008 economic crisis involved cuts to many social policies, including gender policies. Many equality bodies disappeared, budgets shrank, and the situations of those in charge of implementing those policies became even more precarious (Alfama and Paleo 2022).
These factors were operationalized as follows. The size of the municipality was organized into a four-category variable to identify municipalities from 5,000 to 10,000 inhabitants (1—the reference category); from 10,001 to 20,000 inhabitants (2); from 20,001 to 50,000 inhabitants (3); and more than 50,000 inhabitants (4). The party in government became a categorical variable with a value of 1 if the party in government was mainstream right-wing (Partido Popular, PP—the reference category); 2 if the party in government was mainstream left-wing (Partido Socialista, PSOE); and 3 if any other party was in government (Bloque Nacionalista Galego, BNG, a regionalist ideological catch-all party; the left-wing Mareas party; or independent local parties). A dummy variable was included to account for the year of the survey with 2010 as the reference category. Descriptions of the control variables can be found in Table A.5 of the Online Appendix.
The descriptive representation loop
One normative ideal of gender representation studies affirms that increasing the number of women in power will not only promote substantive representation but also foster women's presence in politics as well. In other words, there should be a sort of descriptive representation loop whereby the presence of women in political institutions would encourage the involvement of other women. There is indeed evidence that having a woman mayor is related to a greater presence of women in local executive and legislative offices (Alberti, Diaz-Rioseco, and Visconti 2022). The presence of women also pushes for the creation of gender equality organisms within institutions (Childs 2023) and favors the work of the women's movement (Vijeyarasa 2021). The indicators we used to create our main independent variable (critical mass of women in the local assembly, institutionalized gender equality body, woman mayor, and feminist organizations) may be highly intercorrelated. Is this really the case? Does the presence of women in politics attract other women? Figure 1 presents the percentage of municipalities led by a woman mayor, and in which 40 percent or more of the local assembly were women (dark gray bars) along with the percentage of municipalities led by a male mayor in which 40 percent or more of the local assembly were women (light gray bars), depending on whether there were registered feminist organizations and an institutionalized gender equality body at the local level. Other characteristics were taken into account that have generally been found to relate to the presence of women in political institutions: the party in government (leftist parties are more likely to [End Page 131]
Descriptive representation loop: the presence of women in local assemblies by the sex of the mayor and the characteristics of the Galician municipalities.
promote women's presence than right-oriented parties (Otero-Hermida and Bouzas 2020)) and the size of the municipality (in larger municipalities with larger and more professionalized populations it is more plausible to have women in power (Paleo 2015)).
The small percentage of women mayors in our sample (16.8 percent) confirms that women have difficulty reaching leadership positions (in line with Escobar-Lemmon and Funk 2018). For this reason, figure 1 includes the percentage of municipalities for each pair of categories. For example, the top dark-gray bar on the graph indicates that there is a critical mass of at least 40 percent women in the local assembly in more than 66 percent of the municipalities governed by a woman mayor and which had registered feminist associations. Critical mass was achieved in 56 percent of the municipalities that had registered feminist associations and were led by men (top light-gray bar in the graph).
Figure 1 shows that the Spanish case meets earlier expectations. There seems to be a descriptive representation loop relating women's presence (as mayors, municipal councilors, members of the gender equality body, or feminist associations) to a stronger presence of women, especially when there are favorable characteristics (big municipalities or left-wing parties). Notably, the percentage of municipalities with a critical mass of women is higher where women are mayors than where men preside. This is a constant finding, independently of other kinds of women's presence (in feminist associations or gender equality bodies) and other local characteristics.5 The presence of women in the local assembly is also positively and significantly related to the presence [End Page 132] of feminist associations in the municipality (significant at P < .05): the stronger the presence of feminist associations at the local level, the stronger the presence of women in the local assembly. Similarly, critical mass is more likely in the local assemblies of municipalities that had a gender equality body than in those that did not (significant at P < .1). Having a woman mayor tended to push the presence of women in local assemblies even more when there were feminist associations or a gender equality body.
This study corroborates others with its finding that the party in government and the size of the municipality also appear to be important predictors of the presence of women at the local level. Right-wing PP governments were associated with a lower presence of women in local assemblies, whereas left-oriented governments (PSOE and most parties in the "other" category) tended to have more women in local assemblies (marginally significant differences, P < .108). Moreover, the presence of women in local assemblies increased with the size of the municipality (differences statistically significant at P < .000). Only 47 percent of the smallest municipalities achieved critical mass, compared with more than 90 percent of the larger municipalities. Again, figure 1 shows that women mayors tended to take advantage of these favorable characteristics to promote women's presence in local assemblies.
