UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Gender mainstreaming strategies in the international development context : why practice has not made… Lukatela, Ana Stephanie 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2014_september_lukatela_ana.pdf [ 2.33MB ]
JSON: 24-1.0167468.json
JSON-LD: 24-1.0167468-ld.json
RDF/XML (Pretty): 24-1.0167468-rdf.xml
RDF/JSON: 24-1.0167468-rdf.json
Turtle: 24-1.0167468-turtle.txt
N-Triples: 24-1.0167468-rdf-ntriples.txt
Original Record: 24-1.0167468-source.json
Full Text

Full Text

  Gender Mainstreaming Strategies in the International Development Context: Why Practice Has Not Made Perfect   by   Ana Stephanie Lukatela   M.A., Simon Fraser University, 2005     A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF   DOCTOR OF PHILOSOPHY   in    The Faculty of Graduate and Postdoctoral Studies   (Political Science)   THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)    May 2014     © Ana Stephanie Lukatela, 2014  ii  Abstract The literature surrounding international organizations and policy cycles has overwhelmingly focused on the dynamics of why policies are adopted to the  detriment of asking why they are or are not translated into implementation. This study asks first, what factors explain the differences in adoption and implementation of gender mainstreaming policies in international development organizations, and second, what these dynamics tell us about the reasons for the persistence of such policies in the face of recognized failure. Research was conducted on case studies of UN agencies in Albania, Bosnia and Herzegovina and Kosovo. The findings argue that policy is better studied as a non-linear process where each stage is influenced by different mechanisms emerging from the policy context. Policy adoption depends on the overarching influence of world culture on the attitudes and behaviors of relevant senior management actors who are part of the world polity. However it is also contingent on the dynamics of bureaucratic politics, which can be either personality or policy driven, and can either block policymaking or be used as a tool to overcome differences. The dynamics of implementation on the other hand rely much more on the norms and values of middle management than research up to now has recognized. However for implementation to happen norms are a necessary but not always sufficient factor; they must be paired with organizational behavior protocols that can support implementation among those who are passive towards the relevant norm and there must be ways to sanction programme managers who oppose it. So the support of senior management may be a necessary condition for policy iii  implementation but it is far from sufficient. This study represents a break from the typical IR approach of using one theoretical argument to explain an entire phenomenon and reasserts the importance of opening up the black box of organizations to examine actors at different levels and their interactions.  Improving the success of mainstreaming policies, such as gender, HIV/AIDS or the environment ultimately depends on recognizing these mechanisms and addressing the previously underestimated role of middle management. iv  Preface  The interviews conducted for this project were approved by the Behavioral Research Ethics Board (BREB) at the University of British Columbia (H09 - 02419) on 19 October 2009. In order to preserve confidentiality and anonymity, in accordance with the preferences indicated by interviewees, several interviews are referenced in the following text without their corresponding dates, names or workplaces. Additional information is provided in Appendix A.v  Table of ContentsTable of Contents Abstract ................................................................................................................. ii Preface ................................................................................................................ iv Preface ................................................................................................................ iv Table of Contents .................................................................................................. v Table of Contents .................................................................................................. v List of Tables ...................................................................................................... vii List of Figures ..................................................................................................... viii List of Abbreviations ............................................................................................. ix Acknowledgements .............................................................................................. xi Dedication ........................................................................................................... xii 1. Introduction ....................................................................................................... 1 Research Problem and Justification .................................................................. 1 Summary of Findings ........................................................................................ 6 Significance and Contributions of the Study ...................................................... 9 Case Selection and Sources of Evidence ....................................................... 13 UN Development Agencies ............................................................................. 15 Outline of Dissertation ..................................................................................... 23 2. Theoretical Framework: Policy Adoption vs. Policy Implementation in International Development Organizations ........................................................... 24 Introduction ..................................................................................................... 24 International Organizations and Policy Processes .......................................... 29 Why are Mainstreaming Policies Different? ................................................. 34 The Gendered Nature of the International System, Organizations and Policy ..................................................................................................................... 40 Gender Mainstreaming .................................................................................... 47 Building a Theory of Mainstreaming within the International Policy Process ... 57 Policy Adoption ............................................................................................ 58 Policy Implementation .................................................................................. 71 Alternative Explanations .................................................................................. 77 Bureaucratic Politics Theory ........................................................................ 77 Organizational Behaviour Theory ................................................................ 81 Conclusion ...................................................................................................... 84 3. Methodology ................................................................................................... 86 Research Design ............................................................................................. 86 Content Analysis .......................................................................................... 88 Semi-Structured Interviews .......................................................................... 89 Participant Observation ............................................................................... 91 Observable Implications .................................................................................. 94 Case Selection ................................................................................................ 97 Albania ......................................................................................................... 99 Bosnia and Herzegovina ............................................................................ 101 Kosovo ....................................................................................................... 103 Conclusion .................................................................................................... 105 vi  4. Policy making: World Polity leads to Policy Adoption when Bureaucratic Competition Lines Up ........................................................................................ 106 Declarations of Senior Management Support ................................................ 107 Lack of Accompanying Background Knowledge............................................ 132 Similarity in Rhetoric and Attitudes ................................................................ 139 Decoupling between Values and Action ........................................................ 144 Bureaucratic Politics ...................................................................................... 