Intersectionality Impact Assessment : the European Union 2009-2019 : what are the implications for the quality of intersectionality in EU gender-equality policies?

University of Amsterdam Amsterdam, The Netherlands Faculty of Political Science

M.Sc., Political Science, European Politics and External Relations

‘Intersectionality Impact Assessment:

The European Union 2009-2019’

What are the implications for the quality of intersectionality in EU

gender-equality policies?

Masters Thesis

Author: Mylo Houghton (Student no. 11261803)

Academic Year 2018-19

Supervisor: Afsoun Afsahi

2

_Reader:

_{Franca van Hooren}

Table of Contents

1. Introduction

a. Structure of the Thesis

b. Problem Setting and Relevance

i. Anti-Discrimination Policy in the EU

ii. Gender Mainstreaming and Intersectionality in the EU

c. Research Question

2. Literature Review

a. EU Legal Framework on Anti-Discrimination

b. Framing Intersectionality

c. Gender Mainstreaming

d. Institutionalising Intersectionality in the EU

3. Theoretical Framework

a. Conceptualising Intersectionality

b. Mutually Shaped Inequalities: A Critique of Hancock’s Intersectional

Approach

4. Methodology and Data Collection

i. Critical Frame Analysis: Conceptual Basis

ii. Critical Frame Analysis: Operationalisation

b. Methodological Limitations

5. Analysis: Trends in the Framing of Intersectionality in EU Gender

Equality Policies

a. Explicitness, Visibility, and Inclusiveness of Intersecting Categories

i. Gender-Balance in Decision-Making Positions (DM)

ii. Nonemployment (NE)

iii. Intimate Citizenship (IC)

iv. Gender-Based Violence (GBV)

b. Articulation of Intersectionality

c. The (De)Gendering of Policies

e. Lack of Stigmatisation and Challenging Privileged Groups

6. Conclusion

a. Implications of the Trends in EU Gender-Equality Policies

7. Appendices

8. Bibliography

Introduction

Structure of the Thesis

The structure of this thesis is laid out in the following way: First, the introduction is designed to provide an overview of the issues discussed in this thesis. This background focuses both the social and academic relevance of the thesis. Herein the research question is defined. Second, a literature review is provided in order to give context to the subject of the thesis. Third, the theoretical framework is developed to explore how this research adds to the theoretical debate and how new theoretical concepts can be applied in this thesis. The fourth section details the methodology utilised in this research, how it was constructed, adapted and operationalised, as well as potential

limitations to this work. The fifth section presents the findings of the research and analyses the main trends found across the texts in the sampled documents. Finally, these trends are discussed and concluding remarks plus reflections are then given.

Problem Setting and Subject Relevance:

Anti-Discrimination Policy in the EU

Since its inception the EU has placed gender equality on its political agenda. Yet the establishment of a broader anti-discrimination policy, set to combat discrimination on the basis of racial or ethnic origin, disability, religion, sex, age or sexual orientation, only came into being over the last two decades. Policy measures and legal instruments introduced included the 1997

Amsterdam Treaty, the 2000 Charter of Fundamental Rights, two Directives, and a Treaty of Lisbon article, as well as other non-binding measures (Lombardo & Verloo, 2009a). These changes to the EU equality policy architecture have transformed the approach from addressing gender equality, primarily in the field of labour and employment, to addressing multiple different inequalities.

While approaches to tackling inequality at the EU level have broadened, for the most part they have treated multiple inequalities as separate issues (Lombardo & Verloo, 2009a; 2009b). Such an approach ignores the reality that different types of inequalities often interact with one another, including intersections between gender, the inequalities listed above, and many more.

Gender Mainstreaming and Intersectionality in the EU

Verloo (2007, 23) attempts to “map the diversity of meanings of gender equality across Europe and reflects on the contested concept of gender equality” (Verloo et al. 2007, 23). One of the arguments raised is that the character of ‘gender equality’ is obscured by a tendency to

“homogenise diversity under a dominant norm”, as the EU is accused of doing, or by an explicit “strategic framing” (ibid.) of the concept to allow it easier entry into the political ‘mainstream’ as a common and accepted goal. Verloo et al. (2007) offer several contributions to the study of gender equality in the EU, however the focus in this paper will rest on the two methods of characterisation mentioned above, which will be elaborated upon further.

Official gender mainstreaming strategies can be traced back to the United Nations 1995 Beijing conference, and were adopted soon after by EU institutions (Verloo, 2005). The Council of Europe defines gender mainstreaming as “the (re)organisation, improvement, development and evaluation of policy processes, so that a gender equality perspective is incorporated in all policies at all levels and at all stages, by the actors normally involved in policy-making.” (COE, 1998). Feminist scholars (Shaw, 2005; Squires, 2005; Walby, 2005) have highlighted the transformative potential of gender mainstreaming by challenging existing policymaking structures and setting new gender equality goals across all policy domains. Mainstreaming as a tool to promote gender equality has found some success in the EU, acknowledged by European women’s movements and scholars (Shaw, 2004) despite its goals still being far from realised roughly two decades later (Lombardo & Meier, 2006; Verloo, 2005).

Following the increased use of gender mainstreaming as an important social policy tool by EU institutions, those campaigning to address various other inequality issues also hoped to see their issues addressed across all policy domains (Woodward, 2008). The ‘European Year of Equal

Opportunities for All’ in 2007 highlighted how many equality issues and their advocating groups fought for political attention and economic resources, an issue that continues to weaken EU institutions, civil society organisations, and the rights of marginalised individuals (Lombardo & Verloo, 2009). Woodward (2008) suggests that “in some ways gender mainstreaming has been a victim of its own success”.

Current political and theoretical debates regarding EU equality / anti-discrimination policymaking have begun to recognise the need to address inequalities together in order to avoid such competition. Verloo (2007, 287) argues for “the need to overcome a simple bipolar logic of analysis that treats one type of inequality as compared to another, taking what appears as the dominant one as the norm for comparison, instead of focusing on the point at which the various inequalities of race, gender, class, etc., intersect with each other”. The debate over how to understand and tackle issues of inequality in the EU therein builds on theories of intersectionality from Crenshaw (1989, 1991) and other authors, discussed in the forthcoming chapters.

This thesis attempts to analyse the conceptualisations of inequalities in the EU. More specifically, the analysis focuses on the intersections between inequalities and how these intersections, where present, are framed in EU equality documents from the last decade. The specific area of study is EU gender-equality policy documents. This leads to the formulation of the following research question:

Research Question:

What are the implications of the framing of intersectionality for the quality of

gender-equality policies in the EU?

Literature Review

This literature review will lay out some of the major contributions to the study of intersectionality in the EU. To begin with, the review outlines the EU legal and policymaking frameworks, plus their machinery, developed to tackle anti-discrimination. Major contributions to the issue of how political intersectionality is theorised in the context of the EU anti-discrimination approach are then mapped. Following this, research on the framing of intersectionality in the European policy discourse will be highlighted. Further literature on Gender Mainstreaming and inequalities is then presented, connected to structural issues concerning policymaking and EU institutions.

