Gender and the EU's Support for Security Sector Reform in Fragile Contexts

How does the European Union (EU) include &#8216;gender&#8217; within its support to security sector reform (SSR) programmes? The EU has committed to include gender perspectives by implementing the Women, Peace and Security agenda (WPS) within its foreign security practices. While researchers and practitioners recognise the importance of integrating gender issues into SSR operational effectiveness, there is limited knowledge about how this functions within the EU's security architecture. This article uses Feminist Institutionalism (FI) to understand the process of gender mainstreaming within the EU's support to SSR programmes. It does this by using two crucial theory&#8208;testing cases of SSR programmes &#8211; Ukraine and Afghanistan. It finds that the EU's ability to promote gender inclusive approaches to SSR is limited by the structure of the EU's own assumptions and capabilities, and institutional constraints in third countries. At the same time, the cases underscore the importance of individuals as agents of change.


Introduction
In its broader external security engagements, the EU has committed to including gender inclusive perspectives (European Commission 2016a). The EU aims to address problematic gendered dynamics of women's exclusion, while making visible their experiences within international security. Moreover, as some scholars have argued, these gendered exclusions can be inhibitors to peace in a variety of ways (Bjarnegård and Melander 2011). Yet, presently, security institutions particularly in conflict or fragile contexts are gendered (see Cockburn and Hubic, 2002;Duncanson 2016;Enloe 2014;Goldstein 2006;Karim 2016). Indeed, Security Sector Reform (SSR), which focuses on restoring security institutions, often emphasizes the status quothus reproducing toxic hegemonic masculinities and typically rests on assumptions that remain gender-blind (see for example Atkinson 2016;Brown 2007;Duncanson 2013;Sjoberg 2014). Including gender perspectives during processes of transformation can help to mitigate the harmful effects of those spaces dominated by men and bolstered by patriarchy (True 2013).
However, to address the entry points for gender inclusivity, the role of actors must be understood. Despite the increased roles of intergovernmental organizations in SSR support the operation and effect of political institutions (Mackay et al. 2010, p. 574), we can begin to tackle 'the gendered character and the gendering effects' of political institutions (Mackay 2010, p. 181). For this study, we test how these assumptions of FI operate in the case of EU SSR processes. The two cases thus also help us to further refine theoretical assumptions on the processes of including gender perspectives into the EU security architecture.
The remainder of this article will proceed as follows: in the next section we present the theoretical framework. Section II deals with the analysis of SSR programmes by the EU. Section III presents the methodology we use for our analysis. In section IV we analyze the processes of gender mainstreaming in the EU's security sector reform programmes. The final section concludes and situates the findings within wider debates.

I. Feminist Institutionalism: An Approach to EU SSR Support
Feminist scholars studying post-conflict peacebuilding have often focused on exposing and addressing the role of dominant masculinities in the creation and re-creation of security institutions and insecurities (cf. Cohn 1987;Kronsell 2005Kronsell , p. 1033McLeod 2015McLeod , 2016. To an extent, SSR lends itself to this sort of opening for feminist and gender interventions because it is still dominated by hierarchical and problematic masculinities. Although many studies adopt a gender lens in their analysis, they continue to be very policy (outcome) oriented (see Bastick 2008;Salahub and Nerland 2010;Valasek 2008). At the same time, they lack a theoretical explanation of evolvement of dominant masculinities in the design, implementation and effects of SSR programmes. They thus often exclude knowledge about the institutional processes that allow or constrain gender-sensitive SSR design and implementation. Feminist analysis of institutions, we contend, can fill this gap.
Institutions are at the core of political relations (March and Olsen 1989). Moreover, they have an effect on the outcome of policy practices since they are derived from the particular nature of institutional design and evolution thereof (Haastrup 2013, p. 52). They are essential determinants of political behaviour (Peters 2012, p. 164). However, institutions are profoundly gendered (Thomson, 2018, p. 2). By this, we mean, 'constructions of masculinity and femininity are intertwined in the daily culture or 'logic' of political institutions' (Mackay and Krook 2015, p. 6). It is thus in trying to challenge the hegemonic masculinities of institutions that we rely on a variant of New Institutionalism, Feminist Institutionalism (FI).
