EU responsibility and responsiveness in candidate countries: hypotheses

Accession to the EU comes with strings attached: candidate countries are expected to internalize EU values and establish them as norms, to adhere to the EU’s legal framework and endorse its policy directions. The Commission issues reports setting goals for the candidate countries, and accession is conditional on ‘responsibility’ in implementing EU guidelines (see e.g. Becker Reference Becker2024; Cremona Reference Cremona2003; Kacarska Reference Kacarska and Uvalić2023; Schimmelfennig Reference Schimmelfennig and In Flockhart2005). In this context, ‘responsible’ government means respecting EU legislation and committing to EU partners. These reports not only assess the continuous progress of countries but can also serve as an accountability mechanism for governments that are ultimately responsible for implementing policies and taking action. Governments that do not want to jeopardize their country’s EU candidacy must adhere to specific norms and procedures in certain key policy areas of EU competence. Given that EC recommendations constitute the EU’s understanding of ‘responsible government’ towards accession, parties may incorporate the Commission’s messages for their country into their own programmes to highlight their commitment to the EU accession process. This way, they can signal to both their electorates and the Commission their EU commitment.

However, political parties also need to respond to the policy preferences of electorates. Responsiveness involves parties being attentive and receptive to the needs and desires of the electorate, particularly in the context of representing citizens in decision-making processes (Karremans and Lefkofridi Reference Karremans and Lefkofridi2020; Lefkofridi and Nezi Reference Lefkofridi and Nezi2020; Mair Reference Mair, Schäfer and Streeck2013). Peter Mair (Reference Mair, Schäfer and Streeck2013, Reference Mair2014) alerted us to the potential tensions between EU demands (responsibility) and domestic preferences (responsiveness), what he called the parties’ RR dilemma. Zoe Lefkofridi and Roula Nezi (Reference Lefkofridi and Nezi2020) refined Mair’s (deterministic) framework to allow for variation in how different parties (within the EU) experience the RR dilemma. Drawing on this work, below we will argue why party behaviour is likely to vary in candidate states. We assume that parties are purposeful actors seeking votes for office and votes for policy: parties aim at increasing their electoral shares to attain/maintain executive power and/or to enhance the advocacy of their party ideology (Bartolini Reference Bartolini1999, Reference Bartolini2000; Müller and Strøm Reference Müller and Strøm1999; Pedersen Reference Pedersen2012; Strøm Reference Strøm1990).

To begin with, strategic parties will consider public opinion; the public’s stance on these issues may facilitate or constrain the parties. Even though gender equality is a fundamental principle in EU legislation that is mainstreamed in EU policies (Spehar Reference Spehar, In Kersten-Pejanić, Rajilić and Voß2012) and conditions the admission of new member states (see also Chiva Reference Chiva, In Abels, Krizsán, MacRae and Van der Vleuten2021), political parties within and outside the EU are constrained by their electorates. Importantly, public opinion on gender equality issues may vary across space, time and issue areas. For instance, gender issues can range from equality between men and women to a broader concept of equality that includes equality of other minoritized groups (e.g. LGBTIQ+ rights). These connect to Gwiazda’s (Reference Gwiazda2023) ‘thin’ and ‘thick’ conceptualization of gender. Both conceptualizations should be visible in the EC progress reports for candidate countries, guiding them in implementing laws and policies in line with the EU. This is because the definition of gender has been transformed at the EU level to include a broader conceptualization (see Spehar Reference Spehar, In Kersten-Pejanić, Rajilić and Voß2012; Sümer Reference Sümer2009). While political parties in the Western Balkan region may have increasingly become more pro-EU (see Stojić Reference Stojić and In Stojić2018), compliance with EU policy and endorsement of EU values, and of ‘thin’ and ‘thick’ gender equality in particular, is a different matter. The attention parties give to specific topics may depend on constraints posed by public opinion. Thus, if and how (much) parties talk about these issues depends upon the degree of congruence between what the EU recommends and what the public wants.