The purpose of this review is to provide an overview of how inequalities and

intersectionality have previously been treated within the political institutions of the EU, and how the scholarly understanding of these issues has evolved. This review is designed to synthesise the scholarly works mentioned, so as to provide background to the issues covered. The review also provides part of the qualification for the contributions made in this thesis to the subject area.

EU Legal Framework on Anti-Discrimination

Studies of the EU legal framework on anti-discrimination have been carried out in depth by scholars such as Bell (2002; 2000; 1999) and Shaw (2005, 255; 2004), among others (see Vega, Lombardo & Rolandsen Agustin, 2008, for more on this area). Some of the work conducted by Bell (2002; 2000; 1999) includes: analysis of the anti-discrimination legal framework found in Article 13; Directive 2000/43/EC on race equality; and Directive 2000/78/EC on discrimination and

employment. Bell’s analysis is particularly relevant for the purposes of this research since he argues that the EU legal framework covering equality issues has developed asymmetrically. For example, in

Bell (2002) he claims that the comprehensiveness of cover for different equalities varies

considerably. Across his work, Bell remarks on the ‘hierarchy of equality’, comparing gender to other ‘strands’ of inequality in EU policymaking, where gender is privileged in its level of protection and the capacity of policy instruments built to tackle gender equality issues (Vega, Lombardo & Rolandsen Agustin, 2008). The work of Shaw (2005; 2004) follows that of Bell (2002; 2000; 1999) chronologically and thematically. Shaw investigates the different legal protection the EU has given to different strands of inequalities, including an analysis of EU concepts of non-discrimination,

mainstreaming and equality. Shaw also addresses the use of political strategies by the EU that mainstream equality to include not only gender. In Shaw (2005; 2004) her analysis uncovers how the EU had only begun to apply mainstreaming strategies to inequalities other than gender, in areas such as race and ethnicity, or to some extent disability and age. Shaw concludes however that these policies have been limited in their scope and success, emphasising a disjointed and unclear approach from the EU to addressing the issues of various inequalities.

Vega, Lombardo and Rolandsen Agustin (2008, 8) highlight that to better understand the ways in which inequalities are treated at the EU level, in particular by the European Commission, contributions to the field by Verloo (2006) must be acknowledged. In her analysis of the EC 2004 Green Paper ‘Equality and Non-discrimination in an Enlarged European Union’, Verloo criticises what she views as an overly simplistic approach. She contends that “…the Commission has adopted a ‘one size fits all’ approach that assumes an unquestioned similarity of the social categories connected to inequalities and of the mechanisms and processes that constitute them.” (Verloo, 2006; in Vega, Lombardo & Rolandsen Agustin, 2008, 8). Verloo seeks out the potential applications of structural and political intersectionality by comparing inequalities such as class, race and ethnicity, sexual orientation and gender. Ultimately, Verloo argues that the ‘one size fits all’ approach ignores the differentiated character and dynamics of inequalities, and their unique effects (Verloo, 2006). Her article is considered relevant to this review for several reasons: it offers critical reflections on the EC’s anti-discrimination approach; it highlights the importance of institutional mechanisms and the

context in which inequalities are treated and reproduced; and it questions the terms for the treatment of intersectionality and multiple inequalities in these domains, be they complex or simplistic.

Framing Intersectionality

The European MAGEEQ project (mageeq.net) was developed by Mieke Verloo (2007) and other scholars to research the framing of intersectionality in EU policy discourse. Verloo (2007) explores and compares the different meanings of gender equality in the EU and several member states. Issues raised in the EU discourse include family policies, domestic violence, and gender inequality in

politics, with other issues specific to national contexts. The findings of the MAGEEQ project highlighted the absence of political intersectionality in European gender equality policies, also revealing biases in the treatment of different inequalities. The research emphasised that where any racist, ethnic, or homophobic biases might occur during the creation of equality policies, the mainstreaming of intersectionality in official gender policy could be hindered. Overall the results suggested that the (absence of) framing of intersectionality in gender equality policies was currently a barrier to stretching gender equality to address multiple forms of inequality (Lombardo & Verloo, 2009). Their work described the inclusion of political intersectionality in European policy making as at an “embryonic stage” (ibid.). More can be found on the debate on how gender equality and intersectionality have been conceptualised in theory and practice in the EU in Verloo (2007). Contributions from Verloo’s (2007) development of a Critical Frame Analysis approach to studying intersectionality in the policy discourse will also be discussed in Chapter Three.

Gender Mainstreaming

An overview of the adoption and application of ‘Gender Mainstreaming’ strategies within the EU is included here to better understand some of the specific tools developed to tackle inequalities, and a

review of their success so far. Verloo (2005) and Walby (2005) both address the subject of Gender Mainstreaming as a strategy to combat gender inequality at a structural level, as problematised for EU institutions above. Verloo (2005) signals the potential uses of gender mainstreaming strategies to identify gender biases in current policies and institutional mechanisms that reproduce such

inequalities. She asserts that gender mainstreaming strategies are aimed at “…reorganising policy processes so that the regular policy makers will be obliged and capable to incorporate a perspective of gender equality in their policies…” and that “…this strategy aims at a fundamental transformation, by eliminating gender biases, and redirecting policies so that they can contribute towards the goal of gender equality.” (Verloo, 2005).

Verloo’s research identifies several issues in the implementation of gender mainstreaming across EU countries, due to differences in the definitions of gender equality. She also uncovers a tendency to depoliticise the issue, where gender mainstreaming is framed as a technocratic policy problem that blocks the active involvement of citizens and civil society organisations. Her work also shows how initiatives are very rarely developed at the national policymaking level in some EU member states.

Walby (2005) assesses how the situation of gender mainstreaming in the context of other

intersecting inequalities has also led to neglect by the overly simplistic approach to the treatment of inequalities by EU institutions. She states that “the theorisation of these (intersecting inequalities) and the practical recognition of such intersectionality is a current major concern.” (Walby, 2005). Walby notes that under the processes of mainstreaming, gender must be combined with other major agendas. The question remains whether gender ought to be embedded and thereby

dispersed, or abstracted and analytically separated, so that it remains clearly visible. This has been raised as a policy and theoretical issue. Walby asks whether there can be a treatment of equality that projects sameness in some domains, equal valuation of difference in other domains, plus new hybrid standards in others. Walby (2005) and Woodward (2008) both provide mixed accounts of the

impact and successes of Gender Mainstreaming by and within the EU, stressing that there is still much to be achieved regarding an intersectional approach to the treatment of inequalities.