Feminist Institutionalism or FI (Chappell and Waylen 2013;Kenny, 2007;Kenny and MacKay, 2009;Mackay et al. 2009;Mackay and Krook 2015;Waylen 2014) draws on certain concepts within the historical institutionalist (HI) variant of New Institutionalism (NI). The concepts of HI are useful as they help to illuminate the role of actors or agents, temporal dimensions and unintended consequences of institutional choices (Bulmer 1993;Pierson 1996).
To address how gender perspectives are included within SSR, we draw on FI specifically given its feminist insights. FI contributes four unique elements to NI (Thomson, 2018, p. 2ff.), which make it particularly salient for our analysis. First, it centres gender as fundamental to how institutions function. Institutions reproduce the gender differences and hierarchies found in society. FI thus challenges the neutrality of institutions, which has implications for the ways in which gender becomes understood and prioritized (Acker 1992) in the relations between actors, institutions and practices.
Second, it underscores and elaborates on the notion of power. For feminists, the distribution of power within institutions is important because it potentially has a constraining effect on the inclusion of intended norms and practices aimed at change. Gendered power dynamics determine the decision-making structures and the decisions themselves can reinforce these dynamics within institutions (Mackay et al. 2010, p. 583).
Third, whereas NI acknowledges informal institutions, most works tend to focus on formal structures. FI scholars like Chappell (2006Chappell ( , 2014 argue that a 'gendered logic of appropriateness' informs the practices of formal institutions. These practices are not written down, but where they embolden or sustain gender hierarchies, they are evidence of the power of the informal and they too deserve acknowledgement. Finally, FI also underscores the importance of specific individuals or interest groups within specific institutional contexts in facilitating change (see also Thomson, 2016). In formal institutional settings, observing the work of these individuals or femocrats may help to get a better reading of the opportunities and constraints of the institution to accept change.
Our use of FI is aimed at gaining insight into the gendered character of the EU institution that produces SSR programmes and what effects this may have on the SSR programmes under consideration even when actors attempt to include gender perspectives (Mackay, 2010, p. 181). Thus, in addressing the research question, FI is used to 'question […] the interplay between gender and the operation and effect' (Mackay et al. 2010, p. 574) of the EU's SSR processes.
Our analysis of the EU's support to SSR programmes focuses on the inclusion of gender aspects in two ways. First, we examine the design and conceptualization of the programmes and assess if gender was already included in the concepts and key legal documents of the missions. Second, we assess the inclusion of gender aspects in the implementation of SSR programmes on the ground. In these two contexts, we also examine the ways in which key figures act as advocates of gender mainstreaming in SSR.

II. Implementing Women, Peace and Security through SSR -What is Possible?
According to the EU, 'Security sector reform (SSR) is the process of transforming a country's security system so that it gradually provides individuals and the state with more effective and accountable security in a manner consistent with respect for human rights, democracy, the rule of law and the principles of good governance' (European Commission 2016b, p. 1). Security sector institutions include: law enforcement institutions (police, criminal justice system, armed forces and intelligence services), oversight institutions (parliamentary committees or court of auditors) and non-state security actors including customary authorities (European Commission 2016b, p. 2).
In SSR, the EU is not the only actor in a rather crowded field. However, as Law (2007) argued, regional and intergovernmental institutions will increasingly take on roles within SSR. And despite the limited attention to the EU, the extent of its engagement in SSR support underscores Law's argument and part of our motivation for this article.