The information in figure 1 therefore supports the existence of a descriptive representation loop: women tended to boost other women's presence in local governments, especially where a left-wing party was in power and/or in larger municipalities. This finding may help contextualize the noteworthy critique that the presence of women could be endogenous to the characteristics of the municipality (Ferreira and Gyourko 2011). While figure 1 suggests that women's presence depends on the size of the municipality and the party in power, it also underscores the relevance of increasing women's presence to foster the virtuous circle of gender representation. The existence of a representative feedback loop also implies that the four components of our main independent variable (the presence/alliances of women) are interrelated. This important dynamic is overlooked when each of the components is considered as a differentiated factor. It further substantiates our empirical strategy, which is applied in the following section.
Women's presence/alliances and gender mainstreaming
The main argument in this article is that the increased presence of women and potential alliances is positively related to the development of gender mainstreaming. Figure 2 shows the distribution of the two dependent variables—institutionalization of gender equality in the municipal organization (left-hand quadrant) and gender mainstreaming in the content of municipal public policies (right-hand quadrant)—by type of women's presence/alliances (from [End Page 133]
The development of gender mainstreaming in Galician municipalities by type of women's presence/alliances.
no presence to a very strong presence). The graphs show how gender mainstreaming advances as the presence of women becomes stronger in the municipality. The average number of public policies applying gender mainstreaming increased from less than ten in municipalities with no presence of women to almost twelve in those with a very strong presence of women at the local level (differences statistically significant at P < .10). Yet it is the institutionalization of gender equality in the municipal organization that is most strongly pushed once there is a strong or very strong presence of women at the local level. Figure 2 shows that the average number of actions of gender mainstreaming implemented internally in the municipal organization is twice as large in the municipalities with a very strong presence of women as those with no presence (differences statistically significant at P < .05). In figure 2, the relationship between the presence/alliances of women at the local level and the development of gender mainstreaming is clearly visible.
To further test our hypothesis, we estimated two linear regression models that included several controls. The previous section underscored the importance of the size of the municipality and the party in government as correlates of women's presence in the local assembly. We thus need to control for them to assess whether women's presence/alliances are related to gender mainstreaming independently of municipal characteristics. Table 3 presents the results of the estimates for each of the dependent variables: institutionalization of gender equality in the municipal organization (column 2) and gender mainstreaming in the content of municipal public policies (column 3). The results broadly confirm our expectations: stronger presence/alliances of women were positively and significantly related to a greater development of gender mainstreaming, even after controlling for the size of the municipality and the party in government. In fact, only the size of the municipality is marginally significant. The data also reveal the distinct effect of the global financial crisis on the institutionalization of gender equality in municipal organizations and (even more so) in municipal gender equality policies: gender mainstreaming was less developed in 2019 than in 2010. The economic crisis led to [End Page 134]
Women s presence/alliances and gender mainstreaming: linear regression models.
significant budget cuts at the local level, which apparently impacted municipal capacity to continue developing gender mainstreaming.
Although women's presence/alliances are generally and positively correlated to gender mainstreaming, meaningful differences between the two dependent variables have implications for interpreting the results. First, gender mainstreaming in the content of municipal public policies starts to develop with a moderate presence of women at the local level and increases gradually as women's presence/alliances increase. Table 1 indicated that gender [End Page 135] mainstreaming was highly developed in municipal public policies, with an implementation average of 10.7 policies out of thirteen and more than 70 percent of the municipalities applying gender mainstreaming in ten or more policy areas. These results seem to indicate that municipalities are likely to develop gender mainstreaming across different policy areas even without a strong presence of women. This does not invalidate our findings that women's presence/alliances foster the development of gender mainstreaming in public policies. Rather, it suggests that gender mainstreaming is relatively well-entrenched in local public policies and that a strong presence of women might not be entirely indispensable to reach a fair level of gender mainstreaming.
Quite the contrary, the institutionalization of gender equality in municipal organizations seems much more difficult to achieve and dependent on the presence of women at the local level. Table 3 shows that gender mainstreaming only began to develop when women's presence/alliances were strong or very strong. Indeed, Table 1 displayed weak institutionalization of gender mainstreaming in the municipal organization, indicating that over 60 percent of the municipalities applied a gender perspective in only one of the activities linked to budgeting, evaluation, or training (see Table A.2 in the Online Appendix). This all suggests that the strong presence/alliances of women are needed to promote gender mainstreaming within the organization.