Institutionalising Intersectionality in the EU

Expanding on the literature on the EU legal framework on intersectionality, and on structural institutional analysis of the EU, Lombardo and Verloo (2009) investigate the extent to which intersectionality has been institutionalised in the EU. Their study takes an in depth look at the legal and political architecture organised around intersectionality. They analyse the institutional

structures that have been developed to tackle different strands of inequalities in the EU, across different domains. Lombardo and Verloo’s findings are critical to this review since their research acknowledges that both the European Commission and European Parliament have begun to show interest in concepts of ‘multiple discrimination’ (Lombardo & Verloo, 2009). However, their analysis concludes by stating that an intersectional approach to the treatment of inequalities is not in

evidence in the political practices of EU institutions. In contrast, recognition is given to European civil society organisations for their adoption and promotion of an intersectional approach.

Theoretical Framework

Conceptualising Intersectionality

In this section I will trace the development of theories of intersectionality until the QUING 2011 report, highlighting how the understanding of how to conceptualise intersectional relations has transformed and can also be applied in this thesis. Ultimately the conceptualisation used in this thesis applies an understanding of the relations between inequalities as being mutually shaped. This theoretical contribution is taken from the work of Walby, Armstrong and Strid (2009) in Verloo et al. (2011, 80).

The concept of intersectionality has become increasingly used across several academic disciplines over the last two decades, including gender studies, sociology and economics (see Hill Collins 1998; Brah 2002; Brewer et al. 2002; Beisel & Kay 2004; Brah & Phoenix 2004; Risman 2004; Wekker 2004; Belkhir 2005; Yuval-Davis et al. 2005; Phoenix & Pattynama 2006; Weldon 2006; Hancock 2007; Walby 2007; Davis 2008). The conceptualisation of intersectionality is largely credited to Crenshaw, who introduced intersectionality to “denote the various ways in which race and gender interact to shape the multiple dimensions of Black women’s employment experiences” (Crenshaw, 1989, 139). Crenshaw proposed further distinguishing between ‘structural intersectionality’ and ‘political intersectionality’ (Crenshaw, 1991, 1251).

For Crenshaw, ‘structural intersectionality’ applies to inequalities and their intersections as they relate directly to the experiences of individuals in society (ibid.). According to the theory behind structural intersectionality, inequalities often amplify the effects of one another, they are mutually constitutive (Walby, 2007). Crenshaw defines ‘political intersectionality’ as a way to identify how inequalities and the intersections between them are relevant to political strategies. An

understanding of political intersectionality allows us to identify the interdependencies between intersecting inequalities, since the methods of combatting one inequality rarely have a neutral effect

on other inequalities (Crenshaw, 1991). The concept of political intersectionality highlights the fact that “…feminism can marginalise ethnic minorities, or disabled women, that measures on sexual equality or on racism can marginalise women, and that gender equality policies can marginalise lesbians.” (Lombardo & Verloo, 2009a). In many cases, failure to address political intersectionality inhibits the goal of gender equality.

From a theoretical perspective, Crenshaw’s concept of political intersectionality is particularly relevant for the study of intersectionality in the EU (Crenshaw 1989, 139). Addressing the interdependencies between different intersecting inequalities helps us to understand the limitations of the EU equality policies, and the dynamics of competition and alliances within the EU institutions and civil society (Vega, Lombardo & Rolandsen Agustin, 2008). Crenshaw uses examples concerning the US context, such as domestic violence activists blocking access to information that might reinforce racial stereotypes, but this understanding of intersecting inequalities also offers insight into how political strategies by EU institutions and civil society organisations can benefit some social groups while marginalising others (ibid.).

It is possible that EU gender equality policies can marginalise women from racial or ethnic minorities, based on sexual orientation, age or ability. Similarly, policies designed to tackle racial or other types of discrimination can marginalise women. Vega, Lombardo and Rolandsen Agustin (2008, 7) argue that any analysis of how intersectionality is applied in practice by the EU must reflect on the use of political intersectionality. Such reflections can be accomplished with analysis of policies that might privilege certain inequalities and ignore others that are mutually constitutive, as Crenshaw suggests. Ultimately, such policies reproduce power structures and imbalances among marginalised groups. In order to tackle such an issue holistically, a greater emphasis must also be placed on investigating institutional factors that contribute to such marginalisation (ibid.)

Building on Crenshaw’s concepts, developmental tools to understand how to apply political intersectionality are necessary for researchers and policymakers alike. Hancock (2007) highlights the

differences between ‘unitary’, ‘multiple’ and ‘intersectional’ approaches to studying inequality. The unitary approach addresses individual inequalities (e.g. gender or race only) as the most explanatory or dominant factor. A multiple approach targets more than one inequality at a time and treats them as equally situated, as part of a predetermined relationship with each other. For Hancock, an intersectional approach addresses more than one inequality and how they intersect in their relationship(s), treating such a dynamic as an open empirical question for study (Hancock, 2007). Categories of inequality are seen as intersecting based on “a dynamic interaction between individual and institutional factors”, requiring a more holistic approach to research. Such an approach aims to identify the many unique ways in which complex inequalities might interact, inequalities such as gender, race, class and many more, as well as the impact that these interactions have on individuals, as part of society. More detail on the relationships between inequalities and how they can be conceptualised can be found in the works of Crenshaw (1991, 1997), McCall (2005) and Hancock (2007) among others.

This thesis aims to build on Hancock’s conceptualisation of an ‘intersectional approach’, with some insight from Walby, Armstrong and Strid (2009, 263). In essence, the aim is to view inequalities as part of a dynamic, intersecting relationship, rather than individually or as additive. As mentioned previously, Hancock (2007) advocates for an open empirical analysis of what an intersectional approach should look like, which will be employed in this study. Hancock’s interpretation is highly applicable to the EU case because of the understanding it allows us of the political strategies within EU institutions and competitive group advocacy, as mentioned before. The unitary approach not only treats categories as separately relevant for political explanation, but also assumes that an individual’s membership to each inequality is permanent (Hancock, 2007). The multiple approach takes the separate categories and makes them additive. Ultimately this leads to what Hancock describes as the “Oppression Olympics” (ibid.) which forces groups to compete for political attention and resources. Hancock’s interpretation reinforces the findings of studies on EU level institutional and civil society dynamics, especially in the policy areas of gender equality and anti-discrimination

(see Lombardo & Verloo, 2009; Verloo 2006; Walby 2007). Moreover, Hancock’s interpretation provides a clear path to understanding one of the roots of gender inequality: “the genderedness of systems, procedures and organisations” (Verloo, 2005).

Verloo (2006, 212; 2007, 33; 2011, 80) develops a critical frame analysis approach, designed as a framework to employ open empirical analysis of EU gender equality policy, which draws from Bacchi’s (1999) ‘what’s the problem? represented to be approach’. More information on this framework and its employment in this thesis can be found in the subsequent chapter. Verloo’s research has contributed to the wider QUING (2011) report, which also details some of the relevant weaknesses in Hancock’s intersectional approach, outlined below.