Even if all candidate countries are subject to the same pressures for responsibility (Lefkofridi and Nezi Reference Lefkofridi and Nezi2020) in pursuit of EU accession, the pressures for responsiveness will vary. In the ideal scenario of alignment between the definition of EU responsibility and the preferences of the country’s electorate, parties can send responsibility signals to the EU and respond to their voters.

A problem arises, however, if public opinion opposes the policy direction recommended by the Commission (see also Karremans Reference Karremans2021). In this scenario, parties in candidate countries face the dilemma between signalling a commitment to EU values towards the Commission (responsibility) and responding to what the public wants (responsiveness) (Lefkofridi and Nezi Reference Lefkofridi and Nezi2020; Mair Reference Mair, Schäfer and Streeck2013, Reference Mair2014). As parties compete in elections for citizens’ votes, they cannot ignore public opinion. To minimize electoral costs, parties in candidate countries are likely to refrain from politicizing the issue under the condition of the RR dilemma, which implies low discursive salience. Thus, we formulate our first hypothesis:

All candidate countries are subject to the same pressures for responsibility (Lefkofridi and Nezi Reference Lefkofridi and Nezi2020), in pursuit of EU accession, and all parties within each country are subject to the same pressures for responsiveness to their (country’s) public opinion; however, we argue that the degree of tension faced by individual parties will vary. Previous work has shown that parties often face conflicts between their key goals (Müller and Strøm Reference Müller and Strøm1999): though all parties are trying to increase their electoral shares, some prioritize the attainment/maintenance of executive power, whereas others focus on the advocacy of their party ideology (Bartolini Reference Bartolini1999, Reference Bartolini2000; Müller and Strøm Reference Müller and Strøm1999; Pedersen Reference Pedersen2012; Strøm Reference Strøm1990). We discuss office- and policy-related considerations separately below.

Acknowledging the restrictions of public opinion for all parties within candidate country x, we expect parties in office and parties whose ideology is compatible with EU responsibility to be more inclined to make references to what the EC prescribes. First, we argue that it is important for parties in office to signal responsibility to the Commission. Though the Commission reports should, in principle, motivate all parties of countries aspiring for EU membership towards its policy goals, ‘responsibility’ is particularly expected of parties in government. As responsibility relates to the governing role of parties in fulfilling their country’s obligations towards the EU, incumbents should be, in principle, more eager to send signals of responsibility towards the Commission. Opposition parties are less constrained by pressures to demonstrate alignment with the Commission through rhetorical commitments; in other words, their status offers more leeway to silence/ignore/marginalize the issues raised by the Commission – especially under the conditions of the RR dilemma, given that responsiveness to the public opinion comes with electoral costs. In sum, because EC progress reports matter more for governing parties, incumbents should be more likely to integrate issues related to the Commission’s recommendations in their party programmes – in an effort to show responsibility and commitment to the EU accession process. We thus hypothesize that parties in government will be more likely than opposition parties to mention gender issues in their electoral programmes as signals of a responsible cabinet in pursuit of EU accession. Hence,

Second, the parties’ degree of compatibility with the requested policy (i.e. what the Commission recommends for the country as a whole) is likely to vary (Lefkofridi and Nezi Reference Lefkofridi and Nezi2020), given that EC recommendations in this domain may be (in)congruent with parties’ policy goals. Due to ideological differences between political parties, some party ideologies will be more compatible with EC reports than others. When faced with the RR dilemma, parties whose ideology and policy goals are compatible with EU responsibility in the area of gender equality are more likely to reference gender-related issues than those whose ideology conflicts with what is recommended by the EU. In other words, EC reports are more constraining for some but facilitate others’ policy goals. If the degree of compatibility between party ideology and the EU’s recommendations matters for discursive salience, then the higher the congruence, the higher the likelihood of emphasizing these issues. In contrast, parties whose ideology does not align with EU responsibility are more likely to silence or marginalize the issues in their manifestos.