151 Personality Driven Politics ......................................................................... 152 Policy Driven Politics ................................................................................. 156 Conclusion .................................................................................................... 162 5. Policy Implementation: Normative Commitments Supported by Standard Operating Procedures ....................................................................................... 163 Middle Management Implementers ............................................................... 165 Normative Commitment ................................................................................. 174 Agency Dynamics .......................................................................................... 183 Size and Mandate ...................................................................................... 184 Joint Gender Theme Groups ..................................................................... 192 Leveraging Standard Operating Procedures ................................................. 195 Evasion of SOPs by Norm Resistors ............................................................. 207 Staffing Issues ............................................................................................... 220 Conclusion .................................................................................................... 224 6. Implications, Recommendations and Conclusion.......................................... 226 Theoretical Implications for Crosscutting Policy Processes in International Development Organizations .......................................................................... 229 Understand the Policy Process is Not Linear and Context Has Diverse Impacts ...................................................................................................... 229 Understand the Importance of Diverse Actors ........................................... 232 Policy Recommendations for Gender Mainstreaming ................................... 233 Recruitment in International Development Organizations .......................... 235 Accountability Structures for Crosscutting Policies .................................... 244 Bring the Focus back to Operations and Structural Transformation .......... 245 Reconnect Gender Mainstreaming with the Fulfillment of Human Rights .. 246 Conclusion .................................................................................................... 253 Bibliography ...................................................................................................... 260 Appendix A: List of Interviewees ....................................................................... 281 Appendix B: Semi-structured Interview Questionnaire ...................................... 285 Appendix C: Policies listed in Table 2 and Figure 1 .......................................... 289  vii  List of Tables   Table 1: Characteristics of 13 Largest International Development Organizations based on Annual Budget Size in USD .................................................................. 3 Table 2: Trends in Content and Structure of UN Agency Gender Mainstreaming Policies since 1989. ............................................................................................ 17 Table 3: Annual UNDP and UNIFEM/UN Women funding levels in USD (both regular contributions and voluntary contributions) .............................................. 22 viii  List of Figures  Figure 1: Results of Evaluations of UN Agency Gender Mainstreaming Policies 18 Figure 2: Simple Linear Policy Process .............................................................. 31 Figure 3: Non-Linear Policy Process ................................................................... 33 Figure 4: References to Standalone Women's Empowerment Projects, Mainstreamed Programmes and Mainstreamed Operations in UN Albania Resident Coordinator Reports from 2005-2012. ............................................... 110 Figure 5: References to Standalone Women's Empowerment Projects, Mainstreamed Programmes and Mainstreamed Operations in UN Bosnia and Herzegovina Resident Coordinator Reports from 2005-2012. .......................... 110 Figure 6: References to Standalone Women's Empowerment Projects, Mainstreamed Programmes and Mainstreamed Operations in UN Kosovo Resident Coordinator Reports from 2005-2012. ............................................... 111 Figure 7: Frequency of  occurrence of words "gender" and "women" in UN Albania Resident Coordinator Reports from 2005 - 2012. ................................ 113 Figure 8: Frequency of  occurrence of words "gender" and "women" in UN Bosnia and Herzegovina Resident Coordinator Reports from 2005 - 2012. ................. 114 Figure 9: Frequency of  occurrence of words "gender" and "women" in UN Kosovo Resident Coordinator Reports from 2005 - 2012. ................................ 114 Figure 10: References to Standalone Women's Empowerment Projects, Mainstreamed Programmes and Mainstreamed Operations in UNDP Annual Reports from 2001-2013. .................................................................................. 119 Figure 11: References to Standalone Women's Empowerment Projects, Mainstreamed Programmes and Mainstreamed Operations in UNICEF Annual Reports from 1999-2012. .................................................................................. 120 Figure 12: Frequency of  occurrence of words "gender" and "women" in UNDP Annual  Reports from 2001 - 2013. ................................................................... 121 Figure 13: Frequency of  occurrence of words "gender" and "women" in UNICEF Annual  Reports from 1999 - 2012. ................................................................... 121 Figure 14: Frequency of references to "Gender", "Women" and or Mainstreaming in UNDP Annual Evaluation Reports from 1993-2012. ..................................... 122 ix  List of Abbreviations   BiH   Bosnia and Herzegovina BPfA   Beijing Platform for Action CSO   Civil Society Organization CSW   Commission on the Status of Women  DAW    Division for the Advancement of Women ECOSOC  Economic and Social Council FAO   Food and Agricultural Organization  FCP   Feminist Comparative Politics GFP   Gender Focal Point  GM   Gender Mainstreaming IDO   International Development Organization IFAD   International Fund for Agricultural Development ILO   International Labour Organization IO   International Organization IOM   International Organization for Migration NGO Non-governmental Organization OHCHR Office of the High Commissioner on Human Rights OSAGI Office of the Special Adviser on Gender Issues and Advancement of Women SOP   Standard Operating Procedure TOR Terms of Reference UN United Nations UN INSTRAW United Nations International Research and Training Institute for the Advancement of Women  UN Habitat United Nations Human Settlements Programme UN Women United Nations Fund for Gender Equality and Women‘s Empowerment UNAIDS  Joint United Nations Programme on HIV/AIDS UNCT   United Nations Country Team UNDAF  United Nations Development Assistance Framework UNDP   United Nations Development Programme UNEP   UN Environment Programme  UNESCO  UN Educational, Scientific and Cultural Organization  UNFPA  United Nations Population Fund UNDG  United Nations Development Group UNHCR  UN High Commissioner for Refugees UNICEF  United Nations Children‘s Fund x  UNIDO  United Nations Industrial Development Organization UNIFEM  United Nations Development Fund for Women UNKT   United Nations Kosovo Team UNOPS  United Nations Operations UNPROFOR  United Nations Protection Force UNV   United Nations Volunteers WHO   World Health Organization  xi  Acknowledgements  Many thanks to my committee: Lisa Sundstrom for never giving up on me; and Katia Coleman and Olena Hankivsky for making the final year such an intellectual challenge. The nights and weekends spent drafting and revising this dissertation never felt like a burden, but rather were a true pleasure. Thank you to the Social Science and Humanities Research Council, whose doctoral grant allowed me to travel and conduct field research that eventually led to an exciting and challenging career in the politics of gender equality and women's empowerment and has afforded me the opportunity to meet inspiring women and men from government, civil society and academia in countries around the world. I owe my mother-in-law everlasting gratitude for days and nights of taking care of my children so that I could pursue my travels and my professional and academic ambitions with freedom.  I want to thank my parents for providing a lifetime of support. I must acknowledge the efforts of my Special Advisor on Data Visualization, Hrvoje Lukatela and the efforts of my Chief Linguist, Dunja Lukatela.  I want to thank "The Departed" - in Amman, Brussels, Cape Town, Lima and New York - for always having something to say, no matter in what time zone, meeting, or state of mind I happen to find myself.  