Mutually Shaped Inequalities: A Critique of Hancock’s Intersectional Approach

Walby, Armstrong and Strid (2009, 263) aim to move beyond Hancock’s typology of understanding the relationship between inequalities as unitary, multiple or intersectional. As Hancock argues, the unitary approach examines a single, stable category of inequality, presumed to be primary. The multiple approach addresses more than one category and each of them matter equally; the categories are considered to be stable and have stable relationships with one another. Finally, the intersectional approach also addresses more than one category, the categories matter equally, the relationships between them are open, and the categories themselves are dynamic not stable (Hancock, 2007).

Walby, Armstrong and Strid (2009) agree with Hancock that “different categories related to inequalities should only be temporary stabilised for analysis, while still keeping in mind their flued and dynamic character…”. The criticism levelled at Hancock’s approach is that it considers a category to be either dominant (unitary) or equal to other categories (multiple / intersectional), which ignores the possibility that asymmetrical relationships could exist (Walby, Armstrong & Strid, 2009). The

authors disagree on the symmetry of relations and on the “degree of determination of the interwoven inequalities” (ibid.)

On the symmetry of relations, the authors argue that it is important to treat the issue as one for investigation, instead of making an a priori assumption that all strands are equally important. They consider the possibility of asymmetrical relations between strands as important for empirical analysis, since it may show the importance of one or another related category in the process of policymaking (ibid.). Regarding the “degree of determination of the interwoven inequalities” (Walby, Armstrong & Strid, 2009), they claim that the notion that the relationship between inequality

strands is always mutually constitutive is too limiting.

Walby, Armstrong and Strid (2009, 263) assert that the relationship between inequalities is better understood whereby the inequalities involved mutually shape one another, which allows us to identify their effects on each other while also providing analytic distinctiveness. Such an approach considers both the ontology of, and the relationship between, different categories or ‘strands’ of inequality. The conclusions in Walby, Armstrong and Strid (2009) are complemented by Lombardo and Rolandsen Agustin (2011, 4), who state that: “a research focus on ‘intersectionality-only’ might obscure the analysis of the distinctive ontology of each inequality, which is crucial to develop good equality policies.”. An understanding of intersecting inequalities mutually shaping one another is adopted in this thesis accordingly.

Ultimately, the aforementioned authors argue that focusing on relationships in the

policymaking process “prevents the powerful from fading from view” (Walby, Armstrong and Strid, 2011, 80). For analytical purposes, awareness of the ontological depth of inequalities can highlight the presence or absence of different categories, axes or strands of inequality, across various political frameworks or within policies. The question of “which inequality is the most important” (ibid.) must be treated as an empirical rather than a normative question. Each of the points raised here add to

the theoretical and practical debates on how to study, and further the goals of, gender equality and anti-discrimination policy at the EU level.

Methodology and Data Collection

I do not conduct my research with strong predetermined ideas of how an intersectional approach should look in practice, treating intersectionality as an open empirical question as Hancock prescribes (Hancock 2007, 64). My goal is to assess how policymakers and civil society actors

articulate an intersectional approach, in an EU context, through the analysis of policy documents.

As mentioned before, my research borrows concepts from the QUING Project (Quality in Gender + Equality Policies), which was designed for the purpose of providing knowledge for inclusive gender and equality policies in the EU (see www.quing.eu ). The QUING project also built on the methodology developed within 2007 MAGEEQ Project, which was itself designed to construct a conceptual framework to map the various dimensions of EU gender equality policy frames (see

www.mageeq.net )

Frame analysis will be used in this thesis to identify existing intersectional trends in EU policies. The criteria outlined further in this section can help to assess the quality of such trends by pointing at “what, who, and how policymakers are privileging or neglecting in their policy design” (Lombardo & Rolandsen Agustin 2011, 3)

My focus is on the content of the policies, as opposed to their implementation. I argue that this is sufficient in so much that the identification of intersectionality in the policy documents is a necessary early step in the process of redesigning the EU’s approach to tackling inequalities (Verloo 2007, 21; Lombardo & Verloo 2009, 489). In my interpretation of the data, I will remark on the status

of the EU’s approach to gender equality policy, intersectionality, and inequality policies more broadly.

Critical Frame Analysis: Conceptual Basis

Critical frame analysis is a suitable approach for the study of the diversity of

conceptualisations of gender equality and intersectionality in policy practice in the EU, since the purpose of this methodology is to map an understanding of the ways in which an issue is framed (Verloo, 2007)

As highlighted within the theoretical framework, gender equality and intersectionality are both contested concepts, with dynamic, context specific meanings (Lombardo et al. 2009) Both concepts are open to interpretation and contestation by different actors, across various institutional spaces. As such, Dombos (2011) asserts that the study of such concepts, and such processes of interpretation and contestation, necessitate a discursive approach to politics and policy. One of the methodological tools frequently applied in such approaches is policy frame analysis.

Within the highly contested field of frame analysis, this research draws on the QUING interpretation, based on an extension of Verloo’s definition of a policy frame.

According to Verloo (2005) a policy frame is defined as “an organising principle that transforms fragmentary or incidental information into a structured and meaningful problem, in which a solution is implicitly or explicitly included”. The QUING methodology elaborates on this definition with the concepts of ‘strategic framing’, which centres on the intentional use of frames to achieve an actor’s goals, as well as ‘frame resonance’, the question of which frames are available and can be utilised in a particular context (Snow & Benford, 1988) The latter concept, ‘frame resonance’, is tied to the environments in which actors operate, pertaining to the notion of “deep cultural meanings” that Bacchi (2009) prescribes. From this perspective, when one studies frames in particular institutional settings in relation to particular social issues or policy fields, cognitive

processes such as perception, labelling and structuring reality tend to dominate our understanding (Dombos, 2011).

Critical Frame Analysis: Operationalisation

I will analyse documents by institutional and civil society actors on four policy issues, within which several sub-issues are selected. The document analysed are taken from the period 2009-2019, following from the end period of the 2011 QUING research

The issues covered include:

• Gender-based violence (GBV) which concerns any form of violence rooted in gender inequalities (domestic violence, sexual harassment, trafficking)

• Intimate citizenship (IC) which addresses policies that regulate intimate relations (divorce, marriage, and separation; sexual orientation discrimination; reproduction rights)

• Nonemployment (NE) which considers employment and other related policies (tax and benefit policies; care-work; reconciliation of family and work; gender pay gap and equal treatment) through an inverted perspective on how these policies construct subjects who are considered as legitimately employed or non-employed and what the gender implications of this construction are (Dombos, Krizsan, & Zentai 2008).

I intend to analyse not only the three issues discussed by Lombardo and Rolandsen Agustin (2011, 5), but an additional issue: Gender balance in political decision-making positions (DM)

This issue was chosen as an area of study because of the attention and importance given under the European Commission’s gender equality policy unit (European Commission, 2018-19), and as a major goal of the European Women’s Lobby, detailed in their manifesto of priorities for the coming European term (2019-2024) (EWL Manifesto, 2019).

