Ideology and Party Positions on Gender Issues in Spain: Evidence from a Novel Data Set

Abstract

This article introduces new quantitative fine-tuned indicators to objectively measure political parties’ preferences on gender issues. We assess the validity and reliability of these new empirical indicators by analyzing the relationship between ideology and gender positions in decentralized Spain. Using data collected by the Regional Manifestos Project, which for the first time has incorporated a fully fledged gender domain into its coding scheme for content analysis of regional manifestos, we analyze parties’ gender positions on four fundamental dimensions: welfare and the labor market, violence, representation, and values and identity. The results suggest that there is a persistent left-right divide on the last three dimensions: Spanish left-wing parties score significantly higher than right-wing and regionalist parties. However, ideology does not drive parties’ gender positions on welfare and labor market policies; support for gender equality policies on this dimension is so widespread that it can be considered a “valence issue”—all parties, irrespective of their ideology, endorse the same (positive) position.

Gender equality has become a highly contentious and polarized issue in Western European politics. Populist radical right parties have turned the democratic agenda for gender equality into a political battlefield, making the fight against “gender ideology” one of the fundamental banners of their illiberal programmatic mix. As a result, party competition over gender equality has become increasingly polarized at the local, national, and even European levels. According to some researchers, the future of liberal representative democracy itself is at stake (Graff and Korolczuk 2022).

Despite the relevance of gender politics for understanding the populist shift in Western European political systems (Alonso and Espinosa-Fajardo 2021; Caravantes 2020; Graff and Korolczuk 2022; Kantola and Lombardo 2021), there are no comparative quantitative data with which to measure partisan stances on gender issues (O’Brien 2018). Moreover, only a handful of academic publications analyze party manifestos from a gender perspective (Beloshitzkaya 2020; Greene and O’Brien 2016; Homola 2019; Kittilson 2011; O’Brien 2018; Sanders, Gains, and Annesley 2021; Weeks 2018, 2022). The absence of comparable indicators of parties’ gender positions across space and time has limited the types of research questions that the literature can address. Research on women’s political representation has privileged the use of policy-related indicators, such as the implementation of gender quotas (Verge and de la Fuente 2014) or the extension of work-family policies (Korpi, Ferrarini, and Englundt 2013; Morgan 2013), while much less attention has been paid to the study of parties’ programmatic agendas (Wolbrecht 2002) through a “gender lens” (Alonso and Espinosa-Fajardo 2021, 20; O’Brien 2018).

In this article, we present new fine-tuned indicators to objectively measure gender salience and parties’ positions based on a quantitative content analysis of regional election manifestos. Our indicators measure parties’ stances on four gender dimensions: welfare and the labor market, protection against gender violence, equal representation, and cultural values and identity. We use the 2019 Spanish regional elections to test the newly developed indicators. Our objective in this article is to check the validity and reliability of these new empirical indicators. To do so, we use them to test three widely accepted hypotheses of the relationship between ideology and partisan stances on gender issues—namely, the “traditional” hypothesis, the “catching-up” hypothesis, and the “framing” hypothesis. Our main research questions are as follows: Do our indicators produce robust findings? If they do, what are the implications for a future research agenda studying parties’ manifestos from a gender perspective?

Our results confirm previous findings in the literature. First, the left-right divide is persistent on three of the four gender dimensions that we identify. Left parties score higher than right parties on the scales for protection against gender violence, equal representation, and, above all, cultural values and identity. Second, ideological differences have faded on the dimension of welfare and labor market policies. All parties, irrespective of ideology, defend the public provision of childcare services, the promotion of women’s entry into the labor market, equality in employment, and the encouragement of equal distribution of caring responsibilities between women and men. Third, ideology is still relevant to explain the framing of “gender (in)equality” by political parties. The right frames gender inequality as a problem that can be solved by family and welfare policies and labor market integration, while the left also supports other status-based noneconomic policies among its gender programmatic appeals.

Since our results are in line with the findings of other scholars, we conclude that our methodology is robust and that our newly developed indicators of partisan stances on gender issues are valid and reliable. The advantage of our indicators is that they are measured following the same methodology as the Manifesto Project, the most widely used source of parties’ policy positions by scholars. This means that, first, our data set can easily be scaled up to include more elections and more countries, and, second, our salience and position gender scores can be combined with any other Manifesto Project policy scores. Our study confirms previous findings showing that parties’ policy preferences on gender issues are multidimensional. Therefore, party ideology is not always a good predictor of parties’ gender positions. Moreover, our indicators, when scaled up to include more countries and elections, will allow scholars to test hypotheses on gender politics that remain unexplored from a comparative perspective, including, on the one hand, subnational political dynamics and, on the other, the divides between old and new politics and between radical left and radical right populism, and their impact on the quality of liberal democracy.

The Study of Gender Politics Using Parties’ Manifestos

The Absence of Gender-Specific Categories in the Codification of Party Manifestos

The number of studies that use party manifestos as data to measure partisan stances on gender issues is relatively small (Alonso and Espinosa-Fajardo 2021; Beloshitzkaya 2020; De Lange and Mügge 2015; Greene and O’Brien 2016; Homola 2019; Kantola and Lombardo 2019; Kittilson 2011; O’Brien 2018; Sanders, Gains, and Annesley 2021; Weeks 2018), despite the fact that parties’ manifestos are one of the most reliable sources to estimate their policy preferences. These official documents represent the voice of the whole party (not of a leader or a section), and therefore they cover the collective expression of the policy preferences of the party. In addition, they are easily accessible and published on a regular basis by almost all political parties, allowing comparisons over time and space (Alonso, Volkens, and Gómez 2012).