Political competition concerns party positions on two dimensions: the economic left–right dimension and an increasingly important non-economic, sociocultural left–right dimension pitting libertarian-universalist and traditionalist-communitarian values against each other (Bornschier Reference Bornschier2010). Gender equality issues connect to both, but especially to the latter dimension, which is widely known as GAL–TAN (Hooghe and Marks Reference Hooghe and Marks2009; Jolly et al. Reference Jolly, Bakker, Hooghe, Marks, Polk, Rovny and Vachudova2022; Vachudova Reference Vachudova2021). Research on party positioning within the EU shows that on the sociocultural divide, radical-right or radical-nationalist parties are TAN parties and that green and social-liberal parties, as well as new left parties, are GAL parties (see Chapel Hill Expert Survey classification; Jolly et al. Reference Jolly, Bakker, Hooghe, Marks, Polk, Rovny and Vachudova2022). In detail, GAL (green and left) parties are more likely to voice (positive views on) gender-related issues than TAN (right and nationalist) parties.

Even if public opinion is predominantly against gender-related issues, policy-seeking parties committed to the values promoted by the EU may see the EC recommendations for reforms as an opportunity: they can signal a commitment to their own policy goals and values and EU responsibility to their core supporters (partisans). If this is the case, we should observe that opposition parties, which are less constrained by pressures for responsibility and have more leeway to silence, ignore or marginalize gender-related issues, and will politicize these even under the conditions of the RR dilemma. While this may be interpreted as evidence of a lack of responsiveness to the general public opinion, it would manifest evidence of loyalty towards the party’s values and policy goals – which matter to their core constituencies. As Lefkofridi and Nezi (Reference Lefkofridi and Nezi2020) point out, this behaviour is more likely in proportional systems with multiparty coalition governments; in these systems, the pressure to respond to the average voters is lower than in majoritarian, winner-take-all systems. This leads us to formulate a final expectation:

Case selection, data and methodology

To test our hypotheses about discursive salience, we conduct a comparative study of parties in six Western Balkan countries aspiring for EU membership: Albania, Bosnia and Herzegovina, Kosovo, Montenegro, North Macedonia and Serbia.

Operationalization and data

The policy paradigms of governments and supranational organizations vary over time – and so does public opinion (Hall Reference Hall1993: 279; see also Lefkofridi and Nezi Reference Lefkofridi and Nezi2020). Hence, we need to put empirical content on what the EU understands as ‘responsible’ government; only when this is done can we look at the preferences of public opinion to see whether EU responsibility and the average citizen (in the candidate country under investigation) are in alignment or not.

Pressures for responsibility

To operationally define ‘EU responsibility’ with regard to gender-related issues, we study the so-called Country Reports of the EC on the candidate and potential candidate countries.Footnote ⁶ These reports serve as essential tools on the path towards European integration, providing guidance and encouraging reforms. They also act as a mechanism for monitoring progress in the reform process, highlighting areas that need improvement and providing policy recommendations. We focus specifically on the recommendations on gender-related issues, which we further divide into recommendations focusing more narrowly on the equality between men and women (a ‘thin’ conceptualization of gender) and those referring also to equality of other gender-related marginalized groups, for example, LGBTIQ+ rights (a ‘thick’ conceptualization of gender) (Gwiazda Reference Gwiazda2023). These recommendations serve as the EC benchmark of what is understood as ‘responsible’ on gender-related topics. Although the EC reports are typically published every year, we focus here only on those reports that are published close to a national election and, therefore, close to the election-specific party programmes and thus most suitable for evaluating whether and how (much) parties in candidate countries send signals of responsibility through their party manifestos.

Party programmes may encompass a wide range of issues and, therefore, be less detailed on specific policy areas; nonetheless, they are strategic documents that not only indicate the relevance of specific topics in the overall party agenda but also send signals of responsibility and responsiveness, especially during campaign periods. They are thus useful for understanding agenda-setting and the urgency of some issues compared to others (in the eyes of the EU and public opinion, and their integration in the political agendas during electoral times).