Additionally, I intend to include inequality occurrences that refer to dependency of children, family members or other dependent individuals and the responsible carer. This decision came about as a result of inductive analysis in my data collection, whereby the documents and texts analysed displayed repeated references to the issue of carers, whom suffered discrimination based on the need to care for dependent family members or other individuals. The specific definitions can be found in the EU Law text (NE6, see appendix 2) and the inequality occurrence is coded in the data as ‘carer status’.

Within each of the chosen sub-issues, a sample of documents will be taken that includes “at least one law, one policy plan, one parliamentary debate, and one civil society text”. This is done in order to include the perspectives of each of the most relevant institutional and civil society actors on a particular issue. The justification for this selection is based on the QUING methodology that aims to capture the most important institutional and civil society voices on any given issue (Krizsan & Verloo, 2006):

For my own purposes, I intend to analyse documents from the period 2009-2019 and will identify documents covering or related to the four overarching issues covered above (GBV, IC, NE, and DM).

The method of frame analysis used assesses quality regarding the content of policies, not their implementation, with the criteria for assessing the quality of intersectionality developed further below.

The quality of gender equality policies has been associated with criteria such as:

• Gender explicitness (de Vega, Rolandsen Agustın, & Lombardo 2008a; Dombos et al. 2009); • Understanding of gender inequality (Dombos et al. 2009; Verloo 2007; Walby 2007); • Comprehensiveness (Walby 2009);

• Prioritizing gender equality as an aim in itself (de Vega, Rolandsen Agustın, & Lombardo 2008a; Verloo 2007);

• Inclusiveness of civil society (Dombos et al. 2009); • Intersectionality (Dombos et al. 2009; Walby 2007).

Further criteria are developed to properly assess intersectional inclusion as one of the aspects that could improve the quality of gender equality policies. Weight is also given to inclusive policy-making, by considering EU references to consultations with civil society, since this is an issue that has been identified by scholars as critical to the use of ‘good’ intersectionality in policy, but has been lacking in practice (Lombardo & Verloo, 2009; Lombardo & Rolandsen Agustin, 2011; Verloo, 2006)

According to the methodology used, policy problems will usually include a diagnosis and a prognosis of the issue at stake. To identify the ways in which the problem is represented, the sampled documents are coded on the basis of standardised questions, taken from the QUING Manual listed above (Krizsan & Verloo, 2006) The questions used relate specifically to

intersectionality and gender.

First it is identified whether gender and any other inequality are addressed in a given document. Second, the relationship between the inequalities that appear in the text is addressed, identifying whether the relationship is additive, competing, separate, intersecting, or hierarchical. Here it is also identified whether intersectionality was clearly articulated (as specified later in this section). The purpose of this analysis is to detect the extent to which gender is seen to intersect in the policy measures considered, and if so, how, and with which other inequalities. (An overview of codes and frames on intersectionality in the EU selected documents is given as an appendix) Further questions were asked as to whether the documents referred to consultation processes with civil society actors and whether these included women’s organisations.

Beyond identification of the presence of intersectional frames, attention is also paid to references of inequality dimensions coded in the texts which, despite their lack of articulation as intersectional frames, reveal the occurrence of intersectional ideas.

As discussed in the theoretical framework, such an understanding of intersectionality allows for different degrees of intersectionality to be found in the selected documents. Intersectional ideas may be present in a text without there being reference as to how the relation between the inequality dimensions is articulated (Lombardo & Rolandsen Agustin, 2011). This is defined as “inarticulate intersectionality” and further highlights that certain texts may articulate intersectional dimensions as additive or mutually constitutive, although this is not sufficient enough to claim that intersectionality is made explicit, and occurrences must be cross-checked in the context of the document where they appear. Evidence of explicitness and articulation is identified as greater when intersectional frames can be detected, are found in more than one document and are considered to be particularly relevant according to the quality criteria.

With reference to the work of Lombardo and Rolandsen Agustin (2011, 2), the following quality criteria are adopted to assess whether ‘good’ intersectionality can be seen in EU gender equality policies:

1) Explicitness and Visibility of certain inequalities as well as Inclusiveness of a wide range of inequality categories.

a) In order for a problem to be addressed, it must first be made explicit in some way.

b) The naming of the problem gives it visibility. Crenshaw (1991) identified how the experiences of Black Women were made invisible by feminist and anti-racist activists.

a) The way in which the relations between intersecting categories are understood and the extent to which these relations are explained in an articulated way in the texts. Analysed on a scale.

3) Gendering of policy issues and intersecting inequalities

a) Explicit references to gender, as gender tended to disappear as an explicit category when other intersectional dimensions entered the policy texts (as revealed by the 2011 research)

b) Gender is a fundamental category following the gender mainstreaming approach.

c) (Although the lack of centrality given to gender has been contested in relation to poor quality, gendering enhances the likelihood that gender equality is treated as an aim in itself of any policy)

4) Structural understanding of inequalities

a) The depth and understanding of the problem and the transformative potential this understanding has for changing unequal structures.

b) Whether the policy document looks at the individual and collective (structural) level.

5) Avoiding the stigmatisation of specific groups and challenging privileges among dominant groups (the motivation behind many theories of intersectionality)

a) The naming of particular groups as problem holders can stigmatise them in the policy discourse and in public opinion.

b) To counter-act potentially sexist, homophobic, racist or other biases in policymaking, the lack of stigmatisation of specific groups and challenging of dominant groups is an essential quality criteria.

Methodological Limitations

Although the focus of this paper is tied to the use of theories of intersectionality, discourse and critical frame analysis, scholars in the field have raised concerns that policy approaches to inequality should not “only” be intersectional (Weldon, 2008; Verloo, 2006; Walby, 2007). Such a narrow approach could risk losing any observation of inequalities having autonomous effects, while an “intersectionality-only” approach might also endanger our understanding of the nature of each inequality, which is necessary to develop ‘good’ equality policies.

Ultimately, many scholars studying EU gender equality and anti-discrimination policy recognise that policymakers’ awareness of intersectionality is a crucial factor in developing policies that do not discriminate against or ignore any individual or groups at the point of intersection between different inequalities (Lombardo & Rolandsen Agustin, 2011). As a tool for analysing the nature of gender inequality alone in the selected documents, de-gendering frames in policies that include multiple inequalities is included as a key criterion.

In Lombardo and Rolandsen Agustin (2011, 19), the authors acknowledged that the criteria set for ‘good’ intersectionality in gender equality policies were not exhaustive. These criteria were developed by the two authors and other experts that contributed to the QUING research (2011). Due to the limited scope and resources available in this thesis, I will draw on these criteria, owing to the expertise of the scholars contributing to the work, while recognising that further work can and should be done to assess the quality of intersectionality in EU gender equality policies.

Due to similar resource constraints, the use of critical frame analysis is limiting since the identification of frames within the documents will be influenced by my own (subconscious) biases. Across many of the QUING research projects (2011), documents were cross-checked by multiple researchers. Since this option is not available, this thesis utilises the broad but precise assessment criteria laid out by Lombardo and Rolandsen Agustin (2011) to study intersectionality and gender

equality, which supplements the dynamic theoretical conceptualisations of both. The same issues regarding resource and scope limitations arise when considering the case selection of EU policy documents. To tackle this problem, the methodology developed by the former researchers is followed closely, as outlined in the previous section, in order to capture the most important perspectives on the issues.