The EU has contributed in a variety of ways to security sector reform. We focus especially on those cases fitting within the definition of EU SSR. 2 EU SSR mission types fall into three main categories: military, police and judicial reforms, with the two latter providing civilian components of SSR. In terms of direct engagements, the EU has supported military reform in Bosnia, Democratic Republic of Congo and Guinea Bissau. In its missions, the EU has also supported police reform aimed at making police forces more accountable and whose primary function is to ensure the safety and security of citizens. Examples of these types of programmes include EUPOL COPPS in the Palestinian Territories, EUPOL Afghanistan and EUAM Ukraine. Finally, the EU has also supported judicial reform in Kosovo and Iraq where the goal was strengthening and upholding the rule of law. This article is particularly concerned with the civilian aspects of EU support to SSR missions.
The 'entry' of gender into the EU's SSR programmes is based on the EU's commitment to the implementation of the WPS agenda. The WPS agenda originated with United Nations Security Council Resolution (UNSCR) 1325. Now a collection of eight resolutions, the WPS agenda has allowed for the consideration of the gendered nature of insecurities by bringing to the fore issues such as wartime sexual violence, the different experiences of women in conflict and how these experiences may influence different types of peace and security. The WPS agenda, thus, provides a global normative framework for including gender perspectives within security practices.
Within the EU, a basic guideline was issued in 2008 requiring the integration of the WPS agenda into all EU policies, with a specific sub-policy on integrating the WPS agenda into the CSDP since 2012 (EU Council 2008; Interviewee E, 2017). To fully enact this normative framework, the EU committed to using the strategy of mainstreaming to ensure gender sensitivity in its security programmes including SSR (Interviewee E, 2017).
However, gender (and mainstreaming) can vary based on different applications of feminism, with consequences for the policy and practices of inclusivity. As such, the extent to which the EU actors can advance the WPS agenda is constrained, even when gender mainstreaming should ultimately transform institutions so that gendered power dynamics are broken down.
True and Parisi (2013; see Table 1) offer a useful descriptive typology that elaborates on the implications of these different interpretations of gender within an institution. The application of these different modes is often not deliberate since the EU broadly aims to be transformative. These modes, however, are the results of practice. And it is through our analysis of EU SSR institutional processes that we are able to determine the EU's gender mainstreaming model. Ultimately, knowing what modes are eventually adopted has implications for policy outcomes because this is how they become institutionalized.
There is one relevant caveat to the peace and security sector. True and Parisi (2013, p. 40) argue that institutions may find gender mainstreaming threatening, confusing or irrelevant to what they understand to be their core mission and thus adopt gender mainstreaming so superficially as to be a rejection of the concept. Moreover, institutions can concurrently employ more than one model of gender mainstreaming further inhibiting the likelihood of transformation.

III. Methodology: Case Selection and Sources
As we have noted, we apply FI to two cases of EU support for police reform as a way of exploring what SSR programmes reveal about the EU as a gender and security actor. This study follows a non-comparative research design. Our cases serve as illustrative cases for the aim of testing the assumptions of FI theory on the question of how gender is mainstreamed by the EU within security environments.
Our universe of cases includes all EU SSR support missions (see Table A1 of the online Appendix). Following classical works on case selection for non-comparative research design (George and Bennett 2005, p. 32f; Gerring 2012, p. 95ff), we committed to a case selection strategy that chose one hard and one easy case as a test for our theory. This provides a useful variation on the dimensions of theoretical interest (for the purpose

Core Assumptions & Rationales Implications
Gender-as-sameness (Liberal Feminism) Neutrality between men and women will achieve equality. Adding women into domains typically dominated by men.
More women's representation in previous areas of exclusion.
Does not question implications of, and expects conformity to existing male norm. Gender-as-difference (Difference Feminism) Accepts distinction in the roles of men and women in society but values both equally. Revaluing women/ femininity while promoting women's contributions as distinct will achieve equality.