The Manifesto Project (also called the Comparative Manifesto Project or CMP) has generated one of the most widely used data sets for studying the policy preferences of political parties based on a quantitative content analysis of electoral manifestos. It covers hundreds of parties from more than 50 countries since 1945 and provides estimates of party positions on a wide range of policy issues. However, the coding scheme used by the Manifesto Project does not include specific categories for gender. References to gender-related issues in the parties’ electoral programs are only partially classified under general categories such as “equality” (503), which includes references to the end of sexual discrimination; “traditional morality” (603/604), which may contain positive/negative references to the maintenance and stability of the traditional family’ or “noneconomic demographic groups” (706), which comprises favorable mentions of women. However, these are broad categories that also encompass other policy issues that bear no relation with gender.

The fact that the most widely used source for identifying parties’ policy positions, the Manifesto Project, does not contain specific gender categories is surprising, and regrettable, on two counts. First, it is surprising given the abundance of literature that uses election manifestos to analyze a variety of other topics and policy issues, such as ideology, immigration, welfare, nationalism, and Europeanism (Volkens et al. 2015). Second, it is surprising given the relevance of gender politics to understand the realignment of Western European party systems since the 1980s and the recent surge of left and right populism, which questions the liberal core of established representative democracies (Mudde and Rovira Kaltwasser 2013). Feminist scholarship argues that the recent political polarization over gender equality is “part of a broader conflict, where what is truly at stake is the future of democracy” (Graff and Korolczuk 2022, 3). Several studies have analyzed the illiberal core of populist radical right parties’ anti-feminist and anti–gender equality stances and concluded that this illiberalism represents a danger to democracy (Alonso and Espinosa-Fajardo 2021; Caravantes 2020; Graff and Korolczuk 2022; Kantola and Lombardo 2019, 2021).

Although it may be understandable that the Manifesto Project’s founders could not see the centrality of gender for the political realignment of Western European party systems at the time of the project’s inception in 1979, this is no longer the case. As Graff and Korolczuk (2022, 166, 169) have argued, gender is “[n]ot a marginal controversy, no mere ‘cultural’ issue” but instead “a key battleground for the redefinition of the political scene”; and yet, party manifestos “are less commonly used in the women and politics literature, in large part because the CMP does not code references to women” (O’Brien 2018, 41).

The consequence of this lack of comparative gender salience and position indicators has been the neglect of parties’ programmatic appeals in the study of women’s representation and gender politics (O’Brien 2018). Only a handful of published articles analyze party manifestos from a gender perspective, and, of these, even fewer are quantitative comparative studies (Beloshitzkaya 2020; Greene and O’Brien 2016; Homola 2019; Kittilson 2011; O’Brien 2018; Weeks 2018), while the rest are country case studies (Chaney 2007; Hicks and Buccus 2012; Murray and Sénac 2014; Ono and Miwa 2020; Sanders, Gains, and Annesley 2021; Vetten and Leisegang 2012; Vyas-Doorgapersad 2015) or party case studies (Alonso and Espinosa-Fajardo 2021; Childs, Webb, and Marthaler 2010; Kantola and Lombardo 2019).

Novel Indicators to Measure Party Preferences on Gender Issues

This article draws on the Regional Manifestos Project (RMP) to propose a new coding scheme for measuring partisan stances on women and gender politics. The RMP is a research project originated in 2010 and funded by the government of Spain under the National Plan for Scientific Research. Based on the same methodology as the Manifesto Project, the RMP measures parties’ policy preferences and positions in regional (substate) elections (see Alonso, Gómez, and Cabeza 2013).¹ The RMP team has incorporated a new domain of gender-related categories into its coding scheme (Cabeza, Gómez, and Pérez-Tirado 2023). This new coding scheme allows researchers to track both positive and negative mentions of key issues in gender studies, including equal pay, childcare, violence against women, parental leave, gender quotas, and sexual and reproductive rights.

The new gender domain in the RMP coding scheme

The new domain on gender issues in the coding scheme of the RMP is composed of eight codes that identify party preferences along four fundamental policy areas or dimensions: (1) welfare and the labor market, (2) gender violence, (3) equal representation, and (4) cultural values and gender identity (see Table 1). The reason for choosing these specific categories is twofold. First, the new gender domain exhaustively covers the issues that electoral programs actually address and that coincide with the central areas of government action identified by feminist studies (Mazur and Pollack 2009). In this way, any reference—positive or negative—to gender equality in the parties’ manifestos can be classified using one of the proposed categories. Second, the new gender domain considers the multidimensional nature of gender issues. Previous literature has identified different typologies of gender issues (Annesley, Engeli, and Gains 2015; Htun and Weldon 2010). These studies distinguish between class-based issues and status-based issues. Class-based issues, which would primarily benefit women with fewer resources, are those that address women’s economic integration and the sexual division of labor, while status-based issues, aimed at alleviating the subordinate position of women as a group, are those that deal with women’s integrity and rights. Another categorization has been proposed by Erzeel and Celis (2016), who differentiate between feminist and gendered interests.

Table 1. New gender-specific categories in the RMP coding scheme

Note: See the Supplementary Material for full definitions of each code.

Source: RMP Coding Handbook, 2020 (available at www.regionalmanifestosproject.com).

