Participatory governance or deliberative disjuncture? Exploring the state–civil society policy nexus in the gender mainstreaming programmes of seven Middle Eastern states 2005–2015

To better understand why Middle Eastern states continue to languish at the bottom of world rankings on gender equality, this study presents critical discourse analysis of state and civil society organizations’ implementation of the Participative Democratic Model of gender mainstreaming. A requirement of the 1995 United Nations Beijing Declaration and Platform for Action, the Participative Democratic Model entails state–civil society engagement to embed gender equality concerns in every stage of the policy process. It is in this context that the original contribution of the article is twofold. In methodological terms, it is argued that contemporary analysis of mainstreaming needs to examine the formative phase of policy implementation and the discourse between state elites and civil society organizations. This is integral to effective agenda-setting and coordinated action—and thus to securing successful gender-equality outcomes. In empirical terms, the study findings show how presently, across the Middle East, there are marked contrasts in state and civil society policy framing and issue prioritization. The resulting disjuncture is a hitherto under-examined pathology preventing the realization of the normative vision of gender equality in the region.

P u blis h e r s p a g e : h t t p:// dx. doi.o r g/ 1 0. 1 0 8 0/ 1 3 5 3 0 1 9 4 . 2 0 1 6. 1 1 8 5 9 3 6 < h t t p:// dx. doi.o r g/ 1 0. 1 0 8 0/ 1 3 5 3 0 1 9 4. 2 0 1 6. 1 1 8 5 9 3 6 > Pl e a s e n o t e: C h a n g e s m a d e a s a r e s ul t of p u blis hi n g p r o c e s s e s s u c h a s c o py-e di ti n g, fo r m a t ti n g a n d Thi s v e r sio n is b ei n g m a d e a v ail a bl e in a c c o r d a n c e wit h p u blis h e r p olici e s. S e e h t t p://o r c a . cf. a c. u k/ p olici e s. h t ml fo r u s a g e p olici e s. Co py ri g h t a n d m o r al ri g h t s fo r p u blic a tio n s m a d e a v ail a bl e in ORCA a r e r e t ai n e d by t h e c o py ri g h t h ol d e r s .

Introduction
This article explores state and civil society organizations' (CSOs) policy discourse on gender mainstreaming in the Middle East. The aim is critical analysis to better understand the policy process and how this shapes progress in mainstreaming; the principal international approach to ending gender oppression and securing equality for women as originally set out in the 1995 United Nations Beijing Declaration and Platform for Action (BDPfA). Current attention centres on Bahrain, Egypt, Iran, Israel, Palestine, Turkey and the United Arab Emirates (UAE)states reflecting the diversity of the Middle East in terms of political economy and geography. These constitute an appropriate locus of enquiry because the region continues to languish at the bottom of world rankings on gender equality. Thus, just over a decade ago it was observed that 'women … are on average more disadvantaged economically, politically, and socially than are women in other regions, and certainly more so than in regions with similar income levels or at similar stages of economic development' . 1 More than 10 years on, little has changed. For example, the Global Gender Gap Report 2014 concludes: 'last in order of average scores is the Middle East … with its highest-scoring country placing below the regional averages for all the other five regions' of the globe. 2 The following empirical analysis suggests that it is the nature of the policy process (specifically a disjuncture in state and civil society issue framing and prioritization) that is both symptomatic of and an explanation for the limited progress in the region. This is because the Beijing Declaration requires what is known as the Participative Democratic Model (PDM) of mainstreaming. In other words, effective state-civil society engagement, whereby policy elites listen and respond to the issues and priorities ('situated knowledge') of CSOs. yet the present findings show that instead of the PDM, an instrumentalist, state-led, 'expert-bureaucratic' approach to mainstreaming prevails. The associated disjuncture between governments and CSOs is a hitherto under-examined pathology frustrating the realization of the normative vision of gender equality set out in UN policy. Thus, the overall argument here is that contemporary analysis of mainstreaming in the Middle East (and beyond) needs to examine practice from a wider governance perspective. Specifically, greater attention should be placed on the formative phase of policy implementation and the discourse between state elites and CSOs. This matters because discourse is integral to effective agenda-setting and coordinated action-and, in consequence, the future success of efforts to secure gender equality.
To explore these matters in detail, the remainder of the article is structured thus: following an overview of the literature on gender mainstreaming and civil society, relevant social theory is discussed, the policy context outlined and the methodology summarized. Next, an analysis of state discourse is presented, followed by an exploration of CSOs' discourse on the implementation of the BDPfA. The principal findings and their implications as well as policy recommendations and avenues for future research are then discussed.