The implications and contributions of women's traditional roles are brought into focus with the possibility of mediating the effect of problematic masculinities. Reifying gender difference essentializes women, and potentially reproduces gendered hierarchy. Gender-as- Accounts for the complexity of gender relations. It reflects critically on the social inequalities that come into play with gender.
Not all women (and men) are equal in the sense that insecurity during conflict and post-conflict situations may put women and men of a certain ethnic group at more of a disadvantage than others. Gender interventions could also reinforce these societal hierarchies if intersectionality is not a consideration for gender mainstreaming. Gender-as-Transformation The approach's central aim in including a gender perspective is to challenge existing frameworks, often underpinned by patriarchy.
By upending existing systems of governance and governing, this approach promotes gender equality and justice. This model is radical and calls for the transformation of gender relations to create a new standard for both women and men.
Source: Adapted from True and Parisi (2013) pp. 39-40. of theory testing) and a reasonably representative sample (Seawright and Gerring 2008, p. 296). First, the Afghan case poses a hard test for FI, as the relegation and disenfranchisement of women from economic, political and social life by the Taliban (OECD Social Institutions and Gender Index, 2015) is especially pronounced (Murray 2007, p. 110) in comparison to other cases of EU support to civilian SSR. This context makes the inclusion of gender issues into the realm of security institutions particularly challenging. Afghanistan is deemed one of the most dangerous places for women according to the OECD (2016). And while other studies on SSR in Afghanistan have considered the roles of NATO and the UN (Sedra 2014), the EU's contribution to the civilian component of SSR has had limited exposition (for an exception see Larivé 2012) and is often missing a gender analysis. The focus of our study in addressing the EU's ability to foster gender inclusivity is thus a good test of its own self-regard as a gender and security actor. Other hard cases of EU civilian CSDP missions as a test for the theory would be the DRC, Mali or Iraq, as the situation for women is also particularly challenging in these environments (OECD, 2016); thus making Afghanistan representative in the SIGI categories 'high' and 'very high'. 3 Second, unlike Afghanistan, Ukraine has a reasonable legal gender equality framework (OECD 2016) and formal legislations upholding the principle of gender equality between men and women and the promotion of women's rights. As a legal gender equality framework was already in place before the EU mission started, it can be assumed that the inclusion of gender equality into reform efforts of the security sector might be easier than in other cases of EU SSR support, where no legal requirements were in place before the mission started. The Ukrainian case thus poses an easier test for FI theory (cf. Gerring 2012, p. 95ff). One other easy case of EU civilian CSDP missions as a test for the theory would be Bosnia, as it faces a similar environment for gender equality as Ukraine (OECD, 2016); Ukraine is thus representative for the SIGI category 'low'. Moreover, while some scholarship has evaluated the EU SSR support in Ukraine (see Nováky 2015; Zarembo 2017), these are limited and have not engaged with the gendered implications of the mission.
Further, the EU narratives around interventions in Afghanistan and Ukraine have framed the security situations as urgent and essential to Europe's own security. The EU situates its involvement in Afghanistan as essential to fighting terrorism in Europe. Similarly, conflict in Ukraine, an EU neighbourhood country, is deemed to have an impact on the EU itself. Indeed, the Ukrainian conflict has been framed as a direct threat to the EU's security order. In this sense, the motivations for EU engagement in both Afghanistan and Ukraine are quite similar.
These two case studies of police reform constitute unique cases perceived to have direct and far-reaching implications on the security of the EU itself. Although representing different regions of engagement, they provide the best opportunities for understanding EU practices in civilian missions as a gender and security actor. In analyzing the case studies, we are able to analyze the EU's preferences for gender equality when assessed against the descriptive typology of mainstreaming.
To analyze how gender is mainstreamed in the missions, we use various sources. First, we conducted a number of semi-structured interviews varying in length between 45 and 90 minutes with EU officials in Brussels and in the missions in June 2016 and March 2017. We also use primary and secondary sources such as mission statements, legal documents, reports, official figures, as well as case studies on the missions. Finally, we draw on civil society reports as a means for engaging with alternative explanations.