This thesis focuses on the analysis of policies rather than their implementation, meaning that references made to consultations between EU institutions and civil society organisations does not explore the breadth of these consultations or the diversity of organisations included. As per Lombardo and Rolandsen Agustin (2011) “…the mentioning of consultations is considered a first step towards the design of more inclusive policy-making” (Lombardo & Rolandsen Agustin 2011, 19) Further research has been conducted on the topic of EU-civil society cooperation by Cullen (2015), Ferree (2009) and Lombardo et al. (2009).

Analysis:

Trends in the framing of intersectionality in EU gender

equality policies

In this section, the main trends in the framing of intersectionality in EU gender equality policies will be outlined. On the basis of the quality criteria mentioned in the previous section, the implications of these trends will then be judged. Reference are made to the figures in table 1 (found in Appendix 1) and letter-number combinations (e.g. NE1, DM1, IC1, GBV1) refer to the coded documents for each issue (listed in Appendix 2.)

Throughout the sampled documents, gender does not appear as the only inequality in a single case. However, it is clear that although other inequalities than gender are mentioned, intersectional frames are rare. Moreover, the relations between inequality categories are often poorly articulated. Many of the texts also show a discrepancy in the language used during the diagnosis and prognosis of the issue. Inclusivity of inequalities remains low within many texts and further hinders articulation of the relationship between inequalities. There is also a clear tendency to move towards de-gendered language when other inequalities are mentioned alongside gender. The choice of language in certain texts has helped to remove stigma from certain groups, while it continues to reinforce stereotypes and diminish the effect of policy solutions in others. A structural understanding of the issue and inequality appears within every issue and sub-issue, with several texts offering transformative solutions to the problems. Many of the texts show weakened structural understanding and transformative potential where differences in the quality of the diagnosis and prognosis appear. The imbalanced understanding between diagnosis and prognosis across many issues also presents concerns over the visibility and de-gendering of inequalities. Positive trends can be identified in the references made to consultations with civil society organisations, and women’s organisations in particular, which are present in many texts. Overall, the results of the analysis show that the EU gender equality policies analysed show increased attention to gender and other

inequalities. These improvements are coupled with a greater structural understanding of inequalities that goes beyond individuals and addresses the collective level. However, it remains that many of the texts analysed demonstrated no intersectional understanding or failed to articulate the

relationships between inequalities. As such, it is not possible to say that EU gender equality policies showcase intersectional quality, based on the trends seen in the texts and due to the scope of this research. Ultimately though, improvements continue to be made and the trends identified have many positive implications.

Explicitness, Visibility, and Inclusiveness of Intersecting Categories

In the sample of EU documents, only a small number of frames are present that explicitly refer to or address intersectionality. From the sixty-one coded texts, there are explicit intersectional occurrences in twenty-seven texts, including parliamentary voices. The ‘intimate citizenship’ (IC) issue contains roughly 72 percent of the total explicit intersectional occurrences. This is followed by ‘gender-based violence’ (GBV) with 15 percent of occurrences, ‘nonemployment’ (NE) with 12 percent of occurrences, and ‘gender-balance in decision making positions’ (DM) occurrences comprising just 1 percent of the total.

There is a total of one hundred and fifty-one occurrences of explicit intersectional codes in the sample. Combining all four issues, there is a total of eighty-three occurrences of explicit

intersectional codes in the diagnosis of thirteen texts. This represents an occurrence in the diagnosis of the issue in roughly 21 percent of the total number of texts. There are sixty-eight occurrences of explicit intersectional codes in the prognosis of fourteen texts, around 23 percent of texts.

In total fifteen different frames were detected. The frames found vary in their presence across documents, as well as by issue and sub-issue. Generally, the frames are present in low numbers of texts, with the most widespread frame ‘transform division of labour for parents and

carers’ appearing across seventeen texts, in the NE and DM issues combined. This frame refers

primarily to concerns about the gender-pay gap across EU member states and the inequalities that prevent women from earning equal pay and status for equal work. One of the main inequalities factored into this frame is the role women play as carers for children and other dependents. This is voiced as problematic with regard to the reconciliation of work and private life, especially for women. The frame appears relevant in both issues (see table 1) due in part to its high presence in the diagnosis and prognosis of multiple texts.

In the DM sub-issue ‘gender balance in decision-making positions: politics’, the frame ‘transform division of labour for parents and carers’ is voiced in texts by all three European institutions as well as civil society actors. References are made most prominently to gender and individuals with dependents (carer status), but also include relationships to: age, race/ethnicity, disability, sexual orientation(sexuality), class and regional belonging (regional)(DM5, DM7, DM9, DM10) Texts in the sub-issue DM: politics demonstrate inclusivity of a large range of inequalities, although many of the inequalities other than gender and carer status occur in low numbers and in few texts. In the DM sub-issue ‘gender balance in decision-making positions: business’ the same frame ‘transform division of labour for parents and carers’ is present, though with far fewer inequalities mentioned together. The inclusiveness of texts in the sub-issue DM: business is lower compared to DM: politics and thus fewer intersections are made visible. Moreover, the relationships between the inequalities are most commonly expressed separately or inarticulately in the DM issue. Articulation is discussed in the following section on trends but gives context to the occurrences mentioned here.

One of the most explicit occurrences of intersectionality, the frame ‘Multiple and

intersecting discrimination of at-risk and vulnerable women’ also occurs in DM: politics but is the

only other frame present in the DM issue. The frame occurs in just two texts (DM7, DM9) and illustrates how women face multiple and intersecting discrimination when trying to access certain positions in the labour market. It is used in combination with the previous frame ‘transform division

of labour for parents and carers’ to highlight the gender stereotypes that women face when

burdened with the responsibility of care for dependents such as children. The texts (DM7, DM9) explicitly acknowledge that “stereotypes particularly influence women belonging to under-represented groups”, specifying inequalities such as race/ethnicity, regional belonging, age and class, in combination with gender. The text goes on to state that “they (under-represented women) are made hardly visible in political decision-making and their access and participation in public and political life is even more challenging than for other women” (DM7). Despite the clear explication of mutually shaped inequalities, the frame ‘Multiple and intersecting discrimination of at-risk and

vulnerable women’ occurs just once in two different texts. As such, explicit occurrences of

intersecting inequalities are low in number.

Furthermore, explicit intersectional occurrences were found solely in the diagnosis of the DM texts (DM7, DM9). It should be considered that explicitly defining the inequalities and their relationships does provide visibility to the problem, which is recognised as a starting point for good quality intersectionality. However, the limited number of explicit occurrences, with very minor relevance found in the prognosis of the texts, presents a flaw in the understanding of gender equality and intersectionality.