Gender mainstreaming and civil society
A full discussion of the development of gender mainstreaming is beyond the present purposes. 3 The United Nations defines it as follows: The process of assessing the implications for women and men of any planned action, including legislation, policies or programmes, in all areas and at all levels. It is a strategy for making women's as well as men's concerns and experiences an integral dimension of the design, implementation, monitoring and evaluation of policies and programmes in all political, economic and societal spheres so that women and men benefit equally and inequality is not perpetuated. The ultimate goal is to achieve gender equality. 4 The rapid international adoption of gender mainstreaming (it has been subscribed to by over 180 states) owes much to its holistic and pro-active nature (notably, through the application of key principles, tools and techniques to all stages of the policy process, as well as its democratic potential). 5 As noted, in the latter regard, the PDM places emphasis on involving those targeted by mainstreaming initiatives in both the design and delivery of policy. The 'policy reach' of the Beijing Declaration is wide. It requires all governments to adopt the PDM. This obliges them to involve those targeted by mainstreaming initiatives both in the design and delivery of policy. 6 It asserts that 'civil society cooperation with Governments [is] important to the effective implementation and follow-up of the Platform for Action' . 7 As Petra Debusscher and Anna Van der Vleuten observe, participative 'mainstreaming is constructed, articulated and transformed through discourse. Policy-makers carry the responsibility to push […] equality further by involving civil society and individual activists promoting […] equality' . 8 'Participation' and 'engagement' here can be defined as the full range of formal and informal means employed by individuals and groups to influence the aims, scope, design and implementation of public policy. These include networking, protest, boycott, lobbying and campaigning. It is in this context that sought-after effective practice in gender mainstreaming is dependent upon a number of key factors-including academic research and knowledge-exchange to aid implementation. This is set out in original work by Virginia Vargas and Saskia Wieringa (1998) and Alison Woodward (2004), wherein these authors allude to a 'triangles of empowerment' linking feminist bureaucrats, academics and voices in the women's movement in order to develop gender mainstreaming. 9 Aside from state policies and academic input, the third constituent of the mainstreaming triangle is, as noted, civil society. The term 'civil society' is denotes the realm of dialogue and human relations that is connected to, but distinct from, the state, markets and personal or familial sphere. 10 It is defined as associational activities involving the family, non-governmental organizations (NGOs), pressure groups, charities, community groups, social movements and campaigning organizations. 11 It is a social arena and political space that is of pivotal significance to understanding contemporary gender relations because of its potential to challenge the largely male-dominated character of state institutions, act as a source of pluralism and solidarity around norms of equality and rights, 12 and operate as a space for civility, respect and recognition. 13 It thus constitutes a key social and political nexus with the state through which CSOs may advance women's representation and pursue claims-making cognizant of a history of marginalization and oppression. 14 Not only is civil society's role explicit in the BDPfA, it is underpinned by a series of UN conventions and resolutions. 15 Thus the over-riding objective is to secure 'the participation and contribution of all actors of civil society, particularly women's groups and networks and other non-governmental organizations and community-based organizations, with full respect for their autonomy' . 16 In response, governments across the Middle East have espoused such inter-sectoral engagement. For example: 'the national model for the integration of women's needs in development … is intended to implement the model in partnership between the legislative and executive branches, [… and] civil society'; 17 our approach promotes 'the participation of citizens in the drafting of resolutions and the positive cooperation between the civil society and the government institutions'; 18 and 'government and civil society hold ongoing dialogues in order to ensure full engagement in the different issues, decisions and policies' . 19

Social theory
In conceptual terms, the state-civil society co-working required by the PDM is consistent with four related strands of social theory. First, complementarity theory 20 emphasizes how politicians attempt to cope with complexity by using civil society networks to increase involvement in policy formulation, thus not only strengthening input legitimacy but also policy efficacy through the pursuit of shared goals. The term 'input legitimacy' is allied to pluralist thinking 21 and the idea that, in contrast to elitist modes of governing (where the policies of a political elite are imposed upon the people), 22 decision-making is shared amongst a broad range of groups drawn from across society. In democratic terms it is more legitimate because it reflects the wishes and priorities of a broader, more representative group of the electorate.
In turn, this is echoed in the deliberative democracy paradigm. This is concerned with the shaping of public policy through the participation of exogenous interests. As Jean Cohen explains: 'not simply a form of politics, democracy, on the deliberative view, is a framework of social and institutional conditions that facilitates free discussion among equal citizens-by providing favourable conditions for participation, association, and expression' . 23 In a similar vein, a third strand of theory is Jürgen habermas' Theory of Communicative Action. 24 This makes a powerful statement that validates the current attention to rights discourse and the deliberative input of civil society in relation to mainstreaming in the Middle East: … horizontal political will-formation … An autonomous basis in civil society, a basis independent of public administration and market-mediated private commerce, is assumed as a precondition for the praxis of civic self-determination. This basis preserves political communication from being swallowed up by the government apparatus or assimilated to market structures. lastly, the fourth strand of social theory is linked to the literature on social movements. This suggests that deliberative policy inputs from civil society are a pre-requisite for effective implementation of gender reforms because of what is termed 'frame-alignment' . In turn, 'framing' refers to the language used by policy actors, effectively a 'schemata of interpretation' . 25 Crucially, as David Snow and Robert Benford note, 'by rendering events or occurrences meaningful, frames function to organize experience and guide action, whether individual or collective' . 26 This is done 'by stressing specific values, facts and other considerations, endowing them with greater apparent relevance to the issue than they might appear to have under an alternative frame' . 27 This means that the probability of successful collaborative policy implementation increases at the point at which the frames of key policy actors-such as government and civil society-are aligned (dubbed 'frame resonance'). Applied to the present study, if the policy visions of government and CSOs converge, and if the same issues are prioritized, then the likelihood of successful gender equality interventions is increased. Conversely, if contrasting agendas are pursued, the available capital (human, social, financial) is dispersed and policy intervention lacks coordination-and conflict, protest and resistance may dominate state-civil society relations.
All four strands underpin the present analysis and underline how the PDM may further gender equality in the region. In this way they hold the potential to contribute to what Augustus Norton describes as 'the "rediscovery" of the political uses to which such forms of "informal" association can be put… [This may] contribute […] to a rethinking of the potential of civil society in the Middle East' . 28

The Middle Eastern context
Recent analysis has concluded that 'discriminatory social norms, laws, and practices are still at the heart of gender inequality in the Arab world' . 29 Whilst in Israel some argue that 'at the constitutional level, religious values have pre-empted the introduction of an explicit right to equality for women' . 30 Conflict, religious divisions and state-building exert powerful influences on gender relations in the territories making up the region. 31 For example, as Ruth halperin-Kaddari and yaacov yadgar cogently note: Israel's continuing, violent conflict with its Arab neighbours, is of highly influential relevance to the issue of gender relations.