On the first dimension of the gender domain, we find all social welfare benefits and policies that seek to promote women’s economic independence and equality in the labor market. Broadly speaking, these are the issues that previous literature has identified as class-based, including parental leave, equal pay and working conditions, public childcare provision, work-family reconciliation policies, or the promotion of shared responsibilities within the household, among others. The positive code (801) is applied when the party is favorable to the establishment or maintenance of such policies. The negative code (802), on the contrary, applies to references against their implementation: budget cut proposals, positive references to the male breadwinner model, and so on.

The second, third, and fourth dimensions of the gender domain distinguish between the main subtypes of status-based issues. Thus, the second dimension focuses on gender violence. The positive code (803) includes favorable mentions of the laws on violence against women, the protection and support of victims, or the tightening of penalties for crimes such as abuse or rape. The negative code (804) is applied in the quasi sentences in which the party defends the abolition of laws against gender violence; is favorable to considering it a private matter; justifies or minimizes rape, abuse, or violence against women; questions victim’s credibility; and so on. The third dimension is about equal representation. The positive code (805) is devoted to quasi sentences in which the party supports the establishment or maintenance of gender quotas, denounces the underrepresentation of women in any sphere of public life (political, administrative, businesses, culture, sports, etc.), or advocates for a greater balance between men and women. The negative code (806) is used in cases in which the party opposes gender quotas, parity laws, or women’s presence in any area. Finally, the fourth dimension focuses on cultural values and gender identity. The positive code (807) includes positive references to feminism and the principle of gender equality in general or support for modern family composition, reproductive rights (abortion, contraception), LGBT rights, etc. The negative code (808) is applied when the party stands for patriarchal cultural values and norms, such as support for the traditional family and traditional gender roles of male dominance, opposition to abortion or same-sex marriage, negative references to gender equality and/or feminism, and so on.

Gender salience and position indicators

Based on these codes, we computed indicators of gender salience and position (Table 2). We calculate parties’ positions on gender issues (both in general and in each of the specific dimensions of the gender domain) by subtracting the salience score of the “negative” codes from that of the “positive” codes. Thus, values greater than 0 would indicate a pro–gender equality position, while values less than 0 would reflect an anti–gender equality position. The resulting variables theoretically range from –100 (for a program entirely dedicated to negative categories) to 100 (for a program entirely dedicated to positive codes).

Table 2. Indicators of party preferences on gender issues

Source: RMP data set (available at www.regionalmanifestosproject.com).

Testing the Robustness of the RMP Novel Indicators: Three Hypotheses on the Effect of Ideology on Parties’ Gender Positions

We test the validity of our newly developed indicators on three hypotheses put forward by scholars studying the relationship between parties’ ideology and their positions on gender: the “traditional” hypothesis, the “catching-up” hypothesis, and the “framing” hypothesis. There is wide agreement among scholars that gender equality in general, and women’s political representation in particular, has been an agenda of the left to a larger degree than the right. In fact, egalitarianism is one of the fundamental principles of the left (Bobbio 1996). Traditionally, left-wing parties have represented the interests of the most disadvantaged groups in society, including women. A vast literature on ideology and gender has shown that left-wing parties nominate more women as candidates than right-wing parties (Duverger 1955), have a higher proportion of female parliamentarians (Kittilson 1999; O’Brien 2018), and are more sensitive to women’s demands for gender equality (Caravantes 2020; Kantola and Lombardo 2019; Sacchet 2008).

Based on these points, we propose the following “traditional” hypothesis with which to test our data:

H₁: Left-wing parties adopt a more pro–gender equality position than right-wing parties.

Nonetheless, scholars have begun to question the left’s ownership of women’s interests and to fine-tune the role of political ideology in relation to gender politics and women’s representation (Morgan 2013). Academic scholarship engaged in this fine-tuning exercise argues that the right has been catching up with the left since the spread of postmaterialist values in the 1980s (Graff and Korolczuk 2022; Inglehart 2008; Inglehart and Norris 2003; Lovenduski and Norris 2003; Morgan 2013; O’Brien 2018). New green and new left parties were the political offspring of the spread of postmaterialist values, an adjustment on the part of political supply that reflected changes in political demand. These demand- and supply-side changes forced mainstream right parties (conservatives, Christian democrats, and liberals) to incorporate women’s concerns into their political platforms and to address women’s demands in their actions (Bale and Rovira Kaltwasser 2021). It also forced the traditional parties of the left (socialists and communists) to reformulate the way they addressed gender in their platforms, party organizations, and policies, which, until then, had focused mainly on the material aspects of equality and ignored gender in terms of rights and identity (Keith and Verge 2018; Kittilson 2011).

Ignazi (1992) has further argued that the emergence and success of new radical right parties around the same time that green and new left parties were collecting electoral victories was but a reaction to the expansion of these postmaterialist, emancipative, and mostly left cultural values. Ignazi (1992) labeled this phenomenon the “silent counter-revolution,” a return—with a twist—to cultural values associated with the right (order, traditional family, patriarchy, nationalism, national security). Research has shown that even the populist radical right party family, which is commonly characterized as staunchly anti-feminist and illiberal, is internally heterogeneous and nuanced in its positions on gender (Akkerman and Hagelund 2007; De Lange and Mügge 2015). If they are electorally struggling and suffering from a gender gap in voter support, radical right populist parties are more prone to engage in “strategic descriptive representation”—that is, to elect more women members of Parliament (MPs) to attract women voters (Weeks et al. 2023).