To evaluate whether parties send such signals, and which parties do so, we focus on the salience of gender-related issues in the programmes parties issue in their political competition, which is an indicator of issue politicization; we thus look for references to gender-related issues in party manifestos and analyse the discursive salience of such issues in 77 party programmes from six Western Balkan countries. These programmes were collected mainly by the Manifesto Research on Political Representation (MARPOR) database, which covers all national elections between 2013 and 2021.Footnote ⁷ Notably, the MARPOR collection covers both government and opposition parties across the entire ideological spectrum (see Table A1 in the Appendix in the Supplementary Material for a list of all programmes analysed).Footnote ⁸ The documents were analysed using MaxQDA software for computer-assisted qualitative and mixed methods data.

To analyse the attention parties give to gender issues identified in the EC reports, we code the instances where specific gender issues are mentioned in the documents at the (quasi-)sentence level. We are only interested in discursive salience, so we only count the instances where such topics are mentioned. The coding process resulted in a variety of topics that fall under the thin framework of gender equality, such as women’s employment and education opportunities, political participation and gender-based violence, and some topics that fall under the thick framework, namely support for diverse sexual and gender orientation, support to the LGBTIQ+ community, same-sex marriages and other issues that may be connected to views on LGBTIQ+ such as those pertaining to family or fertility rights (see Table 4 for a comprehensive list of codes).

Table 4. Distribution of General Statements and Policy Areas in the Six Countries (Number of Coded Segments)

Source: Authors’ own data.

Pressures for responsiveness

Testing Hypothesis 1 necessitates data on public opinion to identify whether the conditions for the RR dilemma are present. Since public opinion data on gender-related issues in the Western Balkan countries is sparse, we have to rely on two different sources. First, we use data from multiple years of the Balkan Barometer to better understand public opinion towards equality between men and women (thin conceptualization of gender). The data covers, for example, questions on violence against women, inequality in the labour market and women’s representation in politics. Since Balkan Barometer data is published every year, we select data close in time to national elections and, thus, to party programmes and EC country reports. However, the Balkan Barometer (Regional Cooperation Council 2018) does not include items that are concerned with thick gender-related issues. Therefore, we also use the aggregated data on public opinion towards LGBTIQ+ issues reported in Attitudes towards LGBTIQ+ People in the Western Balkans (Fecanji Reference Fecanji2022) from 2015.

An overview of our data sources is provided in Figure 1. Figure 1 helps to show the sequence of events: when the EC report was issued, when the election took place, and when the public opinion survey we use was fielded.

Source: Authors’ own.

Figure 1. Sequence of Data (Sources and Availability)

Empirics I: responsiveness versus responsibility

To explore Hypothesis 1 (RR dilemma), we examine two sets of issues related to different conceptualizations of gender: On the one hand, a thin concept, referring to equalities between men and women and traditional women’s rights issues; and on the other hand a thick concept, which includes broader categories of inequalities, such as of sexual minorities. We begin by identifying gender-related issues in the EC country reports and examine public opinion on the respective issues to assess if – and how far – EC recommendations might create a responsibility versus responsiveness dilemma for parties.

In the second step, we analyse party programmes to test Hypothesis 1 (RR dilemma) and then look specifically at the differences between the parties in government and in opposition (Hypothesis 2), as well as between GAL versus TAN parties (Hypothesis 3). Using the comparative method, we conduct a stringent test whereby we focus on parties that changed status, from government to opposition or vice versa. By looking at a within-case status change, we can see whether the government/opposition status affects the GAL–TAN party’s attention to gender issues.

‘Thin’ gender-related pressures for responsibility

The EC emphasizes gender equality as a central human rights issue, which is mentioned frequently in all reports across all countries. Most commonly addressed are women’s labour market participation, gender-based violence and women’s political participation, as Table 1 shows. These are all traditional women’s rights issues and refer to a thin conceptualization of gender. While these topics are consistent throughout all country reports, the reports emphasize some topics in specific countries. For instance, gender-based violence and human trafficking are identified as particularly concerning in Bosnia and Herzegovina, as is discrimination in employment regarding maternity rights. On the other hand, public childcare provision is only discussed in the 2016 Albanian country report.