My professional, personal and intellectual pursuits have been indelibly marked by the passion and spirit of Anne Marie Goetz, from whom I have learned everything there is to know about grace under fire, strategic thinking, and lifelong loyalty. Thank you for inspiring me. Lastly, but most importantly, I want to thank my wonderful husband, Andrej Hegedis, a poet warrior, for his courage in choosing a wife who is totally unconventional for his context, and for his strength, unconditional love and beautiful spirit.    xii   Dedication      Dedicated to women and girls around the world who dare to say   no.  And the men and boys who dare to listen.   Audaces fortuna iuvat. 1  1. Introduction  Research Problem and Justification  The concept of gender mainstreaming was introduced more than fifteen years ago in the Fourth World Conference on Women in 1995 and adopted by the UN as a strategy in 1997 by an ECOSOC Resolution. It has grown over time into an almost universal principle embedded in the policy frameworks of a wide range of international development organizations (IDOs).  While there are more expansive conceptions of gender mainstreaming (Jahan, 1995) this dissertation focuses on examining outcomes in gender mainstreaming as defined by the UN. In this definition gender mainstreaming is not a goal in and of itself, but rather is a strategy for achieving the ultimate goal of gender equality. The UN definition of gender mainstreaming that informs this study states: Mainstreaming a gender perspective is the process of assessing the implications for women and men of any planned action, including legislation, policies or programmes, in any area 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 the policies and programmes in all political, economic, and societal spheres so that women and men benefit equally, and inequality is not 2  perpetuated. The ultimate goal is to achieve gender equality.1 The thirteen largest international development agencies (see Table 1), regardless of whether they are multilateral organizations, non-governmental organizations, or faith-based organizations, or whether they focus on poverty reduction or humanitarian relief, all have policies referring to mainstreaming gender in their work and increasing the number of women in their staffing structures.  What is puzzling about the pervasiveness of this concept among international development organizations is the lack of evidence that it has actually been effective in contributing towards more positive gender equality outcomes.  An increasing number of scholars have reached agreement on its ineffectiveness and lack of results (Tolhurst et al, 2012; Powell, 2005; Squires, 2005; Daly, 2005; Lang, 2009; Hankivsky, 2005; Hankivsky, 2009; Hankivsky, 2013; Lewis, 2006; Bendl and Schmidt, 2013; True and Parisi, 2013), however it continues to be the only strategy applied by international development organizations towards achieving gender equality. This disjuncture between evidence and policy is puzzling. Yet few scholarly studies focus on this precise question of why a policy might be proven as ineffective and yet maintain its supremacy relatively unchallenged within organizational policy discourses.                                               1 UN Economic and Social Council Agreed Conclusions (E/1997/L.30) 3   Table 1: Characteristics of 13 Largest International Development Organizations based on Annual Budget Size in USD2 International Development Organization Annual Budget 2012/2013 in Millions of USD Number of country offices Number of staff Gender Mainstreaming Policy UN Development Programme 5300 177 16356 Yes World Bank 5000 97 15312 Yes IFRC (Red Cross and Red Crescent) 1800 187 4285 Yes World Health Organization 1500 147 8000 Yes World Vision International 1000 100 44000 Yes UNICEF 966 190 1197 Yes Oxfam International 918 90 10230 Yes Food and Agriculture Organization of the UN 899 130 3600 Yes Save the Children 617 120 14000 Yes Care International 585 56 200 Yes Catholic Relief Services 440 91 No data Yes International Rescue Committee 396 40 8000 Yes International Relief and Development 355 42 1750 Yes                                              2 Based on data collected from website of each organization and in personal correspondence. 4  Some of the reasons that these scholars identify for mainstreaming's failure include critical flaws in its conception (Tolhurst et al, 2012); that it has relied upon a liberal feminist approach which does not take into account new developments in feminist theorizing that look at identity and the ―interface‖ between gender and race, class, ethnicity, and poverty (Hankivsky, 2005); and that strategies have been designed and implemented by western feminists and practitioners without the participation of local women and gender equality advocates (Hankivsky, 2009). As evidence of its failure Bendl and Schmidt  (2013) argue that in most EU countries that have mainstreaming strategies there has been no change across discourses, structures and processes while Lang (2005) argues that failure is evidenced by a lack of implementation. All of these factors emerged as trends in the research conducted for this dissertation. Interview respondents expressed concerns that gender mainstreaming was simplified to increasing women's participation, as opposed to a transformative approach addressing diversity and local contexts and that there has been little significant implementation or change they can identify and attribute to mainstreaming,  This dissertation will examine this question of why, despite years of experience in attempting to implement this policy, international development organizations still have not achieved the expected results. Correspondingly, why has mainstreaming continued to be maintained as a policy in the largest and most significant international development organizations despite this failure? What causal mechanisms lead international development organizations (IDOs) to 5  make specific policy choices that seem to have no rational or results based foundation and, more importantly, what factors affect how these choices are then implemented? The literature up to now has examined in detail gender mainstreaming (GM) as a concept and its impact (or lack thereof), however few studies have investigated the dynamics behind the adoption and implementation of mainstreaming, how the gap between adoption and implementation can be remedied, and how this relates more generally to questions of why some policies in international development organizations are implemented and others remain words on paper. Mainstreaming policies are different from traditional policy decisions in the way they are adopted and implemented. Mainstreaming refers to the process of applying a specific lens across a range of thematic policies. Examples of mainstreaming policies adopted by international development organizations include gender, HIV/AIDS, and environment. Mainstreaming policies are not implemented by a dedicated programme manager but rather are meant to be implemented by the range of thematic experts in an organization, each in their own policy area. For this reason the responsibility and accountability for mainstreaming policies is different than for other policy issues.  The body of research on gender mainstreaming (GM) presents evidence from the past decade that this concept has not been successfully implemented and has not contributed significantly to its goals and yet it remains almost universally accepted in both development and non-development contexts (Tolhurst et al, 2012; Powell, 2005; Squires, 2005; Daly, 2005; Lang, 2009; 6  Hankivsky, 2005; Hankivsky, 2009; Hankivsky, 2013; Lewis, 2006; Bendl and Schmidt, 2013; True and Parisi, 2013).3 In spite of this recognized failure and the increasing number of academic studies proposing alternatives (Hankivsky, 2009; Powell, 2005; Squires, 2005), no serious discussion has emerged within international development policy circles of alternative policies towards achieving gender equality.  Failure is generally described by these studies as a lack of meaningful change in an organization‘s impact on the social, economic and political outcomes of women since gender mainstreaming‘s ascent to dominance in 1995 (Rai, 2008; 1-10). Why has this strategy been so unsuccessful? What can we learn about mainstreaming by investigating the dynamics that lead to its adoption and implementation? This empirical puzzle leads to the research questions of this study:  1. What explains adoption and implementation of gender mainstreaming policies in international development organizations? 2. What do these dynamics tell us about reasons for the persistence of gender mainstreaming in the face of recognized failure?  Summary of Findings This dissertation examines these questions through a comparative case study of gender mainstreaming in three UN country teams in Albania, Bosnia and                                             3 A comprehensive review of Gender Mainstreaming in UN Operational Activities for Development in 2012 also found little success or progress to report. UN Women, Advancing Gender Equality and Women‘s Empowerment: An Assessment of Gender Mainstreaming in UN Operational Activities for Development, 2012.  