[It] has overshadowed most other civil and social issues, rendering them 'secondary' to the primary concern of securing the safe existence of the state. This has pushed such pressing issues as gender equality and women's rights aside, thus allowing for the perpetuation of discriminatory, sometimes rather repressive treatment of women. 32 In attempting to understand the influence of religion on gender relations in the region, lila Abu-lughod sagely warns against application of a simplistic binary: We also need to think harder about the liberal construction of the opposition between religion and [women's] rights … These are areas that a cursory glance would reveal to be compromised for most human beings on the planet for reasons that are surely more related to structural inequalities of wealth and power and to decisions about priorities than to religion [including] life and health, bodily integrity, employment, mobility and assembly, participation and speech, property and civil capacity …. 33 Thus the relationship between religion, cultural norms and women's rights is a contested one. For example, helen Rizzo et al. observe: In non-Arab Muslim countries, there were higher levels of support for women's rights, and those who supported gender equality were significantly more likely to support democracy. The reverse was true in the Arab Muslim countries. We argue that for a complete and unbiased form of democracy to emerge in the Arab Middle East, a rule of law that would protect gender equality, minority rights and citizen inclusion would need to be instituted. 34 however, others such as Saidat Ilo and Richard Seltzer offer a contrary position: 'whether countries with large Muslim populations will embrace gender equality [? …] findings suggest that across the seven nations surveyed, broad majorities support gender equality but women are generally more likely to endorse equality than men' . 35 The extant literature on the region underlines the way civil society acts as an arena where gender conflict plays out. For example, in the case of Palestine, as Amal Jamal notes, the: centrality of struggles in civil society to gain further access to the state and state power… the women's movement in Palestine has gone beyond a simple struggle for national liberation. … it targets the new political reality developed after the constitution of the Palestinian Authority. It seeks to find its way in the new institutional structure in order to influence policy-making. Most of its activity is directed to the realm of legislation, where the character of Palestinian society and polity will be determined in the near future. 36 yet, he continues, this has led to divisions: 'these efforts have split the women's movement regarding the appropriate strategy to constitute women as equal citizens' . 37 however, as further research by Daphna Golan and Zvila Orr underlines, civil society may also provide the political space to bridge religious divides and promote women's rights. 38 This is part of what Seyla Benhabib terms 'democratic iterations' . In other words, 'complex processes of public argument, deliberation, and exchange, through which universalist rights claims and principles are contested and contextualized, invoked and revoked, posited and positioned, throughout legal and political institutions, as well as in the associations of civil society' . 39

Methodology
A broad and rich social science literature attests to the merits of comparative, regional research on the Middle East and beyond. 40 Compared with single case studies, as Paul Pennings and hans Keman note, 41 the advantages of the approach are manifold: 'both internal and external validity are considered to be enhanced … an important and crucial step in the use and application of the comparative approach is the issue of concept formation which can travel across time, situations or societies' . In this case, as noted, the concept being explored is the PDM of mainstreaming. The present study is therefore the latest in a series of comparative analyses of gender mainstreaming. 42 yet, hitherto, these have overlooked the Middle East. As such, the current study responds to the earlier call made by Sylvia Walby. In making the case for comparative work, she noted that comparison across polities is essential in order to understand: The different forms that gender mainstreaming takes, in different countries and different policy domains, in order to push forward the theoretical debates. Gender mainstreaming often draws on transnational processes, involving transnational networks and agencies and transformations of the discourse of universal human rights, challenging the traditional focus on national processes. These developments are facilitated by the rise of global processes and institutions, such as the UN. 43 It should be acknowledged that comparative analysis does present challenges. Not least, in combining an overall, regional perspective with a sensitivity to the characteristics and issues in the individual states under study. In response, this study discusses issue-salience and policy framing on mainstreaming at both an aggregate regional level and at an individual country level. This allows generalization in terms of variance, framing and issue salience across the region-and understanding of 'local' issues. Throughout, discussion of the use of frames and issue-salience allows for synthesis of the study findings with the extant literature on issues such as gender discrimination in the labour market and gender-based violence. Thus the current methodological approach responds to Benoit Rihoux and Charles Ragin's call for: configurational comparative analysis [… that] deals with a limited number of complex cases in a configurational way. This means that each individual case is considered as a complex combination of properties [thus…] the researcher engages in a dialogue between the cases and relevant theory. 44 Thus, in ecological terms, the current methodology aimed at the regional level in the Middle East is consistent with the aims and ethos of the rich intellectual tradition of 'regional'/area studies. 45 In is in this context that this study applies qualitative and quantitative techniques to critical discourse analysis. In doing so it draws upon the concepts of policy 'framing' (or the use of language) and issue-salience (or level of attention, prioritization of a given topic). As W. Douglas Creed explains, frames can be viewed as 'a necessary property of a text-where text is broadly conceived to include discourses, patterned behaviour, and systems of meaning, policy logics, constitutional principles, and deep cultural narratives' . 46 In the present study, frames in the BDPfA texts were coded twice. Once using an inductive coding schemata 47 based on key frames taken from the BDPfA (including 'equality' , 'participation' , 'empowerment' , 'rights' , 'discrimination/oppression' , 'partnership/cooperation' and 'mainstreaming'), and again according to policy issue (health, economic status, raising children/family life, education, trafficking/prostitution, gender-based violence, genital mutilation, peace/conflict resolution, property rights/inheritance). In addition, the principal frames in the discourse were further analysed to identify tropes. As Frank Fischer and John Forrester state, these are integral to the formative phase of policy and are cross-cutting 'figures of speech and argument that give persuasive power to larger narratives [including frames] of which they are part' 48 (e.g. tropes in the case of the 'equality' frame include 'awareness raising' , 'effective training' , 'robust monitoring arrangements' and so on).
Frame use was quantified by drawing upon the notion of 'issue-salience' . This measures the level of attention to a given topic or frame amongst competing issues and agendas in the discourse. Issue-salience is determined by content analysis-or, the frequency of key words, ideas or meanings in policy documents. 49 This was done by adapting a procedure derived from electoral studies, whereby texts are divided into 'quasi-sentences' (or 'an argument which is the verbal expression of one political idea or issue'). 50 Dividing sentences in this manner controls for long sentences that contain multiple policy ideas.
To operationalize the mixed methodology, electronic versions of the policy documents were analysed using appropriate software. In terms of periodization, the focus is on the decade up to and including Beijing +20 (the latter being the current UN progress review 44 Benoit rihoux and Charles ragin, Configurational Comparative Methods: Qualitative Comparative Analysis (QCA) and Related Techniques (Applied Social Research Methods) (london: sage), p. 183 (2008). 45 Walter isard and iwan J. azis, Methods of Interregional and Regional Analysis (Regional Science Studies) (london: routledge) (1998); and alan Wilson, Complex Spatial Systems: The Modelling Foundations of Urban and Regional Analysis marking two decades since the original Declaration). Official UN translations were used where the original document was in a language other than English. To increase reliability, both phases of coding (i.e. frames and policy areas) were repeated by a research assistant. This revealed a limited number of discrepancies. In total, 15 incidences were identified (under one per cent) and were resolved through discussion between coders.