Table 1. Coverage of Gender-Related Topics in EC Reports for Each Country and Year of the Report

Source: Authors’ own analytical output based on data from EC reports (see Table A2 in the Appendix for data used to create this table).

‘Thin’ gender-related pressures for responsiveness to public opinion

The public appears to agree with the issues highlighted in the EC reports, as shown in Table 2. When asked about the three most urgent inequality issues, the majority of respondents in every country mention violence against women, followed – in slightly different orders – by gender stereotypes, the gender pay gap and women’s representation in politics and the economy. In other words, these issues are important not only for the EC but equally for the public.

Table 2. Public Opinion Data on Most Pressing Inequality Issues by Country

Source: Balkan Barometer (2018).

‘Thick’ gender-related pressures for responsiveness

The position of the EC is in clear contrast to the public opinion on LGTBIQ+ rights, as data from the Attitudes towards LGBTIQ+ People in the Western Balkans report (Fecanji Reference Fecanji2022) shows (see Table 3). Although a majority of citizens in every one of the six countries supports LGBTIQ+ rights as an abstract concept, as the agreement with the statement ‘LGBT people deserve to have rights like any other citizen’ shows, more specific issues receive less support. Only 36% of Montenegrins (fully) accept the marriage of sexual minorities; in North Macedonia and Kosovo, only every fifth person (20%). Similarly, the majority of citizens believe that LGBT couples should not have any children: 57% of the population in Kosovo agrees or fully agrees that LGBT couples should not have any children; in Serbia, it is just over two-thirds of the population (67%). Unsurprisingly, the support for Pride parades is also low, with less than a third of the population (fully) supporting their organization. In other words, support for LGBTIQ+ rights remains low across countries in the region – which is in contrast to the position of the EC.

Table 3. Public Opinion Data on LGBTIQ+ Acceptance

Source: Based on data from Fecanji (Reference Fecanji2022).

Empirics II: from government to opposition and vice versa

We conduct a rigorous test of Hypothesis 2 by zooming in on all cases where the same party has moved from opposition to government (or vice versa) to see if governing status matters for levels of discursive salience (Table 5).Footnote ¹⁰ We have, in total, six such cases, for which we place emphasis on substantive policy areas; our government versus opposition hypothesis is tested against data on both thin and thick issues. In all cases, there is an increase in salience when the party moves from opposition to government, regardless of the GAL–TAN positioning. In one case of a party moving from government to opposition (Bosnia and Herzegovina’s Social Democratic Party), we also notice an increase in salience. However, this may also be affected by the party’s GAL positioning. We notice the opposite effect in the case of a TAN party (the Democratic Party in Kosovo), namely a slight decrease in salience as the party moves from government to opposition.

Table 5. Results for Hypothesis 2

Source: Authors’ own data.

In Albania, the Socialist Party’s move from opposition to government is accompanied by increased salience on gender issues and some adjustments in the most important topics. Educational opportunities, political participation, healthcare for women and the valorizing of unpaid work are taking on importance, all in line with the party’s GAL positioning.

In Bosnia, we have two cases of status change. While we found no content in the Party of Democratic Action programme when it was in opposition, the picture changes upon its obtaining governing power: the party addresses gender issues in more policy areas. For the Social Democratic Party in Bosnia, transitioning from government to opposition meant more attention to gender issues for the party. Specifically, we observe an increase from three to 11 codes across three policy areas (with wages for motherhood being the most salient, followed by labour market participation and maternal leave).