Ukraine
The EU formally launched the EUAM Ukraine mission in December 2014. The mission, a direct response to political tensions, violence and subsequent armed conflict within Ukraine, was the only invited SSR mission targeting the reform of the civilian security sector. EUAM Ukraine is a police and rule of law mission. The political environment within which SSR was being implemented was one that was dominated by an old ruling elite in power, characterized by nepotism. For instance, despite many legal protections, the Ukrainian Ministry of Interior had deliberately kept women out of the police services with unofficial quotas to keep their recruitment to below 10 per cent per year (Denham 2008, p. 19). According to one EU official, women in security institutions 'don't occupy middle and senior management positions' (Interviewee A, Palagnyuk, 2016). Within the society itself, 'violence against women is a huge problem in Ukraine' (Interviewee D, 2017). Using EU commitment to the WPS agenda would thus have been welcome in the case of Ukraine as a way to account for and respond to substantive gender inequalities.
However, in the formal mandate and mission framework, there is no mention of the implications of the gender dynamics necessary for reforms or the role that women play in the concept and design of the mission (Council Decision 2014/486/CFSP, Council Decision (CFSP) 2016/712). The call for Member State contributions to EUAM Ukraine noted that 'the EU strives for improved gender balance in CSDP operations in compliance with UNSCR 1325' (EEAS 2014(EEAS , 2016. However, the eventual make-up of the mission did not reflect parity. The sum total of the staff working for the EUAM Ukraine is 76.1 per cent men, and 23.9 per cent women (Interviewee A, Palagnyuk, 2016). The core EU team supporting the mission by the beginning of 2016 was led by three middle-aged men, defying the EU's own internal commitments to gender equal representation. There is a profound imbalance within the EU's own support architecture (Interviewee A, Palagnyuk, 2016). Thus, the EU's WPS obligation to encourage representation and participation is not reflected in the mission.
From May 2016, however, one gender adviser was embedded within the EUAM in Kyiv. Nevertheless, the dominant view is that including gender issues has been constrained by the initial design of the mission, the EU's commitment and the situation on the ground. The gender advisor of the mission points out some difficulties: 'We are not allowed to influence Ukrainian policies. … [W]e can provide trainings and other information to enlighten Ukrainian politics. But we have to be careful not to push too much, because we don't have that kind of mandate, we don't have an executive mandate' (Interviewee D, 2017).
Indeed, on the general difficulties of the EU's role in Ukraine a former high-ranking member of the Civilian Planning and Conduct Capability is in general agreement. In terms of co-operation with politicians and bureaucrats for SSR (including gender sensitive SSR) they note the following: 'Frankly it is quite difficult in some countries. I am thinking about the Ukraine: The time it took to actually get them to trust you. In theory they give you access, but actually they don't want your advice' (Interviewee C, 2017).
Yet, despite the lack of support for issues around gender related to the EUAM mission, and in spite of on-going challenges, 'Ukraine is in a reform mind-set.
[…] There is a lot of openness, a lot of interest in moving forward especially on gender equality' (Interviewee D, 2017). But, change did not come from the EU. It was the appointment of Ekaterine Zguladze as First Deputy Minister of Internal Affairs of Ukraine that opened the space for including more women in the police forces. Taking on a position traditionally occupied by a man, Zguladze initiated significant police reforms that take gender seriously. For example, she piloted a new patrol police force and ensured a quarter of new recruits were women. With the increase in women police officers, Ukraine has also started to see higher reporting in cases of domestic violence, likely due to higher levels of trust in female police officers.