The sampling frame was as follows. State discourse analysis consisted of content analysis of the seven state reports submitted to the UN Beijing +20 reviews (2014/15) as well as discourse analysis of a stratified random sample of 100 CSOs' reports submitted to the UN Commission on the Status of Women follow-up to the Fourth World Conference on Women. This sample was constructed to reflect population size, geographical distribution and yearon-year changes over a decade (by sampling from CSO reports submitted in each year). In definitional terms, CSOs (alternatively known as NGOs) constitute a diverse range of bodies. Their shared feature is their not-for-profit or charity status. They are funded by member subscriptions, gifts and grant aid. They vary from small, local bodies comprised of a few individuals with limited resources to large international charitable bodies. 51 Their organizational remit also varies considerably. Some are concerned with specific policy areas (such as education or health), specific issues (such as domestic abuse) or particular groups and identities (e.g. older people, single mothers, widows, Muslims, and lesbians).

State and civil society organizations' discourse across the seven states
Analysis of the discourse in governments' Beijing +20 National Reports in seven states drawn from across the Middle East reveals that, when aggregated, the four leading frames were: 'rights' (27.6 per cent), 'equality' (25 per cent), 'participation' (19.7 per cent) and 'empowerment' (12.7 per cent) (Figure 1). From a surface reading this might appear to auger well. yet closer attention reveals a number of issues and shortcomings. For example, in the case of rights, governments and CSOs give contrasting levels of attention to enforceable legal duties (or 'hard rights') compared with 'soft' (or discursive) rights. It is governments' Beijing +20 reports that make the majority of references to 'soft' rights. This is a concerning feature of the discourse because, as existing work highlights, they lack the necessary enforceability mechanisms in order for individuals to secure redress in the face of gender oppression. 52 Without this they are little more than rhetorical devices. Examples of such discourse include: we 'support the rights of working women and help them get the necessary training and promotion opportunities'; 53 'Women as half of the society members can play a major role in determination of their own destiny through participation in … decisions on women's rights and their own empowerment and assist to the realization of gender policies'; 54 and 'promoting community awareness of all the rights and obligations of women in order to provide her with opportunities for excellence' . 55 As leslie london and helen Schneider underline, a 'hard' rights framework is imperative instead of soft rights rhetoric, for it 'creates the space for civil society action to hold public officials accountable … enabling civil society mobilization, reinforcing community agency' . 56 Analysis of governments' discourse related to the second frame, 'equality' , reveals contrasting use and understanding of the concept by administrations across the region. When references to equality are disaggregated by policy topic, the lead areas were: raising children/ family life (27.2 per cent), education (23.8 per cent), health (19.5 per cent) and tackling domestic violence/abuse (13.3 per cent) ( Figure 2). however, a basic concern arising from the state discourse on 'equality' relates to its conceptual underpinnings. Specifically, there is ample evidence of governments' over-reliance on an outdated Aristotelian equal treatment approach to gender. 57 For example, our approach is founded on 'equal treatment to men' 58 and we seek to ensure 'the equal treatment of the children of women citizens without discrimination in education, health or employment' . 59 The problem here is that equal treatment can compound rather than diminish discrimination. 60 A further concern is governments' over-reliance on positive action measures. For example, 'affirmative action shall be taken in favour of women as the new employment opportunities are offered'; 61 and 'these 56 leslie london and helen schneider, 'Globalisation and health inequalities: Can a human rights Paradigm Create space for Civil society action?' , Social Science and Medicine, 74(1) (2012), p. 6. 57 in other words, a basic approach to promoting equality based on treating everyone the same, without taking account of (pre-)existing inequalities or patterns of discrimination. this can actually compound inequalities rather than securing equal outcomes. recommendations advise the universities to enact affirmative action for women in the sciences' . 62 Again, this is troubling because, as an extensive literature attests, 63 whilst such measures may secure short-term gains, they fail to address the underlying structural and cultural causes of inequality. At a more fundamental conceptual level, the hybridized approach to equality in governments' +20 discourse across the region can also be seen as indicative of a lack of clarity and reflective of a non-strategic approach to tackling gender discrimination. In short, it evidences practice that is in conflict with the ethos of mainstreaming. Thus, as Mary Daly observes of such 'hybrid' cases of gender mainstreaming, they: … facilitate a break between the introduction of gender mainstreaming and addressing gender as structural inequality. In other words, gender mainstreaming is introduced in the name of updating existing policy approaches to women rather than as the author of a transformative vision that recognizes gender as a societally embedded and structural problem. 64 A further issue with the equality frame in governments' + 20 discourse is the limited attention paid to 'intersectionality' . This refers to the intersection of two or more axes of inequality or

64
'hybridized' versions of mainstreaming often incorporate equal treatment and positive action. here daly is saying practice is a combination-or hybrid-of elements from two approaches. thus, in contrast to an 'integrated approach' (as seen in sweden) where responsibility for mainstreaming lies with all actors and institutions in a policy, hybridized cases combine elements of: 'limited transversiality' (as seen in Belgium and ireland), meaning mainstreaming signals little more than government implementing a gender equality plan-in these contexts mainstreaming neither covers all policy actor nor institutions; and settings where mainstreaming is fragile and at an early stage and limited to a few policy areas and/or institutions (as seen in france and the united Kingdom). for a full discussion, see Mary daly (2005)  discrimination (e.g. gender and disability; gender and ethnicity, etc.). 65 The BDPfA is explicit in the need for this: governments [must] affirm their determination to intensify efforts to ensure equal enjoyment of all human rights and fundamental freedoms for all women and girls who face multiple barriers to their empowerment and advancement because of such factors as their race, age, language, ethnicity, culture, religion or disability or because they are indigenous people. 66 yet, strikingly, there is just a single reference to intersectional issues in the state reports: … confronting meaningful possible lines of division within the feminist movement, such as issues of peace and feminism, racism within the movement, and intersectionality. In this regard, the movement has brought new understandings and new populations into the arena, including women from lower socio-economic strata, women with disabilities, and other marginalized women. 67 The dearth of attention to intersectionality in the states' Beijing discourse is at once reductive and essentializing-and promulgates the idea that women across the region face identical modes of discrimination and oppression when in fact such processes are inflected by simultaneous and multiple identities. 68 In contrast to the government discourse, analysis of the second data-set-namely, CSOs' reports to the UN 2005-2015-reveals that the four leading frames on the implementation of the BDPfA were 'equality' (33.2 per cent), 'rights' (28.3 per cent), 'discrimination' (13.7 per cent) and 'empowerment' (9.9 per cent) ( Figure 1). The lead policy topics in the CSO discourse were as follows: • gender and economic inequality/poverty (17.2 per cent; e.g. 'women bear the consequences and become economic mainstays, carers and political negotiators. Furthermore, they face homelessness and forced expulsion, widowhood, hunger and poverty and are denied their distinct female identity and psychological needs'); 69 • education (17.1 per cent; e.g. 'education to eliminate illiteracy is the most effective means of propelling girls out of poverty and helping them to realize their full potential'); 70 • gender equality and peace/conflict resolution (14 per cent; e.g. 'the lack of awareness about women's education and training, especially in relation to refugee and displaced women and girls in post-conflict areas'); 71 and • property rights (13.6 per cent; e.g. 'ensuring women's land and inheritance rights') ( Figure 2). 72 The significance of the foregoing contrasts in state and civil society priorities forms the subject of the following discussion and analysis. Mind the gap: state-civil society 'disconnect'?