7  Herzegovina and Kosovo. The collected evidence is examined against an explanatory framework of four different theoretical approaches to policy processes in international development organizations: the constructivist approach, the world polity perspective, the bureaucratic politics theory, and the theory of pathologies of organizational behavior.  The findings lead to the conclusion that gender mainstreaming has become embedded through non-rational mimicry in the discourse of the international civil servants working in the senior management of the UN system. Non-rational mimicry is the mechanism identified by the Word Polity theory for the dissemination of world values and culture. The world polity is defined in full detail in the theoretical framework chapter, however it essentially refers to the set of cultural values and directions that influence actors, including individuals, organizations and states, and has a homogenizing effect over time. However the answer to the second research question above points to the need to separate the causal mechanisms of each phase of the policy process: policy adoption vs. policy implementation. In the case of policy adoption the observed mechanisms are the non-rational mimicry and bureaucratic competition of senior management. However for the latter phase a completely different causal mechanism seems to be at play – the normative commitments of mid-level management coupled with the standard operating procedures they can use to their advantage to influence the behavior of otherwise passive colleagues.  This finding has intuitive validity if we consider that implementing crosscutting mainstreaming policies requires the active participation of diverse 8  sectors of an organization‘s personnel. Thus a normative commitment and strategically applied standard operating procedures can move policy towards implementation, regardless of the behaviours, commitment and relationships of senior management. This was also borne out by the evidence; there were instances where the research showed that a committed staff member could include a gender perspective in their programme regardless of whether senior management approved of this or not. The motivation to mainstream gender into an individual‘s work was born of a reflexive belief in equality and justice and an interest and willingness to identify gender inequalities where they were present.  However it must also be noted that while standard operating procedures can be used by mid-level management to finesse compliance from peers who hold passive norms, they cannot overcome the resistance of individuals with norms hostile to gender equality.  As a result gender mainstreaming has failed not necessarily because of problems with its conceptualization or design, but rather because of dynamics related to the structure and functioning of the mid-level management tier of international development organizations. The implementation of mainstreaming policies requires the commitment and mobilization of a diverse set of actors within this very tier. In the absence of this profile of staff there will be no progress on implementation, even in the presence of declarative commitment and allocation of financial resources from senior management. The literature up to now has not taken into account the importance of this aspect of implementation of mainstreaming policies. The scarcity of actors in IDOs who combine 9  technical/thematic expertise with an internalization of gender equality norms is a major factor in the failure of gender mainstreaming since policy implementation depends largely on their presence. Remedying this gap requires the conscious decision by a set of senior managers to engage in organizational change aimed at transforming the structure and profile of its mid-level management. Significance and Contributions of the Study  This study provides important insights into theories of international organizations by explaining the mechanisms behind policy choices (both adoption and implementation) and also contributes towards understanding ―mainstreaming‖ policies broadly defined. Many of the existing analyses of gender mainstreaming policies have come from the feminist paradigm and have made important and useful critiques. However, few studies have considered gender mainstreaming from an organizational policy paradigm and what the implications of this are for other "mainstreaming" policies now that the approach has gained popularity in international development organizations. The mainstreaming discourse started in international development in the 1990s with gender, however today it is also used widely by a range of international organizations in relation to HIV/AIDS mainstreaming, environmental mainstreaming, governance mainstreaming and disability mainstreaming. In this context these strategies refer to mainstreaming these issues within overall development strategies. More on the development of the mainstreaming discourse will be discussed in Chapter Two. 10  While some scholars posit that the existence of senior commitment to an adopted policy is a sufficient condition for its success (Kingdon, 1984; Mazur, 2002), this study shows that in the absence of a tier of middle management implementers and partners with appropriate normative allegiances, even the best designed crosscutting mainstreaming policy will not be successful. This conclusion in particular offers interesting new research directions for scholars studying the international influences on domestic public policy cycles and public policy diffusion by global governance institutions, as well as the future of gender mainstreaming in both national and international bureaucracies.  The conclusions of this study offer important recommendations for the future of different mainstreaming policies in development agencies. Understanding the dynamics necessary for policy implementation and gender equality outcomes is crucial as the UN is currently in the midst of a protracted dialogue on development agency reform and aid effectiveness  (Paris Declaration, 2005; UN Secretary-General‘s High Level Panel on System Wide Coherence, 2006; High Level Forum on Aid Effectiveness, 2011; Secretary-General‘s High Level Panel on the Post-2015 Development Agenda, 2013). The UN is also engaged in a process of fundamentally rethinking its policy on engagement in development of transitional and fragile states (Dili Declaration, 2010, The Monrovia Roadmap, 2011).   These reform initiatives are important because trends in the UN have an impact on the whole field of international development given that it is the organization with the most funding and reach in terms of staff and country offices. 11  Reform in the development branch of UN agencies will thus have an impact on how the field overall perceives itself and defines its priorities.  The success or failure of gender mainstreaming has become tied to the structure and effectiveness of the UN‘s ―gender architecture reform,‖ however an over focus on structure vs staff may be part of the problem and this dissertation is an attempt to also provide policy recommendations on this issue.  In January 2011 four UN agencies4 working on policy and programmes related to gender equality were formally combined into one new agency, the UN Entity for Gender Equality and the Empowerment of Women (UN Women). UN Women was given a wider mandate, more senior leadership within the UN system, a larger budget and slated to promote the mainstreaming agenda within the UN (General Assembly Resolution A/64/289). Even though all UN agencies are corporately committed to gender mainstreaming, the establishment of UN Women was meant to give new energy to the pursuit of gender mainstreaming in UN development agencies and lead to a new leveraging of partnerships among UN, other international and regional development agencies, national governments and local NGOs. However the findings of this study give cause for concern as to whether the UN system and international development organizations more broadly, will be able to achieve their objectives. Without taking into account the obstacles to policy implementation within organizations, efforts may be futile.                                              4 Division for the Advancement of Women (DAW), Office of the Special Adviser on Gender Issues and Advancement of Women (OSAGI), International Research and Training Institute for the Advancement of Women (INSTRAW), and UN Development Fund for Women (UNIFEM).   