Kittilson’s (2011) pioneering comparative study using party manifestos as empirical indicators of women’s substantive representation was among the first to alert us that attention to women’s interests transcends ideological blocs, as it is present in left-wing as well as right-wing party manifestos. What varies both within and across ideological blocs is the specific meaning that parties give to women’s interests and the policy areas attached to them. For example, Kittilson (2011) found that green, communist, and social democratic parties, but also agrarian parties, were the most likely to emphasize issues of social justice and that social democratic and conservative parties were equally likely to mention education in connection to women’s issues. Relying on party labels, therefore, can be misleading since “party labels obscure a lot of ideological variation between parties and ignore the varying stance that parties might take on different ideological left-right scales” (Erzeel and Celis 2016, 577).

As a result of the political realignment around postmaterialist values, the emphasis given to gender issues on party platforms and the parties’ actions in support of women’s demands are no longer the exclusive prerogative of the left (Celis and Erzeel 2015; Childs and Webb 2012; Sanders, Gains, and Annesley 2021). The spread of postmaterialist values and the changes to political supply in the form of new left and radical right parties has made it difficult for the traditional right to ignore women’s demands in their platforms and actions, for three reasons. First, conservative, Christian democratic, and liberal parties have women within their ranks who are mobilizing on gender issues internally (Keith and Verge 2018; Kittilson 2011). Second, there is a strong feminist movement mobilizing on gender issues in the streets (Molyneux et al. 2021). Third, there is a “contagion effect,” stemming from interparty electoral competition, that cuts across ideological blocs (Cowell-Meyers 2017; Weeks 2018). Among the consequences of this pressure to converge around women’s political representation, we find the increasing adoption of gender mainstreaming in policy making (True 2003) and the widespread establishment of gender quotas around the globe (Norris and Dahlerup 2015) across ideological divides. This leads us to formulate the “catching-up” hypothesis:

H₂: Right-wing parties are as pro–gender equality as left-wing parties. In other words, there are no differences between the left and the right in the position they adopt on gender issues.

This does not mean, however, that the right addresses gender issues using the same ideological parameters as the left. “Gender” and “equality” are “contested concepts that can hold different (often implicit) definitions of the problem and the solution” (Alonso and Espinosa-Fajardo 2021, 9–10; see also Kantola and Lombardo 2021). Right parties frame women’s interests, an issue they do not own, in a way that does not harm their overall credibility as parties of the right (Petrocik 1996). Research shows that right parties, for example, avoid using a feminist critical approach, in contrast with parties of the left, particularly green parties, which collaborate closely with the feminist movement in this respect (Celis and Erzeel 2015). The right tends to defend family values and women’s traditional roles through the use of redistributive policies that benefit women (Graff and Korolczuk 2022; O’Brien 2018) and an emphasis on work-family policies (Morgan 2013).

Although populist radical right parties tend to be staunchly anti-feminist, their framing of gender equality is complex and varies from party to party. While some populist radical right parties defend certain aspects of gender equality that are traditionally defended by the liberal left (such as divorce laws, women’s legal and financial independence, women’s participation in the labor market, etc.) (Akkerman 2015; Akkerman and Hagelund 2007), many parties in this party family reject other relevant aspects of it (the use of gender quotas, the fight against gender violence) as examples of feminist “gender ideology.” It is common among populist radical right parties to argue that gender equality has been achieved (Alonso and Espinosa-Fajardo 2021; Kantola and Lombardo 2019). At the same time, however, many of these parties instrumentalize gender equality against immigration and multiculturalism, arguing that gender equality must be defended from attacks by immigrants, immigration-friendly policies, and the multiculturalist perspectives of the left (Akkerman and Hagelund 2007; De Lange and Mügge 2015). Many more comparative analyses are necessary before we can begin to understand what factors explain the differences and similarities found within this party family, and with what consequences.

Therefore, in the light of these analyses of the frames used by right and left parties to address gender issues, we derive our “framing” hypothesis:

H₃: Parties on the left and on the right address gender issues in diverse ways: left-wing parties adopt a more pro–gender equality position than right-wing parties on issues related to gender identity and women’s rights, whereas there are no differences between the left and the right on issues related to more material aspects, such as welfare and economic conditions.

Context and Cases: Decentralized Spain

Spain offers an interesting context for examining party positions on gender issues, for several reasons. First, its gender regime has undergone a rapid and deep transformation in recent decades (Lombardo 2016). Not only has Spain placed itself at the vanguard of gender equality policies in Europe (Valiente 2008), but also huge strides have been made in women’s political representation. In 2018, Spain had the largest share of female ministers in the European Union. In this context, and after the massive marches of the feminist movement in 2018, gender issues have become a salient topic in the political debate.

Second, Spain is home to two new populist parties on the left and the right of the political spectrum: Podemos and Vox, respectively. Feminist scholars have argued for the need to incorporate a gender lens when studying the influence of populism on established democracies (Alonso and Espinosa-Fajardo 2021; Caravantes 2020; Kantola and Lombardo 2021). Decentralized Spain offers the opportunity to study the differences between left and right populism, and its influence on the behavior of mainstream socialist and conservative parties, while keeping constant a whole set of country-level variables. The arrival of Podemos and Vox has made gender a central battleground of partisan competition and polarized partisan stances on gender issues in Spain. In fact, women’s rights took center stage in both the April and November 2019 Spanish general election campaigns. Prime Minister Pedro Sánchez (Spanish Socialist Workers’ Party) made gender issues a central theme of his campaign. With an openly anti-feminist party—the radical right Vox—on the rise, gender issues also featured prominently in the election campaign of the center-right Ciudadanos and the conservative Popular Party. Thus, we expect Spanish political parties to pay significant attention to gender-related issues in their manifestos.