The key finding from the data analysis is the disjuncture between state and CSO discourse (Figures 1 and 2). This matters because it highlights a hitherto under-examined pathology in the policy process in Middle Eastern states. Specifically, the formative, discursive processes shaping contemporary attempts to implement the Beijing Declaration. It also undermines the future capacity of the civil sphere to challenge traditionally male-dominated power structures and transform gender relations. This is because of the resulting negative feedback loop: over time, limited civil society input to public policy-making means it becomes increasingly reflective of state priorities and the programmes of governing elites-rather than responsive to the needs of communities, individuals and NGOs. Overall, such marginalizing of civil society hampers progress towards the normative vision of gender equality set out by the United Nations. Not only do governments and CSOs have different issue prioritization, they also hold contrasting views on what frames are most important. The result is the prevalence of an 'expert-bureaucratic' mode of mainstreaming. 73 This is when governments' preferred policy agenda is imposed in place of the collaborative process predicted by the PDM. As the following discussion outlines, the state-civil society contrasts in the Beijing +20 discourse apply across the region. The extent of the divergence varies between individual countries-thereby underlining how BDPfA implementation is contingent on local political economy and prevailing socio-economic histories and circumstances in each territory.
A prominent example of the divergence between state and civil society discourse is provided by references to the empowerment of women. This can be illustrated with reference to Arthur Turovh himmelman's (1996) work on empowerment theory. 74 Thus, whereas the state reports tend to extol empowerment by encouraging women to participate in pre-determined government-led initiatives (captured by the term 'collaborative betterment'), CSOs' discourse emphasizes 'grass-roots' action whereby women in CSOs themselves set the terms of their empowerment through self-organization and mobilization (styled 'collaborative empowerment' by himmelman). It is the latter mode of empowerment that is consonant with the sought-after PDM. One where accountability is bolstered through recognition that the state alone cannot address gender inequality.
In the latter regard, the state discourse is typified by 'The National Action Plan on The Empowerment of Rural Women (2012-2016) was prepared by The Ministry of Food, Agriculture and Animal husbandry …' . 75 Whereas examples of the civil society 'collaborative empowerment' discourse include: political and professional associations and labor unions are required to take the necessary measures to guarantee wider women participation. These include set[ting] measurable strategies, and monitoring towards the increase of women participation in general, empowering them, in specific, to assume leadership, executive and managerial positions in the structures of these institutions. the promotion of collective initiatives through civil-society organizations, networks and women's cooperatives, will contribute to the empowerment of women to participate fully in economic life across all sectors and at all levels. It will also lead to the strengthening of rural women and the establishment of more stable and fairer societies ' . 77 The key contrasts in the state and civil society discourse are further illustrated by the emerging patterns in each of the seven countries comprising this study. From an international perspective the first state to be considered here, Bahrain, is a member of the highest-income group of countries. yet it is also one of the worst performing in terms of gender equality. A fact that attests to current failings across a range of domains. According to the World Economic Forum, 78 Bahrain's global rankings (from a total of 142 states) are: economic participation and opportunity 126th, educational attainment 90th, health and survival 132nd, and political empowerment 116th. As the latter indicator suggests, gender inequality and discrimination links to deep-set problems and crises of democracy and governance. These are complex and are far from a solely domestic concern. They relate to wider geo-politics, not least because as Oz hassan notes, 'the transatlantic democracy promotion strategy is complicated by a conflict of interests problem … in Bahrain, transatlantic democracy promotion is itself muted by the strategic interest in containing Iran' . 79 Reflecting the longstanding discrimination and marginalization of women in Bahraini politics, the discourse on the implementation of the Beijing Declaration not only shows statistically significant differences between state and civil society across policy frames (P < 0.001), 80 it also reveals contrasts in the prioritization of policy areas (P<0.001). 81 This finding shows how government is pressing ahead with actions and priorities that contrast with the demands of CSOs. Notwithstanding official acknowledgement of the 'disconnect' (e.g. 'the challenges encountered in the integration of the needs of Bahraini women in the Government's agenda … will require further cooperation and coordination between relevant sectors in ministries, governmental bodies, private sector institutions and civil society organizations'), 82 such statements offer little explanation as to how the disjuncture will be addressed.