Conclusions
The gendered dimension of representation is a very relevant topic in gender studies. In this article, we attempted to contribute to this strand of research by revisiting the link between descriptive and substantive gender representation in relation to gender mainstreaming. Three main innovations were introduced in this study. First, we focused on gender mainstreaming rather on than the subsectors that make up gender policies. This allowed us to provide new empirical data on the implementation of the gender mainstreaming strategy from a holistic perspective, an approach that needs to be further developed. Second, we proposed the notion of women's presence/alliances to qualitatively measure women's presence beyond a head count of women in politics. Third, we looked at the local level—the closest to citizens—and the specific case study of Galicia in Spain. The focus on the local level is particularly relevant from a gender perspective because it is the level of government where gender equality can be put into practice most effectively, at least in Spain.
Our study revealed a descriptive representation loop, or virtuous circle, in which women's presence in local government promotes the presence of other women. Women's presence in politics seems thus to be an important entry point for other women. Crucially, we also found that descriptive representation relates positively to substantive representation. The stronger the presence/alliances of women at the local level, the more gender mainstreaming is [End Page 136] developed. However, gender mainstreaming in the content of public policies appears to be more developed and less dependent on women's presence/alliances at the local level than institutionalization of gender equality in the municipal organization. Our results showed that the latter was relatively weak and highly dependent on the strong presence/alliances of women.
These findings indicate that women's presence and alliances within municipal organizations—as mayors, members of the local assembly, or "femocrats" in gender equality organisms—and outside the organization in the women's movement are fundamental to overcoming resistance and removing obstacles that prevent the implementation of gender policies and the progress of gender mainstreaming within the organization. Our study suggests that women's presence/alliances can counteract some of the main problems for the successful advancement of gender mainstreaming: lack of political will, resources, technical resistance, and so on. Women's presence and ability to form alliances and coalitions is thus key to defending gender equality interests (Harder 2023), probably because women are generally more connected to different interests, including those of the marginalized (Celis and Childs 2023). Even if cautiously, this idea is further supported by our data. Table 3 clearly shows that gender mainstreaming suffered an important setback after the economic crisis in Spain. Apparently, women's presence/alliances were also fundamental to preserving gender mainstreaming in times of crises. Further analysis of our data suggests that, at least in the Spanish case, gender mainstreaming seems to have been better protected from the economic crisis in municipalities with a strong presence/alliances of women than in those with a null or weak presence/alliances of women.6
Although generally convincing, this study has certain limitations. Perhaps the most important is that the results are based on a single case study. We described the specificities of the Spanish case, which suggest that the pursuit of gender mainstreaming might be more dependent on the (lack of) political will than in other contexts which have enjoyed a long tradition of gender equality policies. A comparative study would be needed to fully assess whether our findings transfer well to other contexts (see Erzeel and Rashkova 2023). Another constraint is that presence does not equate to alliances. Our theoretical assumption was that the presence of women favors their alliances. However, the qualitative measure of women's presence we introduced did not distinguish between cases of presence only and those involving both presence and alliances. Further qualitative efforts are needed to fully understand when and how women's alliances relate to achievements in gender mainstreaming.
Supplementary Material
Notes
1. There is variation concerning the threshold that identifies when the critical mass is reached. Here we align with the most conservative threshold of 40 percent suggested by Kanter (2008).
2. As reported by the person who took the survey.
3. In Spain, municipalities usually have an official register of the associations that are constituted in their respective territories and what their purposes are. In this study, the municipalities were asked whether there were registered associations in their territories that self-identified as feminist associations.
4. The analyses were replicated using multilevel mixed-effects linear regression and the results remained fundamentally unchanged (see Table A.4 in the Online Appendix).
5. If the number of women and men mayors were balanced, the percentage of municipalities with a critical mass would increase by almost 200%. See Table A.6 in the Online Appendix.
6. To test for the differential impact of women's presence/alliances before/after the economic crisis, an interaction term was included in the linear regression model between our main independent variable (women's presence/alliances) and the year of the survey. The interaction term was significant for "strong presence/alliances of women" and "very strong presence/alliances of women" categories.
Funding
This work was supported by the Spanish Government (MCIN/AEI/10.13039/501100011033; grant MICIN-AEI PID2020-115568RB-I00). The financial institution had no role in the design, implementation, analysis, and interpretation of the data or in the writing of the paper.
Conflict of interest
The authors—Mónica Ferrín and Natalia Paleo—declare there are none.
Data availability
The data underlying this article are available in the article and in its online supplementary material.
References
This work is licensed under a Creative Commons Attribution 4.0 International License.