Nonemployment (NE)

The frame ‘transform division of labour for parents and carers’ is voiced in nine separate texts in the nonemployment issue (documents NE2, NE3, NE6, NE7, NE8) including multiple voices in the parliamentary debates. The frame is present in both sub-issues, ‘reconciliation of work and

private life’ as well as the ‘gender-pay gap’. In contrast to the frame’s presence in the DM issue,

within NE there are multiple occurrences in both the diagnosis and prognosis of the texts, establishing more thorough understandings of the intersectional inequalities.

It is clear from the inequalities and intersections that can be found in the NE texts that the sub-issues are inextricably linked. Across the NE issue as a whole, age plays a more inclusive role and intersects with the dominant gender and care axes. The intersectional occurrences between gender, care, and age go one step further here than in the DM issue, specifically addressing the way in which older women are affected by the burden of care for children and dependents.

The intersectional frame ‘remove gender-pension gap’ centres the relationships between age, gender, and care (NE2, NE3, NE4, NE6) by connecting the two NE sub-issues. The texts in these documents make visible the problems faced by women who have responsibilities of care over dependents, resulting in and combined with other barriers when accessing the labour market. As a result, the gender-pay gap thus becomes a gender-pension gap for older women. Other inequalities intersect in these issues, such as class and disability. All of the inequalities mentioned are made explicit in the texts when referencing the disproportionate poverty rates among older women.

In texts on NE, visibility is given to intersectionality through the frame ‘recognition of diverse

family structures’ that expresses the need for greater awareness and measures to address “the

impact of heteronormativity in legal parent recognition” (NE8). Just two texts in the NE issue (NE6, NE8), adopt this frame, which is formed from the relationships between several inequalities, including sexual orientation (sexuality), marital/family status, gender and carer status. Age and

disability are also voiced. Similar in terms of inequalities, the frame ‘intersectional

non-discrimination’ can also be found in NE (NE8). Although the frame occurs in just the single text, its

relevance is major within the text and showcases some of the most explicit occurrences of

intersectionality of all the documents and texts sampled. The frame examines the rights to parental leave that same-sex couples are afforded, as well as the intersectional discrimination faced by gay and lesbian same-sex partners, related to legal rights as spouses and/or parents and guardians. Overall the NE texts incorporate lesser inclusivity of inequalities than the DM texts, but the

inequalities that receive focus in NE documents occur more commonly in multiple texts, making the inequalities more visible. The greater occurrence of explicit intersectional occurrences in the NE issue is also of note, served by the visibility and explicitness of inequalities in multiple texts.

Intimate Citizenship (IC)

The IC issue is comprised of the sub-issues ‘sexual orientation discrimination’ and ‘women’s

sexual and reproductive health rights’. Mirroring the NE frames ‘recognition of diverse family

structures’ as well as ‘intersectional non-discrimination’, frames found within the intimate citizenship

issue contain several similarities. For example, the intersections between gender, age, sexuality, and marital status occur in high numbers in several IC frames. The frame ‘remove civil status restrictions

for LGBTQI people’ intersects each of these inequalities, further addressing the “multiple” or

“accumulated” discrimination that affects LGBTQI individuals (IC2, IC4, IC5). The frame includes inequality relationships that address: the discrimination and abuse faced by “young LGBTI

individuals” among other vulnerable groups (IC2) (gender, age, race/ethnicity, sexuality); the barriers placed in front of LGBTQI partners and families seeking reunification (gender, sexuality, marital status, citizenship status); the barred access to inheritance or pensions, among other benefits and legal rights, based on “multiple discrimination” (IC5) or “double invisibility” (IC4) (gender, age, sexuality, marital status, citizenship status, disability). The texts IC4 and IC5 offer some of the most

explicit intersectional occurrences to be seen in the entire sample of documents, including intersectional frames that cross over multiple texts.

‘Support ageing LGBTI communities’ is an IC frame closely tied by subject to the previous

frame. Whereas citizenship and marital status were prominent intersections with sexuality in the previous frame, here the inequalities of sexuality and age are intersected, though gender and disability axes are also incorporated (IC4, IC5). The significance of age across the IC issue is reinforced by the presence of the frame ‘empower older women’ which combines intersectional elements from IC, NE and DM frames to express how women of a certain age suffer from multiple discrimination. They form one of the most vulnerable groups in society. Inequality occurrences found in these texts include gender, age, carer status, marital status, sexual orientation, disability, and class. The clear links between not only the inequalities but also the issues and frames make visible the vulnerable position of women belonging to intersecting at-risk groups. Despite the explicit framing of the problem, age still only occurs in either the diagnosis or the prognosis of seven of the fourteen IC texts, the vast majority of which are found in just two texts, IC4 (civil society) and IC5 (European Parliament FEMM committee). The lack of explicit gender-age and any other intersection in half of the texts reduces visibility of the problem considerably.

A central part of the discourse on citizenship in the EU continues to revolve around migration, asylum seekers and refugees (European Agenda on Migration, EC 2015). Hence the occurrence of these themes in policy documents within the IC issue and sub-issue ‘women’s sexual

and reproductive health rights.’ The most frequently occurring inequalities captured in the sub-issue

are gender and citizenship status, often intersecting. Other inequalities such as age, marital status, carer status, sexual orientation, regional belonging, and disability are also captured. The frame ‘adapt asylum procedures for multiple marginalised individuals’ includes references to each of these inequalities, many of which are voiced as intersecting. The frame is made visible and explicitly addressed in two texts (IC6, IC8) but is weakened since very few inequalities cross over between the

two texts. The inequalities with particular relevance in both IC6 and IC8 are age, gender, and

citizenship status. ‘Adapt asylum procedures for multiple marginalised individuals’ is voiced as a tool to tackle problems such as the citizenship status of female migrants, refugees and asylum seekers being tied to their marital status, and also being made vulnerable due to their economic dependence on such spouses for this reason. There is also recognition in the texts that asylum claims made for reasons of persecution suffered by an individual for their sexual orientation are not given the attention they require. This intersection is then given visibility through its framing in this light. There is also explicit reference to the needs of girls as asylum seekers and refugees, whose unique needs are often subsumed in policy texts under terms such as “children”, “women and girls” or

“unaccompanied minors” (IC8). One of the strengths of the intersectionality found in this frame is that it occurs explicitly in diagnosis and prognosis.

The frame ‘non-discrimination in healthcare provision’ has specific relevance to the texts on migration, but also throughout the IC issue. This frame is dominant in the diagnosis and prognosis of six IC texts (IC4, IC5, IC6, IC7, IC8) including parliamentary voices. Concerning the needs of female migrants, refugees and asylum seekers, the frame captures the various intersections between gender, age, citizenship status, carer status, disability, race/ethnicity and regional belonging. In this context, the frame highlights the need for specialist social and healthcare assistance for many women and girls when they enter refugee processing centres. The other texts referencing this frame cover a range of issues, including access to healthcare for women and girls in poverty, women and girls of all ages, available in rural areas, specialised based on sexuality and gender identity, accessible to women with disabilities or dependents and dependents with disabilities, plus access for

individuals at the intersections of any and all of these categories. The ‘non-discrimination in

healthcare provision’ frame is just one of two frames of the fifteen total that is dominant in the

diagnosis and prognosis of multiple texts, holding major frame relevance. The other dominant frame ‘transform division of labour for parents and carers, is discussed above. The inclusivity of a wide range of inequalities is exemplary in the IC texts. IC is the only issue of the four that voices every

inequality occurrence in more than one text, when combining the diagnosis and prognosis occurrences.