Overall, on whether the EU's role enhances gender sensitive perspectives in the SSR support programme, an EEAS Official, a member of EUAM mission, had this to say: 'There is no doubt that the EU is a very strong, motivational force in Ukraine. But whether the mission as such, the activities carried out by the mission make a difference, I am doubtful' (Interviewee D, 2017). At its best, the EU's involvement in Ukraine has opened the space for considering WPS issues, including responses to gendered violence and the participation of women. However, the active implementation of the WPS agenda by or through the EU is very limited. This underscores FI claims about the production and re-production of dominant masculinities within the processes of change in security institutions (cf. Kronsell 2005Kronsell , p. 1033. Moreover, in considering how the processes of (non) inclusion map on the True and Parisi's mainstreaming typologies, the EU's impact in terms of the typology is limited. We would consider, however, Zguladze's drive to include women in previous male dominated spaces provides a tacit support for the first mode within the typology, gender-as-sameness.

Afghanistan
The European Union Police Mission in Afghanistan (EUPOL Afghanistan) is the EU's contribution to rebuilding Afghanistan in the aftermath of the 2001 American led invasion. It was established in 2007 as a transformation of previous German and Italian bilateral programmes. The EU's role was parallel to those of other bilateral and multilateral actors: the US was leading in military reform, while Japan led in disarmament, demobilization and reintegration of ex-combatants (DDR). NATO (including some EU Member States and the US) also took over some of the training for the Afghan National Security Forces (ISSAT 2012). The EU was thus not a lone actor in the realm of SSR or the promotion of gender equality in this context. However, its role in Afghanistan is comparatively less known.
The goal of EUPOL Afghanistan is to build a civilian police service that operates within an improved rule of law framework and with respect for human rights. It aimed to do this through three priority areas: support the institutional reform of the Ministry of the Interior; professionalize the national police; and connect the 'national police to the wider justice system' (European Court of Auditors 2015, p. 12).
At the time the EUPOL mission was established, the international community consensus was that achieving gender equality, through more active participation of women in political life, was a priority. There was a lack of gender representation in its security services with no female personnel, which created practical challenges for dealing with genderspecific crimes. For example, performing a body search on an Afghan woman in a burkah was out of question for any male officer and yet security was jeopardized if men disguised themselves as women to smuggle weapons or bombs (Bastick 2008, p. 15). Thus, for the EU in Afghanistan, gender was easy to inscribe into the design and conceptualization of the mission: 'One of the major efforts where gender really has been in the fore has been EUPOL Afghanistan', underlines the former Director of Civilian Crisis Management, Leinonen (Interviewee B, 2017).
While other actors in Afghanistan, like the US appeared to reinforce 'a 'militarization' of some aspects of civilian efforts, particularly police reform' (Peral, 2009, p. 332f), it was the goal of EUPOL to support the transformation of the Afghan National Police (ANP) towards a civilian force. This helped increase awareness and use of human rights as a guiding ethos for criminal justice resolution; however, the ANP still remains highly militarized.
Nevertheless, EUPOL's support for gender and human rights concerns made some important gains. For instance, EUPOL helped the Afghan government to establish Family Response Units within the ANP, who deal specifically with domestic violence crimes where women are the victims, crimes against women and children, and crimes committed by women. Further, ANP officers received training from EU personnel on violence against women and children including sexual violence. Thus, the push to include women in the ANP is not just about gender balance, but also about gender justice in society. The former Head of the Mission confirms this: 'I am certain of this [gender justice]. Especially women police can now demand their rights which was not possible before' (Interviewee F, Stjärnvall, 2017).
Through this EU mission, a strategy was developed on the sustainable recruitment of, career development opportunities for, and retention rate of female police officers. A former official within the Civilian Planning and Conduct Capability explains: 'We started to be very ambitious about numbers, but then we realised that it was not the right angle, because if the conditions were not there for women police officers to join the police forces, they would not join, so first we had to work on the conditions, and then on the numbers. Provide facilities, in some cases you had to provide women-only classes/ classrooms, because it is easier for them to engage, team and trainers as well, so it has a lot of practical implications' (Interviewee C).