The greatest disparity between state and civil society framing relates to the discourse on women's rights. Just over two-thirds (66.3 per cent) of all CSO quasi-sentences are concerned with greater rights for women compared with just under a quarter (24.4 per cent) of the state discourse (P < 0.001). 83 For example: a legislative gap before the jurisprudence of the Bahraini judiciary… this results in a number of legal ambiguities, because of jurisprudence on behalf of judges, that is different and sometimes contradictory. The absence of this law in Bahrain is considered a stark discrimination against Bahraini women and a prejudice against their human rights' . 84 Further significant disparities between state and civil society relate to the gendered impacts of poverty and economic inequality (15.9 per cent compared with 10.8 per cent) and tackling domestic abuse (17.8 compared with 11.5 per cent). The latter gap resonates with research by Elena Ambrosetti et al., who underline the need for state-CSO engagement in order to tackle gender-based violence: change in gender relations requires a questioning of hegemonic patriarchal norms and family values, influencing women's freedom to choose their spouses, their status in their family and the couple, and their participation in decision-making within the household… there is a need for cooperation between state and women's rights groups in order to promote sustainable prevention policy against gender-based violence. 85 Notably, the discourse of Bahraini CSOs details the nature and extent of the gender-based violence problem: There is a shortage of field surveys, research and systematic statistics that monitor incidents of violence against women especially that the official side does not admit to their presence as a phenomenon but rather insists that they are simple incidents that do not represent a social phenomenon. 86 In the case of Egypt, its global rankings also attest to gender oppression and major failings across a range of domains. Notably, the country is ranked 131st in relation to economic participation and opportunity, whilst for educational attainment it is 109th, health and survival 57th, and political empowerment 134th. 87 Recent political turmoil has compounded long-established patterns and processes of gender oppression. As Maya Morsy notes, the transition government of the Muslim Brotherhood saw 'a backlash against women and their agenda prompted conservative calls for amendments of legislations that were passed during the Mubarak era' . 88 She proceeds to outline 'the importance of women's involvement in transitional processes for a successful democracy in Egypt' . yet cogently notes that comparison of: Egyptian laws both before and after the revolution with the international human rights conventions to which Egypt is party, point at inconsistencies and outright violations, including the treatment of gender-based violence (GBV), the debate around female genital mutilation (FGM) laws, and issues of gender and Sharia in the draft Constitution.
Against this backdrop, the present analysis reveals statistically significant differences in state and CSO discourse (P < 0.001) 89 on the implementation of the Beijing Declaration across policy frames.
The scale of the problem in Egypt is underlined by the principal state-CSO disparity: over a third of CSOs' discourse (35.9 per cent) was concerned with making the basic case for gender equality compared with a fifth of state discourse. In other words, contemporary social attitudes and practices constituting gender relations are so unequal that attention centres on the fundamental case for equality, rather than the means for its achievement. Reflecting the recent turbulent history of the Arab Spring and frustration at the limited progress in women's rights, the greatest state-CSO disjuncture across policy areas (P < 0.001) 90 was CSOs' emphasis on the gendered effects of civil conflict and women's role in peace-making. 91 This accounted for almost a third of quasi-sentences in the CSOs' reports (31.6 per cent) compared with just 2.7 per cent of the state discourse. For example: 'conflict resolution and in post-conflict peace-building … the characteristic role of women … she is one of the big losers in wars and one who really pays for destruction'; 92 and 'inequality is particularly gendered in war and conflict which severely compromises women's right to sustainable development … inequality is intensified during conflict since peacekeeping infrastructure is often destroyed' . 93 A further significant contrast relates to the gendered impact of economic inequality and poverty. This is indicative of a failure on the part of governing elites to fully recognize and address the gendered dimensions of economic inequality. 94 Thus, this frame was the subject of a quarter of CSO discourse overall (25.8 per cent), compared with just 14.9 per cent of state discourse. For example: The government has to design a program to promote economic initiatives for poor rural women to integrate them into local markets and improve their income. Encourage more private investments to create jobs in rural area.
[Furthermore] There are severe constraints faced by rural women in accessing financial resources, therefore there should be a coordinated delivery programs of financial resources through government. 95 As Goli Rezai-Rashti observes, 'the Islamic Republic of Iran has consistently enacted discriminatory policies regarding gender relations since 1979' . 96 Its global rankings attest to this and reveal a plethora of current failings: on economic participation and opportunity the Republic is ranked 139th, educational attainment 104th, health and survival 89th, and political empowerment 135th. 97 Unsurprisingly, a state-civil society 'disconnect' is evident in the policy framing associated with the implementation of the Beijing Declaration (P < 0.001). 98 The greatest difference is CSOs' emphasis on tackling gender-based discrimination and oppression (17.7 per cent of quasi-sentences, compared with just 5.6 in the state discourse). here CSOs highlight the state's failure to meet its basic international obligations. For example: the Islamic Republic of Iran has ratified the Convention… and has incorporated the Strategic Objectives of Beijing Platform for Action… Despite such measures there are inadequacies in all fields and a coordinated action plan is necessary to be implemented by government and NGOs in all levels of society… [and to secure] women's equal access to health, education and employment. It also strives to empower women against gender-based violence, social discrimination and socio-psychological harm. 99 Across policy areas too there are statistically significant differences between government discourse and civil society (P < 0.001). 100 For example, in a country where incidence of domestic abuse of women is downplayed or denied, 101 CSOs place significantly greater emphasis on this frame than the state (15.3 per cent of quasi-sentences compared with 6.2 per cent). Moreover, as Roksana Bahramitash observes ,102 women in Iran suffer badly from the gendered consequences of economic inequality and insecurity. This is evident in the discourse where it is underplayed in the government's Beijing +20 review (accounting for just 5.0 per cent of quasi-sentences) compared with 19.4 per cent in the CSO discourse. The latter is typified by the following statement: The Unequal share of women from the national income, which itself is the result of discrimination in employment, wages and financial laws, has a great effect on increasing the discrimination and violence against women in Iran… The government propaganda in recent years has concentrated on stressing only a mother's role for women and on persuading them to stay at home. 103 Israel contrasts with other states in the region in that it has made significant gender equality reforms. As Frances Raday observes: in areas of law not directly related to religious values and norms, a strong concept of gender equality has been developed both in legislation and in the courts, combining social accommodation for maternity and parenthood, with equal opportunity guarantees for women's participation in the labour force and the military, with affirmative action in public sphere economic activities and with protection against sexual violence as an inherent part of women's right to equality and human dignity. 104 Accordingly, Israel's overall gender equality ranking is higher than the other states in the region (65/142), yet issue-specific rankings indicate that a series of challenges remain: on economic participation and opportunity it is ranked 90th, educational attainment 49th, health and survival 96th, and political empowerment 49th. 105 Despite the country's higher international standing compared with other regional states, its +20 Beijing discourse exhibits statistically significant differences between government and civil society both in policy framing (P < 0.001) 106 and policy areas (P < 0.001). 107 Greatest disparity is evident in relation to women's rights. This constitutes almost a third of CSOs' discourse (31.0 per cent), compared with under a quarter of state discourse (22.7 per cent). Examples of the former include that there is a need to: protect the rights of women and young girls… Israel, ha[s] made significant progress in recent years toward the elimination of human trafficking. however, there is much work still to be done.