In Kosovo, we also have two cases: the Democratic Party shows a slight decrease in salience when moving from government to opposition, but the overall difference is small. Maybe, once a party has had governing power, it carries on with its agenda, also because in the less advanced countries in the accession process, basic gender issues are still unresolved. For the Self-Determination Movement, once governing power is obtained, the salience of gender issues increases, and the party focuses more on topics that are salient both to public opinion and the EC in the thin dimension, like women’s labour market participation and domestic violence.

In the case of Serbia, for the Alliance of Vojvodina Hungarians, we see very little attention to gender issues. Specifically, we coded only the topic of wages for motherhood in 2016 and the labour market participation and healthcare for women in the 2020 programme, as the party was part of the government coalition. When the party was in opposition, we found no gender-related content. Despite the small number of codes, there may be a slight government effect as the party addresses one of the most consistent topics of the EC report, which is also endorsed by public opinion: women’s labour market participation.

Empirics III: the role of ideology and gender-related advocacy

We argued that the RR dilemma (i.e. where EC reports and public opinion diverge) would stifle competition on thick issues of gender equality; Hypothesis 3 posited that under these strained conditions, attention to gender-related issues would be most likely among GAL parties, as opposed to TAN parties. A rigorous test of Hypothesis 3 concentrates only on the thick gender-related issues, where parties face a tension between responding to the EC or the broader electorate. We are particularly interested in whether GAL parties will refer to LGBTIQ+ issues despite the RR dilemma.

Focusing on LGBTIQ+ issues, we identify which parties express support for LGBTIQ+ rights (in accordance with the reports) and which are against them (in accordance with public opinion). Table 6 shows all the LGBTIQ+ statements in our sample of documents. Very few parties explicitly focus on issues concerning the LBTIQ+ community; the few that do are mainly opposition parties (against Hypothesis 2). In detail, we found related issues only in the programmes of eight parties, either centrist or left-leaning parties (partially in support of Hypothesis 3). The case of North Macedonia is worth mentioning: the majority of statements come from a GAL party (a green party in a coalition government) but it is also the only case where an opposition party on the right (the centre-right Alliance of Albanians) is openly against same-sex marriages (a clear TAN message).Footnote ¹¹ We interpret this as some first evidence of the politicization of LGBTIQ+ (both GAL and TAN refer to them and have competing positions).

Table 6. Party Statements Mentioning Specifically the LGBTIQ+ Community and Issues

Source: Authors’ own data and authors’ translations.

Discussion and conclusion

A recent report about women’s rights in the Western Balkans by the Kvinna till Kvinna Foundation found that ‘rather than progressing, however slowly, indicators show that women’s rights are indeed regressing’; the evidence ranges from ‘declining access to social and economic rights’ to ‘increased violence and threats against women’s human rights defenders’ and ‘rising femicide rate’ (Danaj et al. Reference Danaj, Čolak, Dičić Kostić and Vrbaški2023: 7). This is troubling, given that gender equality is a condition in countries’ paths to EU accession. Gender equality is not just part of the EU acquis but a whole set of values and principles of the Copenhagen criteria, based on the charts of fundamental human rights and freedoms. Beyond the domain of legislation, gender equality concerns a deep and complex change in attitudes that is difficult to achieve even in many EU member states.

Against the background of new gender lines of conflict in Europe and alongside the ever-increasing gender mainstreaming efforts into EU social policy, this study focused on party attention to gender-related issues in Western Balkan countries aspiring for EU membership. We limit our analysis to the Western Balkan region’s third decade into liberal democracy and market economy; our selection of country cases sample manifests relatively similar levels of development, institutionalization and political maturity, although some considerable differences remain between them. We expected that pressures to align with the EU agenda could lead to higher levels of attention to gender issues, along the lines of government and opposition or GAL–TAN orientation. We relied on EC country-specific reports published regularly, serving as a definition of EU responsibility on gender-related issues – and a benchmark for the alignment of candidate countries with the acquis. Parties in candidate countries need to show commitment to the country’s EU trajectory: the endorsement of EU values, legislation and policies includes gender equality, so making references to gender equality can send positive signals to the EU. Since political parties have to behave responsibly towards EU membership requirements but are also mindful of citizens’ attitudes, we argued that they might be faced with a responsibility versus responsiveness dilemma (Mair Reference Mair, Schäfer and Streeck2013, Reference Mair2014) and tested specific hypotheses on how discursive salience would depend upon the presence of the RR dilemma (Hypothesis 1), but also on parties’ governing status and their ideology. As reports are typically aimed at governments, we anticipated differences between parties in government and those in opposition, with governing parties making an effort to appear more responsible towards the EU by aligning their gender equality agendas with the Commission’s recommendations (Hypothesis 2). We also hypothesized that TAN parties would be less likely to refer to gender-related issues than GAL parties (Hypothesis 3).