Yet, despite the strategy for change within the police, an audit revealed that some trainers skipped subjects like 'corruption, human rights and gender' because they were deemed to either be 'unnecessary' or 'sensitive' (European Court of Auditors 2015, p. 25). An audit conducted 7 years after the mission further revealed that only 2 per cent of the new police force were women (European Court of Auditors 2015).
Individual leadership within the EU team, particularly Pia Stjärnvall, and several highranking officials within the Afghanistan governmental architecture, was instrumental to some of the changes observed. Stjärnvall, prior to being head of mission, was instrumental in Finland's support to Afghanistan in developing Afghanistan's National Action Plan (NAP) to implement UNSCR 1325, and was thus familiar with the necessity of implementing the WPS agenda within SSR commitments. As Head of the mission, she initiated a Female Police Officers Conference in collaboration with Brigadier General Hekmat Shahi Rasooli, the Head of Human Rights, Women and Children Affairs within the Ministry of Interior (MoI), Ms. Fawzia Koofi, a Member of Afghan Parliament, and Ms. Sima Samar, the Head of Afghanistan Independent Human Rights Commission (EUPOL Afghanistan 2016).
While in Afghanistan, Stjärnvall strived for gender mainstreaming by building partnerships: 'We worked closely with the MOI [Ministry of Interior in Afghanistan] and also with some civil society organizations. We also gathered political support for different areas, especially for community policing and for the female police' (Interviewee F, Stjärnvall, 2017). Moreover, EUPOL Afghanistan directly employed a local Gender and Human Rights expert as a direct adviser to Brigadier General Hekmat Shahi Rasooli. This Afghan experience reinforces the theory that femocrats are essential to ensuring gender perspectives are kept on an institution's agenda.
Afghanistan makes the FI case for the ways in which the complexities of oppression interact with gender. Arguably, the EU considered the insecurities brought on by the conflict itself, but also the prevailing sources of gendered inequalities within Afghanistan in how it approached gender inclusivity in its support to SSR. At the same time, the recruitment of female police officers brought women into a space typically inhabited by men. To an extent then, True and Parisi's first and third models within the typology were enacted simultaneously. While mission planning considered the intersectionality of oppressions, implementation is in line with True and Parisi's gender-as-difference mode.
However, gender-based violence is still pervasive. Distorted and harmful traditional customs and practices have deep roots and the persistence of weak governance and rule of law has made transformation slow. Indeed, both men and women remain insecure, but women and girls are disproportionately affected by multiple violence. Institutionally then, EU SSR does not achieve significant change.

Conclusion
In this contribution to debates on the EU's inclusion of gender perspectives in its external security practices, we adopted a feminist institutionalist approach to understand and assess the execution of EU's SSR programmes. Specifically, we sought to understand the opportunities and constraints of gender inclusion into the EU's SSR programmes. This approach assumes that most formal and informal institutions of the security sector reify patriarchal and male-dominated hierarchies. Further, we identified how individuals or femocrats played an influential part in including gender perspectives within SSR programmes. Femocrats acted as agents of change and promoted gender equality within security sector institutions by challenging the status quo to different degrees. While studies have often excluded gender experts within the EU, the findings, like those of Thompson and Prügl (2015), confirm that gender experts are constrained by their institutional environment.
To re-cap: Afghanistan and Ukraine present a hard and easy case, respectively, of gender mainstreaming within EU's support to SSR. They helped us test the four theoretical assumptions proposed by FI: that institutions reproduce gender differences and hierarchies found in society; that gendered power dynamics determine the decision-making structures; that 'gendered logic of appropriateness' informs the practices of formal institutions; and that change and innovation is fostered by agents of change such as femocrats. Across the two cases of EU's support to SSR, we found that the inclusion of gender perspectives in each case was dependent on the security sector institutional contexts including societal constraints and the role of certain individuals. From these, we observed the conception of gender within the mainstreaming typology developed by True and Parisi (2013). Thus, our analysis tested and confirmed the assumptions of FI theory. Moreover, our application allows for a better understanding of the processes of gender mainstreaming within the EU's security architecture using cases that represent 'high' and 'very high', as well as 'low', restrictions on women's and girls' access of rights.