Gender-based violence

Four intersectional frames are present in the GBV issue, though none occur in more than two texts, signalling low visibility. The explicit intersectional occurrences found in this issue provide some balance to this, since intersectional language dominates individual texts, but without

transference across texts.

In the GBV sub-issue ‘sexual harassment’, the frame ‘targeted gender-age harassment’ is present (GBV5), displaying explicit intersections between gender and age, but with very few occurrences of this relationship throughout the single text it appears inside. As such, the frame registers less dominant diagnostic relevance and only minor prognostic elements, since most of the inequality occurrences, including intersectional occurrences, appear solely in the diagnosis.

‘Availability and accessibility of specialist women’s support services’ is another frame that also appears exclusively in one text (GBV8) in the GBV sub-issue ‘sexual harassment’. The frame is inclusive of a wide range of inequalities. Intersections of gender with race/ethnicity are mentioned, as are relations between gender, carer status, disability and class. Both of these intersections are further mapped related to the need for specialised support services such as healthcare, be they for the women or their dependents, who may belong to a vulnerable group. Identifying the needs of women belonging to communities in rural areas is also factored into the frame. In spite of its minor relevance across texts, the frame is of major relevance all through the text (GBV8) and the inclusivity of inequalities is showcased.

The final two GBV frames identify similar intersectional inequalities and problems, but due to the importance of the language identified they are presented separately. The frame ‘intersecting

forms of discrimination’ is present in two texts (GBV2, GBV4) and is one of the most explicit

occurrences of intersectionality in the sample of documents. The frame is used most explicitly in the text GBV2, in reference to the goal of gender equality and equal rights for all women, and for example “including women from religious and ethnic minority groups who face intersectional inequalities” (GBV2). The visible intersecting inequalities here are gender, race/ethnicity and religion/belief. Mention is then made of “the rights of the most vulnerable minorities” which incorporates further inequalities of gender, sexual orientation (sexuality) and class. Other

inequalities including disability, age, citizenship status, and carer status are also prevalent in the text. The weakness of the frame used here is similar to that of the frames found in DM7 and DM9.

Although the GBV frame ‘intersecting forms of discrimination’ utilises explicit intersectional concepts, it does so primarily in the diagnosis of the issue. Trends in the articulation of inequalities and their relations become clear from this distinction, discussed in the forthcoming section on articulation.

The final frame to be discussed is ‘multiple discrimination of at-risk and vulnerable women’ that appears in two GBV texts (GBV4, GBV6) and explicitly discusses the relationships between inequalities that make certain individuals and groups vulnerable. The frame is inclusive of the following inequalities: gender, race/ethnicity, religion/belief, sexual orientation (sexuality) age, disability, marital status, citizenship status, regional belonging. Specific language includes inequalities such as gender, age and marital status which intersect to make visible how “elderly people, especially older single women” are a “particularly vulnerable social group” (GBV6) according to the data on sexual harassment. Here the framing and use of intersectionality is again weakened by a minor prognostic relevance.

Articulation of Intersectionality

As explained in the methodology section, the articulation of intersectionality is considered greater when the relation between inequalities appears as mutually shaped. Additive and separate articulations are also counted. Where the relationship appears as inarticulate in the text, the

explanation or understanding of the nature of the relationship between inequality categories may be inexplicit or lacking, however intersectional ideas may still be present.

Intersectionality is articulated in varying quality in the selected EU gender equality documents. Mutually shaped relations can be found in the diagnosis of sixteen texts (26 percent) and in the prognosis of twelve texts (19 percent). Despite the articulate understanding and use of mutually shaped inequalities in these texts, the numbers establish evidence that articulate intersectionality is not present across a wide range of EU gender equality documents. Inarticulate relations between inequalities occur in the diagnosis of twenty texts (32 percent) and the prognosis of twelve texts (19 percent). Although the total number of inarticulate occurrences is lower than the number of mutually shaped relations, an inarticulate expression of the relationship between

inequalities is present across a greater number of texts. Moreover, the majority of occurrences of mutually shaped relations can be found concentrated in just a single issue, intimate citizenship (see table 1). These occurrences can be found in documents from only civil society organisations (IC4, IC8) and the European Parliament institution (IC2, IC5, IC6).

Some of the most explicit intersectional frames do not detail the relationships between the inequalities in the text but use language such as “multiple and intersecting forms of discrimination” (DM7) that demonstrates a deeper understanding of discrimination. Other frames such as ‘support

ageing LGBTI communities’ do not use explicit terms such as ‘multiple’ or ‘intersectional’

discrimination however the relationships between the inequalities are explained so articulately that the occurrence can only be listed as mutually shaped. For example, the civil society text IC4

discusses how the lack of legal recognition for same sex couples in some countries is a barrier to certain entitlements. This is a particular source of insecurity for older people in same-sex couples, who are denied entitlement to pensions or establishing inheritance. The text further describes how “neither the LGBTI nor the older populations are homogenous groups; therefore, older LGBTI individuals often face multiple discrimination that extends beyond homophobia, transphobia and ageism.” (IC4) and is also inclusive of other inequalities, forming new intersections. Even though a large proportion of mutually shaped occurrences are concentrated in IC, mutually shaped relations also occur the greatest number of times in the NE and GBV issues.

Inarticulate relations are less concentrated, occurring in multiple texts across all four issues. These occurrences suggest that the understanding of relations between inequalities in EU gender equality documents is mostly inarticulate, though intersectional ideas are present in some of the occurrences. Examples can be found in documents GBV1, IC6, DM7 and NE6, where multiple inequalities are presented together, often expressing the marginalising effects they can have, but without clear expression of the relationships to one another and new mutually shaped intersections that may form. In NE6, inarticulate occurrences include the inequalities gender, age, carer status, disability and marital status. In some texts, it is made clear how these inequalities intersect and create problems for the reconciliation of work and private life for many women, which ultimately leads to gender-pension gaps. In NE6 however, the relationships between the inequalities are not always expressed clearly, leaving any marginalising effects to interpretation. Inarticulate relations often occur in conjunction with the de-gendering of policies and other inequalities. Whereas explicit mutually shaped occurrences intersect gender, age and other inequalities to frame the gender-pension gap, inarticulate occurrences may refer to “parental leave” and “parents in particularly disadvantaged situations related to disability…” (NE6) but also reference gender inequality elsewhere.