Third, Spain offers a perfect setting for studying the relevance of subnational political dynamics. Scholars have described Spain as a “federation in all but name” (Elazar 1991, 227). Although it is formally a unitary state, Spain functions as a “de facto federal system” (Hueglin and Fenna 2006, 19) with 17 regions—called Autonomous Communities—that have a high degree of political power. Gender equality policies are largely decentralized in Spain. Each regional government develops its own policies on gender equality. These policies mostly fall under the responsibility of the Autonomous Communities. The regions also lead policy innovation in this field: they have acted as pioneers in the introduction of several equality policies that were later adopted at the national level (Alonso and Verge 2014). For instance, the Equal Status Act (Organic Law 3/2007, for the Effective Equality of Men and Women) that came into force in 2007 in Spain with the purpose of eradicating discrimination in the workplace was largely inspired by a pioneering law adopted in the Basque Country (Law 4/2005, for the Equality of Women and Men).

We consider the focus on (subnational) regional elections advantageous for two reasons. First, existing research suggests that decentralization facilitates women’s political representation: subnational political arenas tend to be more accessible to women, which, in turn, may lead to the formulation of policies that are more responsive to women’s needs at lower tiers of government (Kantola and Lombardo 2019; Khan 2011; Stockemer and Tremblay 2015; Vengroff, Nyiri, and Fugiero 2003; Vickers 2011). Thus, we can expect gender issues to be especially salient in party manifestos for regional elections. The second benefit of focusing on the regional level is methodological. We have enough cases and interregional variation to test the literature’s hypotheses on how ideology affects party positions on gender issues, while we keep “constant” some exogenous and country-specific factors that may interfere with our dependent variable, like those related to the social, economic, and political context.

The 2019 regional elections in Spain, which are the focus of this article, were held in 13 of the 17 Spanish Autonomous Communities: Aragon, Asturias, the Balearic Islands, the Canary Islands, Cantabria, Castile and León, Castilla–La Mancha, Extremadura, La Rioja, Madrid, Murcia, Navarre, and the Valencian Community. The RMP provides data on the gender positions of 76 parties. We include all of them in our analysis. The number of parties varies from one region to the other, depending on how many parties obtained representation in the regional parliament: from as many as nine parties in the Balearic Islands to only three in Castile–La Mancha.²

Since the beginning of the 1980s, Spanish politics has been characterized by the alternation of power between the two big mainstream parties: the center-left Spanish Socialist Workers’ Party and the conservative People’s Party. This has also been the case in most Spanish regions. Spain’s imperfect bipartisan system began to collapse in 2015 with the rise of two new parties on the left and center-right, Podemos and Ciudadanos, respectively (Orriols and Cordero 2016). A few years later, in 2019, the emergence of Vox put an end to Spain’s exceptionalism as a country without a radical right party (Turnbull-Dugarte 2019). Broadly speaking, the Autonomous Communities replicate the Spanish party system, with a crucial difference: in the regions, the regional branches of the Spanish statewide parties compete with regionalist parties—that is, non-statewide parties that present candidates in elections (national, regional, local, or European) only within their respective regions and whose main goal is the defense of the interests, distinctiveness, and identity of the Autonomous Community (Alonso, Cabeza, and Gómez 2017). The internal organization of Spanish statewide parties is highly decentralized. Regional branches of statewide parties in Spain have full autonomy and draw up their own regional electoral manifestos, resulting in a high degree of ideological and programmatic divergence within the same party (Alonso, Gómez, and Cabeza 2013). The following parties are included in this article: the regional branches of the traditional Spanish parties—the Spanish Socialist Workers’ Party (PSOE), the People’s Party (PP), and the United Left (IU); the new political parties—Podemos, Ciudadanos, and Vox; and the main regionalist (or non-statewide) parties in each Autonomous Community (see Table 3).

Table 3. Cases in the data set

Source: RMP data set (available at www.regionalmanifestosproject.com).

The RMP’s Novel Indicators: Evidence for the Spanish Case

Is Gender on the Agenda?

How salient are gender issues in parties’ manifestos? Do our data show the high relevance that other scholars have reported in their analyses? In line with the manifesto research, we assume that issues that are emphasized by political parties in their manifestos are largely consistent with the policies they advocate in parliaments and governments.

On average, gender issues make up about 10% of the total number of quasi sentences of the manifestos included in the analysis. This figure is relatively high compared to the level of attention devoted to other issues (Table 4). One of the most salient issues in the manifestos for the 2019 Spanish regional elections was the economy (26.5%). Gender issues (10.1%) were more salient than technology and infrastructures (9.7%), the environment (9.6%), the center-periphery divide (4.2%), or even education (6.8%), one of the most relevant policy areas in the hands of regional governments in Spain. As illustrated in the previous section, the relatively high salience of gender issues should not surprise us if we consider the centrality of gender issues in Spain in recent years.

Table 4. Attention to gender compared to other issues

Note: Average salience scores, percentage over the total number of quasi sentences in the manifesto, N = 76.

Source: RMP data set (available at www.regionalmanifestosproject.com).