The EU did not achieve parity in representation within the Ukraine mission. Moreover, the inclusion of a gender adviser was secondary to the initial deployment of the mission, thus underscoring the implications of institutional design and also path dependency for including gender inclusive norms. As feminist institutionalists have claimed, however, the informal practices and advocacy of femocrats do matter; they serve as opportunities for change within institutions. Change within the Ukrainian police was achieved not through formal frameworks but the tenacity of First Deputy Minister of Internal Affairs of Ukraine, Ekaterine Zguladze. In her, position, she was able to affect a gender inclusive reform drawing on the existing legal provisions and the opportunities for change in the post-conflict environment.
In Afghanistan, societal constraints continue to impact on the implementation of positive gender norms and this serves as a constraint on institutional change. Yet, the EU can be said to have been successful in terms of including gender at the point of mission design. Moreover, the role of a femocrat and local buy-in beyond elites including civil society supported the progress made on gender issues. From the perspective of the EU, while acknowledging the challenge that Afghanistan poses, positive change requires the interplay between early inclusion in institutional design, political willingness, and the advocacy of femocrats.
These case studies expose the difficulties of enacting change, but also possibilities for overcoming these difficulties. A feminist institutionalist approach gives us ways of understanding how internal or external actors facilitate and promote gender equality in their support for SSR programmes, even in those cases where the societal context conditions are not necessarily in support of gender mainstreaming.
In re-considering True and Parisi's typology, we find that for the most part, the EU enacts a double understanding of gender in its mainstreaming approaches across the cases, perhaps unsurprisingly. While focusing on its gender representation within the mission itself, the EU favours the gender-as-sameness modelpromoting the idea that neutrality between men and women will achieve equality. For instance, the EU does not consider that difficult living and working conditions in the missions do not always favour women (Interviewee C). This is evidence that the EU's security institutions themselves have not implemented gender mainstreaming. Moreover, it betrays a tendency to focus more on outputs by simply emphasizing the recruitment to security institutions rather than paying attention to their material conditions within those institutions and especially their abilities to influence processes. This is not unique to SSR programmes but emblematic of the current structure of the EU's security architecture.
When the EU promotes representation without problematizing the hegemonic masculinities within the security architecture, challenges are bound to persist. It is thus unsurprising that when CSDP planners request gender advisers and gender-balanced teams from Member States, these are not as forthcoming, and consequently the missions remain gender imbalanced. The everyday practices that reinforce this situation and the lack of change underscore the power of informal institutions.
Externally, the EU's emphasizes shifts. For instance, while planning for Afghanistan appears to consider the intersectional oppressions in programme design (gender-as-intersectional), the achievements appear to reify the gender-as-difference mode within the typology. The societal context within which women are included in the ANP emphasizes the difference in the roles of men and women despite their perceived equal value. As in the Ukrainian case, the lack of political willingness on the part of the EU impacts on its ability to meet its WPS obligations.
The findings of the article are highly significant as they link scholarship on EU foreign and security policy with literatures on gender and peacebuilding, SSR and the WPS agenda. By adapting a FI theoretical framework for studying the EU's support to SSR, we are able to gain novel insights into the external role of the EU as a gender and security actor. We also show how contradictions in the EU's gender mainstreaming models emerge from its policy practices in the security sphere. Through this analysis, we also contribute to those debates on the role of specific types of agents in internalizing new normative frameworks notwithstanding difficult constraints that otherwise reproduce the status quo.
Thinking ahead, this research leaves room for future research to focus not only on the impact of the EU on gender agendas but also to address the gendered power dynamics between donors like the EU and third countries. Moreover, while this analysis focuses on the role of elites, the impact of civil society, including women's groups, would reveal fascinating insights.