[for example, we] urge all nations to make the elimination of the trafficking of women and girls a priority by strengthening anti-trafficking legislation. 108 Allied to this there is a further significant CSO-state disconnect in relation to tackling gender-based discrimination and oppression (12.6 per cent compared with 7.7 per cent). In terms of policy areas, divergence is evident in relation to the gendered dimension to economic inequality in a society where overcoming gender oppression is intimately linked to income. 109 Thus, this frame constituted 21.7 per cent of quasi-sentences in CSOs' discourse compared with just 8.9 per cent in that of the state. Existing work has also highlighted gendered aspects to peace and conflict resolution. In the case of Israel this is to the fore in the +20 Beijing discourse. Matters covered under this frame include women's military refusal and participation in the organized conscientious objection movement. 110 Thus, Israel presents a context where, as Cynthia Cockburn concludes: 'in general, hopes have been dashed… women and the politics of gender-let alone any idea of a transversal politics of "empathy without sameness, shifting without tearing up your roots"-have been largely out of the picture for policy makers' . 111 This is reflected in the statistically significant difference in the level of attention that CSOs place on the 'peace and conflict resolution' frame in their +20 Beijing discourse (15.1 per cent) compared with government (2.9 per cent of quasi-sentences).
Accounts of gender relations in Palestine point to a political context shaped by conflict and crisis. Penny Johnson and Eileen Kuttab put this in historical perspective describing a present in which '"authoritarian populism" tends to disallow democratic politics and participation' . They continue: the seeming absence of women and civil society from the highly unequal and violent confrontations [… contrasts] with the first Palestinian intifada , that occurred in a context of more than a decade of democratic activism and the growth of mass-based organizations. 112 It is against this troubled backdrop that the present analysis reveals statistically significant differences across frames in the +20 Beijing discourse (P < 0.001). 113 Such disparity is most prominent in the case of the basic espousal of gender equality (this accounted for 38.7 per cent of CSOs' quasi-sentences, compared with 22.2 in the state discourse). It is also evident in relation to tackling gender-based discrimination and oppression (15.7 per cent compared with 5.6 per cent). For example: Palestinian women living in towns and villages in the West Bank and in East Jerusalem as well as in Gaza face discrimination in the fields of land, housing, property and water rights as a result of discriminatory laws and practices by the State Party. Many Palestinian women are affected by forced evictions and displacement. 114 Significant differences in CSO and state discourse (P < 0.001) 115 are also evident across policy areas-including the gendered impact of conflict and women's role in peace-making in Palestine. 116 Notably, the current data show how CSOs give this issue greater attention than the official state discourse on the implementation of the Beijing Declaration (17.3 per cent of quasi-sentences compared with just 1.9 per cent). A further key difference in the discourse relates to the gendered impact of economic inequality and poverty (22.7 per cent compared with 15.9 per cent). For example: Palestinian women continue to live under foreign occupation and to live with ongoing violations of their human rights … the occupation is a major obstacle to women's advancement and their involvement in the development of Palestinian society, [there is need for greater] efforts to increase their [women's] role in decision-making with regard to conflict prevention and resolution and to ensure their equal participation and involvement in all efforts for the maintenance and promotion of peace and security. 117 As yeşim Arat observes, in Turkey women's rights are seen by some as 'a means to pursue the goals of secularist modernization. They shape the power struggle over what the role of religion in public life should be and what secularism should entail' . 118 It is a country whose global gender (in)equality rankings attest to a series of failings across a range of domains. On economic participation and opportunity Turkey is ranked 132nd, educational attainment 105th, health and survival 98th, and political empowerment 113th. 119 Analysis of the +20 Beijing discourse shows that the most pronounced of the statistically-significant differences across frames (P < 0.001) 120 relates to CSO's greater emphasis on tackling gender-based discrimination and oppression (31.9 per cent compared with 7.2 per cent). For example: Under the current Government's second term (since 2007), there has been little progress in Turkey in terms of the necessary legal and institutional reforms for gender equality … The increasing conservatism poses a threat to women's ability to enjoy their rights and freedoms. The already acquired legal rights of women are subject to backlash and efforts that aim to eradicate existing discrimination are usually met with resistance. 121 Across policy areas there are further significant differences between the civil society and state discourse (P < 0.001). 122 A key disjuncture relates to the emphasis placed by the state on gender equality in the context of family life/raising children (this accounted for 22.3 per cent of the government discourse compared with 6.8 per cent of the CSO discourse). This reflects ongoing Islamism-secularism debates over preserving traditional family relationships versus extending women's rights. As Ebru Erdem-Akçay notes, at the heart of this is the question as to 'whether those rights should be defined according to traditional customs or the norms accepted by international agreements' . 123 Indeed, the intersection between gender (in)equality and religion receives far greater attention in the civil society discourse compared with that of the state (9.8 per cent as opposed to 1.3 per cent), with many feminist campaigners pressing for a rebalancing of gender and religious rights. Further differences relate to the gendered impact of economic inequality and poverty (19.5 per cent compared with 4.8 per cent), and the trafficking of women and prostitution (21.5 per cent compared with 0.3 per cent (e.g. 'although nearly all identified victims of trafficking in Turkey are women, policies that address human trafficking lack a gender perspective, and fail to analyse of the impact of gender inequality'). 124 As Serra Kirdar observes, 'Emirati women are undergoing a transition as their society, exposed to foreign influences, adapts to changing identities while protecting cultural and religious traditions' . Nevertheless, ' continue to pose a significant barrier to advancement' . 125 Thus, the UAE's global rankings attest to deep-set patterns and processes of gender discrimination and oppression across domains. In relation to economic participation and opportunity the country is ranked 123rd, educational attainment 83rd, health and survival 132nd, and political empowerment 96th. 126 The present data analysis reveals significant differences between state and CSO framing of the implementation of the Beijing Declaration (P < 0.001). 