Apparently, there are differences in attention to gender issues by region. The parties in the Community of Madrid are the ones that, on average, pay more attention to gender issues (13.6% of the total manifesto), while the average salience of these issues in Cantabria, Castilla–La Mancha, and Asturias is below 9%. However, taking all the parties together, the differences between regions are not statistically significant, as indicated by the results of an analysis of variance (F (12, 63) = 0.959, p = .496).

In fact, the variation does not depend so much on the region as on the party. Figure 1 looks at the average salience score of (positive and negative) gender codes by political party. The new parties (Ciudadanos, Podemos, and Vox) dedicate significantly more space in their programs to gender-related issues than both the traditional statewide parties (PP, PSOE, and IU) and the non-statewide parties (regionalists). The differences between the parties in the attention paid to gender issues are statistically significant, as confirmed by an analysis of variance (F (6, 69) = 5.333, p = .000).

Figure 1. Attention to gender issues. Average salience scores, percentage over the total number of quasi sentences in the manifesto, N = 76. * Non-statewide parties.

Source: RMP data set (available at www.regionalmanifestosproject.com).

This divide between old and new politics suggested by the data is in line with the literature. Previous research has shown that new parties tend to embrace postmaterialist values to a larger degree than traditional parties (Dalton 1991). In Spain, the emergence of Podemos and Ciudadanos in 2015 is mostly linked to the political crisis and the growing disaffection of voters with institutions and traditional parties (Orriols and Cordero 2016; Rodon and Hierro 2016). These new parties have brought more pluralism to the Spanish de facto two-party system and opened “a window of opportunity for gender equality policies” (Alonso and Lombardo 2018, 84). The new radical right party Vox also dedicates more space on its manifestos to talk about gender-related issues than the traditional and the regionalist parties (Figure 1). Yet, it does so in the opposite direction with respect to Podemos and Ciudadanos. Like most radical right populist parties in Europe, Vox’s opposition to feminism is a central element in its political discourse (Alonso and Espinosa-Fajardo 2021; Bernárdez-Rodal, Requeijo Rey, and Franco 2020; Kantola and Lombardo 2019).

Figure 2 shows the position of the different parties on the general gender scale. If Figure 1 clearly reflected the division between the new and old politics in the level of attention paid to gender issues, Figure 2 shows that within each of these two large blocs, the greatest differences in positioning respond to party ideology. Among the old mainstream parties, PSOE and IU score higher than PP on the general gender scale. Similarly, in the new politics bloc, Podemos shows a higher score than Ciudadanos and Vox (Figure 2). In other words, the old and new left are more pro–gender equality than the old and new right. The next section will delve deeper into the effect of ideology.

Figure 2. Position on gender issues. Average position on gender issues, N = 76. * Non-statewide parties.

Source: RMP data set (available at www.regionalmanifestosproject.com).

To sum up, these descriptive findings suggest that gender is clearly on the agenda. In fact, gender issues make up an important part of the electoral pledges of all the parties included in the analysis. Yet not all the parties emphasize all gender issues alike, with the same intensity, and, most importantly, in the same direction. Ideology, together with the divide between old and new politics, seems to play a crucial role in parties’ position on the general gender scale. Next, we test whether the effect of ideology is consistent in all the dimensions of the gender domain.

How Does Ideology Drive Party Positions on Gender Issues?

To investigate the effect of ideology on parties’ gender positions, we employ linear OLS regression. We use robust standard errors, accounting for the clustering of observations at the regional level (parties nested in Autonomous Communities). The dependent variable is party position on gender issues, calculated as shown in Table 3, both in the general scale (Model 1) and in each of the specific policy dimensions of the gender domain: welfare and labor market policies (Model 2), gender violence (Model 3), equal representation (Model 4), and values and gender identity (Model 5).

We test our hypotheses on the effect of ideology by including in the models the variable RILE, which measures party position on the left-right scale (Budge and Laver 1992). Although its validity has not gone uncontested (see Mölder 2016 for a brief overview of criticisms), the “standard” RILE index of the Manifesto Project is the most popular and widely used estimate of parties’ left-right positions (Budge 2013; Volkens et al. 2013). The RILE index of the RMP relies largely on the original scale proposed by the Manifesto Project, with some minor differences in the set of left and right categories.³ The RILE index suits our research purposes for two reasons. First, it performs well in terms of face validity for the Spanish case (Alonso, Gómez, and Cabeza 2013). Second, and most importantly, none of the categories of the gender domain are part of the RILE scale.⁴ The variable RILE ranges from –100 for a manifesto containing only left issues to +100 for a manifesto devoted only to right issues.

As discussed earlier, the left and the right are not homogeneous blocs. Within each ideological bloc, we expect to find differences in positions on gender issues depending on the type of party. As shown in Figures 1 and 2, there are clear differences in gender positions between new, old, and non-statewide parties. Thus, we include a categorical variable denoting the type of party: regional branches of traditional statewide parties (PP, PSOE, and IU), regional branches of new statewide parties (Podemos and Ciudadanos), non-statewide parties (regionalist parties), and the radical right (Vox). We expect new parties, the reference category, to adopt a more pro–gender equality position on gender issues than traditional statewide parties, non-statewide parties, and the radical right. We opted to keep the new radical right party Vox in a separate category. The reason is that we expect opposite effects on the dependent variable: while Vox opposes gender equality policies in some areas, Podemos and Ciudadanos are expected to have a positive effect on the gender scales. In fact, although gender issues receive a similar degree of attention in the manifestos of the radical right and the radical left (Figure 1), Vox and Podemos take highly polarized gender positions on most of the gender dimensions.