12  The study chose case studies based on the varying strength of UNIFEM's presence on the ground, in anticipation that it was a factor in the presence of gender mainstreaming. However the findings demonstrated that this was not necessarily the case as the variation in strength of the UN's gender equality agency did not make a difference to the outcome. It is critical for the success of not only UN Women, but also the whole UN development agency system, that the impact of the mechanisms described in this study are considered before unrealistic expectations are placed on the ability of UN Women to renew the vigour in mainstreaming efforts in the UN system. A recent policy paper, written after the formation of UN Women by the former Deputy Executive Director of UNIFEM, argues that units ―tasked with promoting gender equality in mainstream organizations face pervasive institutional discrimination‖ and that without a transformation of the overall system, including the overall UN system with regards to its relationship with UN Women, gender mainstreaming cannot be achieved. While she attributed the concerns with UN Women‘s formation to senior leaders trying to protect their own power within a patriarchal and elitist bureaucracy, her evidence also clearly showed examples of strong instances of targeted discrimination against leaders who openly self-described as feminists (Sandler, 2012). Suggesting that while everyone mimics the discourse of gender equality and gender mainstreaming, there is no clear advantage tied to promotion or career advancement in showing results on gender mainstreaming. This study suggests that it is critical to look in detail at how policy adoption and policy implementation can differ drastically. It argues that a large part of what 13  IOs do is in fact based on the norms of their middle management, the implementers, and not necessarily based on the institutional rules designed to control or limit their actions, or the political will of their Member States or senior management. Establishing why policy implementation can fail despite the presence of corporate level commitment is extremely valuable knowledge for organizational theories of change and international relations theory because it runs counter to common understandings that the most important factor for policy implementation is commitment at the highest levels or Member State support for policy adoption. Developing more effective procedures and mechanisms to identify and reward candidates who display the right normative commitments paired with the relevant competencies needed for mid-level management appears to be critical. The lessons of this study can also be applied to improving the follow-through of many other crosscutting or mainstreaming policies, such as HIV/AIDS, environment, disability, sustainability, adaptation, etc.   Case Selection and Sources of Evidence The dissertation will contribute to empirical and theoretical knowledge on how international organizations adopt and implement policies; under what conditions crosscutting/mainstreaming types of policies can be effective; and also on how progress towards gender equality can be achieved by international development organizations. The study links International Relation's (IR) interest in the functioning of international organizations with the policy process, by investigating the adoption and implementation of gender mainstreaming. 14  Sources of evidence for this dissertation include content analysis of approximately two hundred UN agency policy documents, implementation reports and evaluations. The data is disaggregated by year and (where possible) by country in order to examine trends in policy adoption and policy implementation over time and location.   In addition more than 75 semi-structured interviews were conducted from 2010-2013 with UN, civil society and government staff across New York, Tirana, Sarajevo and Pristina. The author also engaged in participant observation of the UN country teams in BiH, Kosovo and Albania during  a period of employment in UNIFEM in Southeast Europe and with UN Women in New York (all together from 2009-2013). Data from work experience prior to the 2009 ethics approval for this research was not included. During this period of 2009-2013, events, phenomena and observations were collected and recorded for the purposes of the dissertation. More details on the sources of evidence and cases will be presented in the methodology chapter.  The dissertation uses the evidence collected and the observable implications of the theoretical framework to conduct process tracing and identify the mechanisms of interaction operating between cause and effect (Bennett, 2004; Bennett and George, 2005; King, Keohane, and Verba, 1994: 85-65; Checkel, 2004). By describing causal processes involving policy adoption and policy implementation separately, this dissertation is able to deepen understanding on policy processes in international relations and to build a theory that explains the non-linearity of policy.  15  UN Development Agencies The UN has a number of different agencies that deal with social and economic development issues in countries around the world. These agencies cover every sector from agriculture to sustainable energy. Programmes are implemented by country offices of these UN agencies, which may or may not be housed together, and often are also implemented in partnership with national authorities of the country. The major UN development agencies are the United Nations Development Programme (UNDP), the United Nations Population Fund (UNFPA), the Food and Agriculture Organization (FAO), the International Fund for Agricultural Development (IFAD), the International Labour Organization (ILO), the United Nations Educational, Scientific, and Cultural Organization (UNESCO), the World Health Organization (WHO), the United Nations Children‘s Fund (UNICEF), the United Nations Entity for Gender Equality and Women‘s Empowerment (UN Women), the International Organization for Migration (IOM) and the United Nations Human Settlements Programme (UN HABITAT). The collection of UN development agencies present in a particular country is led overall by a Resident Coordinator who is responsible for developing a coordinated UN strategy to cooperation with government on key priorities, which happens through the negotiation of a UN Development Assistance Framework (UNDAF).  The UNDAF describes the collective response of the UNCT to priorities in a national development framework, which in principle are jointly identified by government, UN and other international stakeholders.  16   Official policy documents on gender mainstreaming issued by UN development agencies started to appear in the 1990s and became increasingly similar over time in terms of both content and structure.  Table 2 below lists the policies5 developed by major development agencies on gender mainstreaming and displays the trend over time of convergence of the elements in their structure and content. Early on in the 1990s policies focused on mainstreaming gender into programming (Prog) and establishing accountability procedures (Acct) for doing so. Over time and in several waves (1996, 2001-2002, 2008) elements in mainstreaming strategies began to include funding (Fund), mainstreaming human resources (HR), using indicators to track progress (Ind), setting up institutional mechanisms for mainstreaming (IM), conducting research (Res), mainstreaming operations (Ops), training staff (Tr) and establishing partnerships and cooperation (Pshps).  Figure 1 on the following page lists those policies that underwent independent evaluations commissioned by their agencies, as well as the conclusion of the assessment (none of which were positive). The figure presents a matrix of the various issues identified by each evaluation as factors in the lack of results. While the list of factors is not broad, each evaluation tended to focus on different issues. Many of the issues appear in the results of this study, such as the over-reliance on individual commitment, but interestingly some of the most significant factors identified in this research did not come out prominently in these                                             5 See Appendix C for the full list of policy titles. 17  independent evaluations - such as the impact of organizational culture and individual norms.  Table 2: Trends in Content and Structure of UN Agency Gender Mainstreaming Policies since 1989. 18  Figure 1: Results of Evaluations of UN Agency Gender Mainstreaming Policies   19  Normally UN country presences are referred to as ―UN Country Teams‖ (UNCTs). However the case of Kosovo is specific; it is not a Member State of the UN (only 105 of 193 UN member states have formally recognized Kosovo‘s independence as of November 2013); for the purposes of the UN mission and UN agencies it is a UN administered territory of Serbia under UN Security Council Resolution 1244. In this context, referring to agencies in Kosovo as a UN Country Team is technically incorrect, as they self-reference as UN Kosovo Team (UNKT) to avoid political problems. For the purposes of this study and ease for the reader, UN country team may be used when referencing all three cases simultaneously, while standalone references to Kosovo will be written as UNKT.   