127 Key contrasts exist with regard to the espousal of fundamental gender equality (36.1 per cent of CSO discourse is concerned with this, compared with 14.2 per cent of state discourse), as well as tackling gender-based discrimination and oppression (9.8 per cent compared with 3.4 per cent). In the case of policy areas (P < 0.001). 128 The greatest divide between CSO and state discourse relates to tackling gender-based violence and domestic abuse (15.7 compared with 3.0 per cent). For example: The UAE does not collate or maintain official statistics on the scale of domestic violence, either against Emirati women or domestic workers. Its failure to take action to combat domestic violence, coupled with discriminatory legislation, has had disastrous consequences. UAE criminal laws prohibiting physical or verbal violence do not apply to acts committed within a family home; in fact, the Penal Code gives men a positive legal entitlement to 'discipline' their female family members and children as they see fit, up to and including physical violence. 129 A further clear disparity is on the gendered impact of economic inequality and poverty (31.3 per cent compared with 9.8 per cent). For example, 'Under the UAE's penal code, men are allowed to use violence against female and minor family members. Even in countries where some legal remedies are available for domestic violence victims, the law is mostly unenforced' . 130

Discussion
Complementarity theory emphasizes how politicians attempt to cope with complexity by engaging CSOs in policy formulation, thus not only strengthening input legitimacy but also efficacy through the pursuit of shared goals. Such thinking underpins the PDM of mainstreaming required by the Beijing Declaration. Instead of being an imposed political 'project' , according to this view mainstreaming is essentially democratic and pluralist in nature. The involvement of a broad range of groups and individuals from civil society is viewed as a means to address the longstanding marginalization of women in representative structures and public policy-making. however, notwithstanding political elites' espousal of civil society engagement, this study's critical discourse analysis of contemporary practice across the Middle East paints a different picture. It reveals a significant disjuncture between government and CSOs' envisioning of the implementation of the BDPfA. This is in stark contrast to the notion of complementarity which asserts: 'governance networks when predicated on the basis of deliberative and other democratic practices … engender both a democratic ethos and consensual decision-outcomes that transcend and accommodate partial preferences' . 131 Instead, the present data show that government is prioritizing aspects of policy and framing issues of gender equality in ways that contrast with the discourse of CSOs.
The result is the application of an 'expert bureaucratic' model of mainstreaming across the region. This is principally founded on the input of state policy elites and gender experts forging ahead with their own priorities, as opposed to joint authorship of policy founded on civil society engagement. 132 Whilst this may reflect a legacy of mutual mistrust in state-CSO relations, 133 it is deeply problematic for realization of the Beijing objectives. Not least because, as Jo Beall notes: advancing gender equality demands striking a balance between the essentially political project of ensuring women's social and economic participation and political representation, and the more technical project of institutionalizing or mainstreaming a gender perspective in policy and practice… it is essential that the national machinery for the advancement of gender equality … does not forget its foundation in civil society, and that there is an on-going commitment to 'doing gender' from bottom to top, as well as from top to bottom. 134 knowledge and awareness of the BDPfA amongst CSOs across the region. In general terms, the current data also show significant civil society capacity and resourcefulness to engage in the Beijing +20 review. In terms of representativeness, CSOs in the region are diverse in nature (ranging from small voluntary groups to well-established international charities). In summary, the general picture in the region is broadly comparable with other international contexts. however, the present study does point to a distinctive issue affecting Middle Eastern states: the nature of the policy process and, specifically, governments' political will to engage with CSOs.
To address this requires a shift in the way the BDPfA is implemented across the region. To overcome the power asymmetry between government and civil society, the study findings underpin the following policy recommendations. First, in each state, national representative bodies from civil society need a fuller role in UN monitoring of the BDPfA. Also, in each polity there also needs to be more effective collation of all CSO submissions, with a single synthesis of the key demands presented to government. Third, governments should then be required to detail how they intend to address each of these at the beginning of each five-year review cycle-and, at the end of it, to report on progress against the demands. Crucially, this should be done with reference to measurable targets (rather than general, declaratory statements of intent). lastly, the UN monitoring machinery should be better able to compel states to act on CSO recommendations where implementation is seen to falter.
In addition to the foregoing policy recommendations, the current study also suggests a number of avenues for future research: qualitative analysis of leadership, agenda-setting and management within CSOs in order to better understand how, at an organizational level, they respond to the BDPfA; critical analysis of communication/information flows between government and civil society in each state in order to explore the most effective means of raising awareness (and thus CSO engagement) in gender mainstreaming; analysis of the internal governance procedures of CSOs and how effective (and democratic) they are in engaging with their memberships as they feed into BDPfA reviews; and mixed-methods sectoral analysis across civil society combining survey data and qualitative techniques to better understand CSOs' views on how government could better engage with CSOs in furtherance of the Beijing Declaration.
In summary, whilst contemporary espousal of mainstreaming by governments across the region is undoubtedly positive, the present study raises a number of key, ongoing concerns about its implementation. Analysis of the seven Middle Eastern states shows that mainstreaming practice is contingent on local, country-specific factors (such as conflict, religion and Arab culture). however, at a more fundamental level, progress across the region is also linked to the nature of the policy process itself. here the study findings reveal a pronounced asymmetry in the power of government and civil society. This manifests itself in marked contrasts in policy framing and issue prioritization. The overall effect is state-driven policy delivery. This is significant and constitutes a hitherto under-examined pathology in the formative, discursive processes shaping policy implementation; one that undermines the capacity of the civil sphere to challenge traditionally male-dominated power structures, thereby hampering progress towards the normative vision of gender equality set out in UN policy. Downloaded by [Cardiff University Libraries] at 03:29 13 July 2016