We include in the statistical models some control variables that may impact party positions on gender issues. The first control variable is the gender of the party leader. Since Phillips’s (1995) seminal work on the “politics of presence,” a growing body of feminist research has shown that the gender of politicians makes a substantive policy difference: women leaders or MPs enhance the representation of women’s interests (Celis 2006; Espírito-Santo, Freire, and Serra-Silva 2018; Lovenduski and Norris 2003; Phillips 1995; Swers 2020; Wängnerud 2009; Wängnerud and Sundell 2012). According to this literature, women’s attitudes, values, and policy priorities differ from those of men, and therefore, once engaged in political life, they raise issues of particular concern to women. Thus, women’s political presence, especially within the highest ranks of parties, translates into more responsiveness to women’s needs and demands. In fact, it is now well established from a variety of studies that female party leadership impacts party policy agendas in several ways. Female leadership is associated with the inclusion of a broader range of issues on party manifestos (Greene and O’Brien 2016), a stronger emphasis on social justice (Kittilson 2011), party position change on the sociocultural dimension of political competition (Kroeber 2021), and the adoption of gender quotas (Kittilson 2001). Therefore, we account for women’s leadership with a dichotomous variable that takes the value of 1 when there is a female candidate running for regional office and 0 otherwise. Twenty-five of the 76 parties in our sample (32.9%) have a female leader. In all the regions but Castile-León, at least one electoral list for the 2019 regional elections was headed by a woman. In some regions, there were as many as four female candidates for president of the regional government. Podemos and the regionalist parties had higher numbers of regional women leaders.

We also control for the degree of competition in the party system. Previous research has found that multiparty systems are more responsive to gender issues (Kittilson 2011) because parties in these systems are under greater pressure to respond to their competitors’ innovations. We expect that the larger the number of parties in the subnational party system, the higher the parties’ score on the gender scales will be.

Finally, government and opposition parties may differ in the positions they adopt on gender issues. Being in government or in opposition has been found to be a crucial element in the conformation of parties’ political agendas (Green-Pedersen and Mortensen 2010). It not only affects parties’ position on the left-right scale (Greene and O’Brien 2016), but also women’s descriptive representation (Murray 2008; Shair-Rosenfield 2012). Although we do not have any expectation regarding the direction of the coefficients, we include incumbency as a control variable in our models. This variable adopts the value of 1 when the party is in government before the elections and 0 otherwise. Table 6 reports the results of the statistical models. Summary statistics for all variables in the regressions are shown in Table 5.

Table 5. Summary statistics of all variables in the regression models

Source: RMP data set (available at www.regionalmanifestosproject.com).

Table 6. Results of OLS regression models explaining party positions on gender issues

Note: In parenthesis, robust standard errors (adjusted for 13 clusters in region). *** p < .001; ** p < .01; * p < .05.

Source: RMP data set (available at www.regionalmanifestosproject.com).

It is apparent from Model 1 that there is a negative relation between parties’ left-right position and their general position on gender issues (Table 6). An increase in the RILE score leads to a more negative position on women’s issues. That is, the more to the right parties are, the lower their score on the gender scale. At first glance, this seems to support the traditional hypothesis that left-wing parties adopt a more feminist position in their party platforms than right-wing parties. However, a more nuanced picture emerges when we analyze the results of the other models. Interestingly, closer inspection of Table 6 shows that the effect of ideology is not consistent on all the dimensions or policy areas of the gender domain. While ideology clearly influences party positions on gender violence (Model 3), equal representation (Model 4), and values and gender identity (Model 5), it does not affect the position of parties on welfare and labor market policies (Model 2). In contrast with what we found on the other gender dimensions, the coefficient of the RILE scale in Model 2 shows a positive sign and is not statistically significant. This partially confirms the catching-up hypothesis, at least with regard to the welfare state and labor market policies. There are no differences between left-wing and right-wing parties in their position on this specific dimension of the gender scale. In fact, the R ² associated with Model 2 is rather low, indicating that other explanatory variables are needed to better predict parties’ position on gender issues in this dimension. We find the higher R ² in Model 5, where ideology explains a substantial amount of variation on the position of parties on the dimension of values and gender identity.

Figures 3 and 4 graphically show the relation between ideology and position on gender issues in the dimensions of welfare and labor market policies, and cultural values and gender identity, respectively. While Figure 4 indicates that there is a close relation between left-right position and the score on the cultural values scale, the scatter plot in Figure 3 does not reveal a clear relationship between ideology and position on welfare and labor market policies. According to our data, all political parties in Spain regardless of their ideology, adopt a pro–gender equality position on issues related to welfare and the labor market. There is broad consensus among all parties on issues belonging to this dimension, such as the public provision of childcare services, equality in employment or the promotion of equal sharing of responsibilities between women and men.

Figure 3. Ideology and position on gender issues: Welfare and labor market policies.

Source: RMP data set (available at www.regionalmanifestosproject.com).

Figure 4. Ideology and position on gender issues: Cultural values and gender identity.

Source: RMP data set (available at www.regionalmanifestosproject.com).