In general UN country teams function in similar ways, they are headed by a UN Resident Coordinator (RC), who leads the work of the family of UN agencies and represents the UN to the national government. The Resident Coordinator is responsible for the actions of the UN entities in the host country. The Resident Coordinator is typically also the Head of Agency of one of the UN agencies in country and in effect plays a dual role. Often the entity that the Resident Coordinator heads is UNDP or UNFPA – one of the larger UN entities. In this respect Resident Coordinators often struggle to maintain a neutral role and make significant attempts to not be seen as favoring their agency over the other agencies in the UNCT when it comes to, inter alia, resource mobilization and negotiations with host governments on funding and programming priorities.  20  UN country teams also have different types of theme groups through which different entities cooperate and coordinate programming on issues of mutual interest. Members of the theme groups are typically programme officers or focal points for that particular issue at the mid-management level. Examples are gender, communication and advocacy, monitoring and evaluation, policy and research, poverty, democratic governance, communities, etc. Of the three cases only Albania had a functioning UN Gender Theme Group to coordinate UN interagency gender equality programmatic activities throughout the country. UN Kosovo team had a Human Rights and Gender theme group; however it suffered from lack of leadership and motivation and eventually ―fizzled‖ out with meetings becoming sparser and sparser.6 Bosnia had a UNCT gender theme group; however it was clearly not a significant forum as it did not meet regularly. The focal point from each agency that attends the Gender Theme Group meetings is expected to report back to his or her agency on progress made towards achieving gender equality related ―outputs and outcomes.‖7 It serves as a vehicle for joint inter-agency programme initiatives and tries to ensure that progress towards the gender equality targets in the UN and national government‘s jointly agreed development assistance framework (UNDAF) are on track.  Many agencies in the cases had a ―Gender Focal Point‖ (GFP). The gender focal point is responsible for guiding the organization‘s gender                                             6 UNFPA Official, Interview by Ana Lukatela, Pristina, March 2010. 7 UN development agencies adopted a ―results based management‖ approach to their programming where all activities must feed into outputs, which feed into outcomes, which feed into an overarching broad goal meant to contribute towards an impact at the societal level.  21  mainstreaming strategy and its monitoring. The strategy may refer to mainstreaming gender into all policies, programmes and structures, or may only focus on mainstreaming gender into programming. Establishing a Gender Focal Point is a corporate requirement for UNDP‘s country teams and is reflected in the UNDP Gender Equality Strategy of 2008-2011 (UNDP, 2008). While UNDP Gender Focal Points are ideally supposed to have a ―Terms of Reference‖8 (TOR) stipulating their responsibilities, this is often not the case in practice and instead the duties of gender focal point are piled on top of already existing portfolio duties.   Case studies looking specifically at UN agency teams were selected because the UN development agencies can be considered ―crucial cases‖ in the development context. They were among the first donor and development agencies to officially institute gender mainstreaming policies and in any given development context the family of UN agencies is almost always the most powerful development presence on the ground financially, materially and in terms of number of personnel. For this reason how UN teams institute gender mainstreaming is more critical than how a smaller donor or development agency, which does not have such a universal presence, institutes gender mainstreaming. UN agencies on the ground in development scenarios are both significant leaders and actors and provide signals to others on how to institute policy and agendas. This also means that these cases will be useful when drawing mid-level generalizations from the results to the wider development                                             8 Terms of Reference is the UN phrase used for a job description. 22  context. The generalizations are mid-level to the extent that they are meant to travel to policy issues in international development organizations; they are not meant to reflect on implementation of policy in the West and developed countries.  Of the key UN development agencies it is important to note that UNDP has the largest annual budget, while UN Women (formerly UNIFEM) has the smallest. This is an important reflection also of the gendered nature of the structure of UN development agencies.  UN Women has missed its annual fundraising target of voluntary contributions from Member States for each year since its establishment. While its senior management planned for a goal of 500 million USD in 2011, it raised not even half of this - only 230 million. While funding levels have been affected by the global economic crisis since 2010, it is also clear that the way funding is channeled is a reflection of the relative priority of the issue in the development discourse.  Table 3: Annual UNDP and UNIFEM/UN Women funding levels in USD (both regular contributions and voluntary contributions)9  2008 2009 2010 2011 2012 UNDP 5 billion 5.3 billion 5.5 billion 4.8 billion 4.6 billion UNIFEM / UN Women 0.21 billion 0.18 billion 0.18 million 0.23 billion 0.22 billion                                              9 Compiled from UNDP and UN Women annual reports 23  Outline of Dissertation Chapter Two presents the dissertation‘s theoretical framework. It develops the key concepts, highlights current gaps in the literature, proposes a theory and elaborates on alternative explanatory frameworks of mainstreaming policy adoption and policy implementation in international organizations.  Chapter Three outlines the methodology used to examine the theoretical frameworks and explains the case selection. Chapters Four and Five examine the dynamics of policy making and policy implementation. Chapter Four explores the relationship between policy adoption, world polity theory and the bureaucratic politics approach. Chapter Five examines the relationship between policy implementation and the constructivists‘ focus on norms and values as well as the significance of standard operating procedures explained by organizational behavior theory. Chapter Six discusses the theoretical implications of the findings for all types of crosscutting/mainstreaming policy processes and policy recommendations for the future of gender mainstreaming. It also suggests directions for future research and a summary and conclusion.  24  2. Theoretical Framework: Policy Adoption vs. Policy Implementation in International Development Organizations Introduction The theory and the alternative explanations presented in this chapter look at the values, motivations, norms and knowledge of those actors responsible for adopting and implementing policies in international development as well as the influence and impact of the policy context within which international organizations exist. These four explanations are all grounded in theoretical approaches that accept the agency and autonomy of international organizations.  In particular there has not been a debate in the literature of how the policy process might specifically be different for mainstreaming policies (those policies intended to crosscut a wide range of other thematic policies and programmes) and why there has been a gap between adoption and implementation.  What are the conditions under which a mainstreaming policy will be adopted and yet not be implemented? This issue has suffered from a lack of conceptual clarity in the existing literature. The different stages of the policy process are not distinguished and surprisingly little work has been done to develop a robust theory of how these two discrete processes unfold in international organizations.  This is all the more surprising since in related fields there has been an investigation into this question. For example Harrison and Sundstrom in their examination of international environmental treaties recognize that the mechanisms that lead to treaties being signed and policy adopted on paper are very different from the national-level mechanisms that influence implementation 25  (Harrison and Sundstrom, 2007: 2). In the same vein Bacchi and Eveline (2009) recognize in their study of Aboriginal politics and mainstreaming in Australia that there is a clear gap between policy and ―doing.‖ This chapter will discuss the literature on international organizations and mainstreaming policies from the perspective of policy adoption vs. policy implementation and will outline the dissertation‘s theoretical framework as well as relevant alternatives. The dissertation recognizes the gap between policy adoption and policy implementation and uses it as a starting point to propose a theory of the impact of mainstreaming policies in international organizations.  