The broader message of the present study is that gender-related issues are generally not politicized in this region; when they are, they are usually of the thin type, namely focusing on women’s rights. This is where public opinion and the EC agree, and parties are not faced with an RR dilemma. Even in the ideal world, where EC reports and public opinion align, most political parties in the Western Balkans do not show substantial commitment to gender equality agendas beyond promises for the future and vague rhetorical commitment. This refers mainly to a thin conceptualization of gender issues that focuses more on advancing women’s rights. As expected, government status matters (Hypothesis 2).

In contrast, thick gender-related issues, where public opinion and the EC are not aligned, are brought into the political debate by some centrist parties, which offer somewhat general support to gender identities, and, more forcefully, by GAL parties, which make more concrete references to LGBTIQ+ issues (partial support for Hypothesis 3). These findings from the Western Balkans echo research on party positioning from within the EU that indicates that, in terms of the sociocultural divide, radical-right and radical-nationalist parties align with TAN positions whereas green and social-liberal parties, as well as new left parties, fall under the GAL category, as classified by Chapel Hill Expert Survey (Jolly et al. Reference Jolly, Bakker, Hooghe, Marks, Polk, Rovny and Vachudova2022). While TAN parties value order, tradition, authority and the nation, GAL parties value the expansion of personal freedoms (divorce, abortion, same-sex marriage) and cherish cosmopolitanism and liberal policies.

To be sure, party programmes offer a snapshot of the party’s political agenda and may be less telling of more elaborate positions across policy areas. These party documents may, for instance, only tell us very little about the values parties hold about women and LGBTIQ+ people and remain silent as to how parties respond to specific events (e.g. gay pride parades) or legislative proposals related to women’s rights. Nonetheless, manifestos are useful for capturing which issues are prioritized in the parties’ official electoral messages and how much space they devote to different issues across policy areas. Crucially, they are appropriate for comparative analyses as they are comparable across countries, parties and within countries over time. Given our expectation that gender issues will be further politicized in the course of accession processes, we hope that future research will revisit our findings with empirical analyses of different types of data (e.g. parliamentary debates, media discourses or parties’ communication channels) and party responses related to specific events, legislative proposals or contentious issues, which relate to gender issues.

As countries progress in their accession process, they are expected to expand their references to gender equality, from women’s issues (thin) towards thicker diversity agendas (embracing LGBTIQ+ rights). Our analysis shows that very few parties do so and bears implications for understanding the interplay between EU demands and domestic politics. Due to the polarizing potential of gender-related issues, especially in societies where patriarchal norms persist to some extent, our analysis of EC reports, public opinion and party programmes according to the thick and thin conceptualization implies that EC pressure for more progressive agendas may cause tensions between EU responsibility and responsiveness. The politicization of gender-related issues may, in turn, activate political competition along ideological lines on these issues, as well as along pro/anti-gender lines in the years to come. Certainly, political will to comply with the EU’s recommendations in the area of gender equality differs significantly depending on how gender is understood. This has repercussions for the EC’s assessment of these countries’ future in the area of gender equality. With regard to this fundamental value of the EU, change in response to EU conditionality and the (slow) progress on the EU accession path is reflected neither in our findings nor generally in human rights trends that are negative in the region as a whole.