This finding is consistent with that of earlier qualitative research on the Spanish case, which showed that since the 1990s, the conservative PP has converged toward the PSOE in a more progressive view on gender and family policy issues (León et al. 2019). In fact, when the PP took power for the first time in Spain in 1996, it followed the path of “defamilialization” opened by the PSOE by investing in childcare and the promotion of women’s access to the labor market (Estévez-Abe and Naldini 2016). Interestingly, the radical right Vox is not very different from the other Spanish parties in its position on the welfare and labor market dimension. Vox has negative scores on the scales of protection against gender violence (–1.05) and values and gender identity (–4.48). In fact, the regional manifestos of the Spanish populist radical right contain regressive proposals in these areas such as the abolition of the law on violence against women, the opposition to bills in defense of LGBT rights or the end of subsidies to feminist associations. However, Vox is in favor of gender equality on the welfare and labor market dimension (6.17). Family-work reconciliation policies, public care services, or policies to promote mothers’ labor force participation after childbirth are included among the programmatic proposals of the Spanish radical right. This finding matches those observed by Graff and Korolczuk (2022), who also found that Polish and Hungarian right-wing populist parties promote family policies to attract voters hit by the economic crisis.

Taken together, our results provide support for the framing hypothesis. Left-wing parties score higher than right-wing parties on the dimensions of equal representation, protection against gender violence, and values and gender identity. By contrast, there are no differences between the left and the right in their positions on the welfare and labor market gender dimension. The widespread support for gender equality policies in the labor market or in the field of family care suggests that these policies can be considered in Spain as a “valence issue”: all political parties, independently of their ideology, endorse the same (positive) position.

With regard to the control variables, only the type of party is a significant predictor. As expected, traditional parties, the radical right, and regionalist parties have lower scores on the general gender scale than the new parties. It is interesting to note that although they adopt a pro–gender equality position, the scores of regionalist parties are below the average on three of the four dimensions of the gender domain. Regionalist parties form a very heterogeneous party family, which encompasses parties on the left, center, and right of the political spectrum. However, our data suggests that on average they pay less attention to gender issues than other parties in our data set. A possible explanation might be that regionalist parties devote a significant portion of their manifestos to defend the regional territory’s distinctiveness, which leaves less room for other issues. It may seem trivial, but more attention to one issue necessarily implies less attention to another. Regionalist parties are not single-issue or niche parties but their priority is the center-periphery cleavage (Alonso, Cabeza, and Gómez 2017). According to our results, governing and opposition parties do not adopt different positions toward gender issues. Finally, female leadership⁵ and the number of parties in the subnational party system have no statistically significant effect (Table 6).

Conclusions

This article has made a twofold empirical contribution. First, and foremost, we have presented a new proposal for measuring parties’ policy preferences on gender issues (salience and position) in a systematic way. Our approach is based on quantitative content analysis of election manifestos. By introducing a new domain on gender issues in the coding scheme of the Regional Manifestos Project, we have obtained novel data to analyze parties’ preferences on four fundamental policy dimensions: welfare and the labor market, protection against gender violence, equal representation, and cultural values and gender identity. Second, we have assessed the degree of validity and reliability of the new indicators by analyzing the effect of ideology on parties’ gender positions in decentralized Spain.

Our findings shed new light on how political parties have integrated gender issues into their political agendas. First, in Spain, the left-right divide is persistent on three of the four gender dimensions that we identify. Left parties score higher than right parties on the scales of protection against gender violence, equal representation, and, specially, cultural values and identity (i.e., from positive general references to feminism or the feminist movement to specific policies to guarantee reproductive or LGBT rights). Second, there is general consensus among all parties on the dimension of welfare and labor market policies. Irrespective of their ideology, all parties adopt a pro–gender equality programmatic position on this dimension, which includes childcare, family policies, promotion of women’s entry into the labor market, equal working conditions, and equal distribution of caring responsibilities. Differences between the left, which has traditionally been more likely to embrace feminist claims, and the right have faded. In fact, right-wing parties tend to score higher on the gender scale of welfare and labor market than left-wing parties. Third, ideology is still relevant to explain the framing of gender (in)equality’ by political parties. While the right frames gender inequality as a problem that can be solved through family and welfare policies as well as labor market integration, the left also emphasizes status-based noneconomic programmatic proposals.

Although this article focused on Spain, our findings may well apply to other cases. Center-right parties are the ones that in recent years have embraced significant welfare state reforms and promoted the expansion of work-family policies in many Western European countries, such as Germany, the Netherlands, and the United Kingdom (Morgan 2013). Similarly, recent qualitative research on countries as diverse as Belgium, Austria, Portugal, and Italy has suggested that party attention to work-family issues is widespread across the political spectrum (Weeks 2022).

The most obvious general conclusion that emerges from our findings is that parties’ policy preferences on gender issues are multidimensional: what drives party positions on each dimension is not always the same. Parties’ positions on the dimension of cultural values and gender identity are largely explained by their ideology. However, being a left-wing or a right-wing party does not seem to be determinant for the position of the parties along the welfare and labor market dimension. Other explanatory variables are needed to better predict parties’ positions on gender issues on this dimension. This would be a fruitful area for further work.

Many other questions still remain to be answered. How do contextual factors such as the socioeconomic and political situation of a country impact parties’ positions on gender issues? Under which conditions do political parties include gender issues among their programmatic stances? Does parties’ attention to gender issues in their manifestos translate into more and/or better gender equality policies? The lack of comparative data on parties’ policy preferences on gender issues has limited the type of research questions that the literature can address. We make the new data we have presented here on the Spanish case available to the research community on the project’s website (www.regionalmanifestosproject.com). Only by collecting data from a significant number of cases and countries will we be able to advance in our knowledge of how, when, why, and with what consequences political parties incorporate gender issues into their political agendas.