The findings of this study have interesting implications for international organization (IO) institutional and neo-institutional theory. Institutionalism argues that IOs ―do what they are designed to do‖ (Haas, 1964), while neo-institutionalism argues that IOs ―do what is in their interest‖ (North, 1990; Barkin, 2006:36). This research, on the other hand, suggests that both of these approaches are missing an important factor by not looking in more detail how policy adoption and policy implementation can drastically differ. This study argues that a large part of what IOs do is in fact based on the norms of their middle management, the implementers, and not necessarily based on the institutional rules designed to control or limit their actions or the political will and commitment of their Member States or senior management.  The puzzle with gender mainstreaming is directly related to this gap. Why has it been persistently accepted by senior level policymakers as the sole policy option on gender equality, despite the agreement among scholars that it has 26  suffered from a lack of implementation and a lack of mid-term impact (Tolhurst et al, 2012; Powell, 2005; Squires, 2005; Daly, 2005; Lang, 2009; Hankivsky, 2005; Hankivsky, 2009; Hankivsky, 2013; Lewis, 2006; Bendl and Schmidt, 2013; True and Parisi, 2013); and despite the academic discussions on policy alternatives to gender mainstreaming (Hankivsky, 2009; Powell, 2005; Squires, 2005)?  Instead of applying one theory to the whole policy process, this dissertation argues that it is necessary to break down the policy process into its distinct parts – the policy context, policy adoption and policy implementation – and then to propose mechanisms relevant to each of these stages.  The dissertation also argues that global policies that touch on strongly held norms about individual behavior in traditional societies face a different adoption and implementation path than other types of policies that deal with standalone thematic issues. The state-centric theoretical approaches within international relations give relatively little insight into this puzzle. They do not concern themselves with intra-organizational details of policy adoption and implementation as they assume the elements of policy are uniformly affected by the will of Member States. They focus their attention on state-centric mechanism of power dynamics with the assumption that a policy adopted at the behest of Member States is a policy implemented at lower levels. However as Avant, Finnemore and Sell argue, theories of International Relations (IR) that ―focus only on states are poorly equipped to understand the kaleidoscope of activity by such a wide range of actors‖ ongoing in the global governance sphere (Avant et al, 2010: 2). This 27  dissertation takes the independence and agency of IOs working on political, social and economic issues as a central assumption. For this reason the theoretical approaches that are the focus of this dissertation are world polity theory, constructivism, organizational behavior theory and bureaucratic politics theory.   IR theories of international organizations focused for decades on regime theory, rationalism and realism (Krasner, 1983; Grieco,1988; Keohane,1984; Habermas,1962; Fearon,1995) and studies of the UN have overwhelmingly focused on the dynamics of negotiations between members of the Security Council on specific issues (Kennedy, 2006: 210; Barkin, 2006). However with the end of the ideological split of the Cold War the dominance of realism in international relations theory began to fade as states began to behave in previously unforeseen manners (Rother, 2012). Approaches in the realist tradition had previously not looked at IOs as autonomous actors, which resulted in an incomplete explanation of many phenomena in the international relations arena, such as the way international organizations function.  Realist, neo-realist and neo-liberal approaches were not considered within the scope of this research as they do not make specific claims about the independent influence of international organizations. These approaches do not open the ―black box‖ of international organization functioning (Broome and Seabrooke, 2012). Realists do not accept that NGOs, IOs and expert networks can independently influence world politics in significant ways (Mearsheimer, 1994; Morgenthau, 1960; Wapner, 2002: 37), while neo-realists and neo-liberals 28  argue that states maintain membership in IOs only as long as there is mutually beneficial cooperation (using a rational design model) (Keohane, 1993; Koremenos et al, 2004) or that IOs are used as hegemonic tools of the most powerful member states (Krasner, 1991; Gilpin, 2001). These authors do not consider IOs to have authority independent of the state not to be actors with agency.  Barnett and Finnemore (1999) point out that scholars in the realist tradition generally assume that IOs exist in a competitive environment that exerts strong pressure for efficient and responsive behaviour and sets clear criteria for success (Barnett and Finnemore, 1999: 704). However, in accordance with the observations of the author of this dissertation, international organizations operate with vague directions of mandate, little external interference and no serious threat of dissolution. Within these conditions it is not difficult for their staff members to exert influence based on their expertise and ―legal rationale authority‖ that runs counter to state interests or preferences.  This is the case with UN development agencies, which have complicated structures and whose staff employ strategies to buffer themselves from hostile member states in their Executive Boards and in the organization as a whole. One anonymous respondent in this study described the difficulties associated with Pakistan‘s membership on UN Women‘s Executive Board. The respondent explained that it just meant giving less visibility to some initiatives in Southeast Asia or purposefully refraining from mentioning any potentially controversial 29  activities in reports to the Executive Board.10  In the author‘s many years of observation of the UN system, member states have very little actual involvement in the day to day functioning of UN development agencies. This chapter will present the theoretical framework of the dissertation as well as alternatives, while the following chapter will outline in detail the methodology and observable implications guiding the study.  International Organizations and Policy Processes The number of international organizations has proliferated – while there were only 67 in 1930, by 2000 there were more than 330 (Beckfield, 2010). These organizations differ – some are transnational civil society organizations, others are scientific/cultural professional associations and others are multilateral intergovernmental organizations. Based on a study of the annual data collected by the Union of International Associations Beckfield concludes that these networks persist and that the density of their connections continues to grow as world polity theory (described below) predicts (Beckfield, 2010).  Traditional international organization research has looked at the external influence of international organizations; how these "regimes" can influence the behaviour of states (Krasner, 1983; Grieco, 1988; Keohane, 1984; Habermas, 1962; Fearon, 1995). However the regime theory treats the internal workings of IOs as a ―black box‖ (Ellis, 2010). As Benner et al argue (2007) ―mainstream work in the discipline of International Relations (IR) has produced surprisingly                                             10 UN Women Official, Interview by Ana Lukatela, New York, 2012.  30  few studies on the everyday workings of international organizations.‖ Studying the UN‘s bureaucracy can make a significant contribution to ―opening up the ―black box‖ of international organizations‖ (Benner et al, 2007: 4). In his book International Organizations: Theories and Institutions, J. Samuel Barkin argues that the extent to which development assistance IOs set their own agendas, independent of Member State influence, is unclear (and he particularly references UNDP in this context). If IOs are indeed setting their own agenda, then it is also unclear whether the ―agenda is being set primarily at the political level or at the operational level‖ (Barkin, 2006: 108). In addition, Barkin argues that different IOs have different levels of centralization, with some more focused on the ―development priorities of the central IO bureaucracy‖ than others (108). Standing (2004) argues that policy failure from adoption to implementation is due to the inability to understand how bureaucracies at the mid level truly function in terms of translating a political agenda into a technical one. In particular when it comes to international development organizations stakeholders are naïve about the causal links between policy intention and policy outcome. These examples clearly set out the question that this dissertation examines - the differing impact of the political and operational levels of IOs on policy adoption and policy implementation.  The classic rational paradigm of public policy analysis argues for a linear or cyclical model that emphasizes formal procedures and ignores complex and value-laden aspects of the process (Howlett and Ramesh, 2003; Weingast, 1979; Riker, 1990) (see Figure 2). Linear refers to the direct progression from context 31  to adoption to implementation, where implementation is dependent only policy adoption. It is an effort to understand policy as a sum of progressive and sequential stages: ―agenda‐setting, decision‐making, implementation, evaluation, and termination‖ (Howlett, Ramesh and Pearl, 2009). However, as Nelson argues, the most ―significant contributions from research on the policy cycle have been insights into the workings of the separate stages of the policy process‖ (Nelson, 1996: 567). Eyben argues that policy must be understood as a site for resistance and contestation and that interpreting the policy process as a ―top-down linear policy implementation can seriously constrain an imaginative search for more appropriate understandings of the context and possible responses to the [policy] context" (Eyben, 2010: 55).  Kingdon‘s classic work on agenda-setting has focused on the first part of the policy cycle – policy adoption – and examines how ideas enter the problem process, the policy process and the political process to eventually end up as legislation (Kingdon, 1984). However he does not focus on the difference between adoption and implementation and how the two may differ.  Figure 2: Simple Linear Policy Process  This study agrees with Nelson that the observation of policy processes rarely leaves the impression that it functions in a linear path. Instead a non-linear and non-cyclical approach to the study of policy in international organizations 32  provides a much richer and comprehensive analysis and insight of the workings of the separate parts of the policy process. Most importantly, this approach allows for better explanations of crosscutting policies, such as mainstreaming, that due to their complex nature may be more likely to function in a non-linear and non-cyclical fashion. This dissertation proposes that mainstreaming policies are heavily dependent on context not only at the beginning of the cycle but at each point along the way.  In a non linear approach to policy processes there are discrete causal mechanisms influencing both the adoption stage and the implementation stage. Adoption is not a necessary condition for implementation since the crux of mainstreaming relies not on a veto by senior management, but on middle management attitudes and behaviour. A middle management practitioner who is sensitive to gender inequality and diversity may be implicitly gender mainstreaming his/her programme without even thinking of it and without a policy instructing him/her to do so on paper – but rather doing so through their own ―unconscious competence.‖11  In juxtaposition to the simple process in Figure 2, Figure 3 provides a simple example of what a non-linear policy process might look like. In Figure 2 A and B and in Figure 3 A, B and C represent the proposed causal mechanisms                                             11 In psychology, the four stages of competence, or the "conscious competence" learning model, relates to the psychological states involved in the process of progressing from incompetence to competence in a skill. An individual learning a new skill progresses through unconscious incompetence, conscious incompetence,  conscious competence and reaches unconscious competence once they are so practiced that the skill becomes natural to their everyday life. Developed by psychologist Noel Burch in the 1970s. 33  between the stages. For each of these mechanisms the dissertation will propose a theoretical framework with an accompanying set of observable implications. Figure 3: Non-Linear Policy Process  This conception of a non linear policy process follows Falleti and Lynch‘s (2009) argument that ― ―relational concepts‖ are useful in proposing explanations of why institutions endure or change. Mechanisms are the types of generalizable processes that connect parts of the process to each other. The lack of attention to mechanisms within the non-linear approach to policy processes has meant that the nuances of policy adoption vs. policy implementation has not been addressed with due attention. Falleti and Lynch argue that the relationship between context and mechanisms is critical and that ―mechanisms tell us how things happen: how actors relate, how individuals come to believe what they do or what they draw from past experiences, how policies and institutions endure or change, how outcomes that are inefficient become hard to reverse, and so on‖ (2009: 1145-7). In the case of mainstreaming policies different patterns and interactions are present in relation to policy adoption and policy implementation. Due to the 34  different relational interactions between the policy context and the mechanisms no single approach can propose a full explanation of both adoption and implementation on its own.  The expectation of the linear policy process is that in a hierarchal organization a policy adopted is a policy implemented. Policy decisions are translated down the hierarchy and implemented. However, based on the author‘s own observations of international development organizations this is not necessarily the case and in his study of hierarchies Cooley (2005) argues that scholars of policy process place too much focus on the politics behind a policy‘s adoption and too little on how different kinds of hierarchies then implement these policies.   Why are Mainstreaming Policies Different? ―Mainstreaming‖ first appeared in the practitioner and academic literature in the mid 1980s referring to childhood education; it referred to the policy of transferring students from separate special education classrooms into general education classrooms (Hunt, 2010). In the 1990s ―mainstreaming‖ became synonymous with inclusion policies in general, which eventually led to its being used in reference to other social issues such as gender equality and HIV/AIDS.  Mainstreaming policies are different from traditional policy decisions in the way they are implemented, but also in their adoption. This dissertation examines why they are different and what we can learn from gender mainstreaming to 35  improve the implementation of mainstreaming policies in general in international development organizations. Mainstreaming policies are typically articulated through a corporate-wide policy document developed at headquarters, which is then translated into a ―local‖ directive in each country office. Programme managers are then tasked to implement this policy, often without any further direction or oversight.  This problem of how to translate the goals stated in a mainstreaming policy into action has been the subject of research (Eveline et al, 2009; Elsey et al, 2005). However, in the author‘s observations there has actually been very little communication between senior and middle management of how policies are to be implemented. Research on mainstreaming has also highlighted the problem of ―conflicting agendas‖ (Rathgeber, 2005). This refers to the competition that can exist between mainstreaming and other policy priorities; Rathgeber argues that in these cases mainstreaming is sacrificed as it is the secondary priority.  This dissertation argues that in the case of mainstreaming policies, which are implemented by managers with thematic expertise in one area but who are required to be ―sensitive‖ to the issue they are supposedly mainstreaming the critical mechanism is a normative one. In particular this is because issues being mainstreamed are often contentious ones – gender, HIV/AIDS, environment – that require commitment by an individual to the underlying concepts. A practitioner cannot mainstream gender if he/she does not ultimately believe that women‘s empowerment is a good thing; or mainstream HIV/AIDS without believing that harm reduction for drug users or tolerance towards homosexuality 36  is important; or mainstream environment unless one believes global warming and climate change are a threat to our way of life (despite its scientific basis, many people continue to be skeptical of both) (Diethelm and McKee, 2009). Some scholars criticize gender mainstreaming as a ―vague‖ concept of ―non-specific‖ character (Beveridge and Nott, 2002; Bendl and Schmidt, 2013; Caglar, 2013) and the lack of understanding among practitioners of the operational link between gender mainstreaming as a strategy and gender equality as a goal.  At the same time, the vagueness of the concept means there is little accountability in the way in which organizations implement gender mainstreaming. Paradoxically, this led to the rapid diffusion and almost universal acceptance of the concept in the development discourse, as it could be operationalized in any fashion with impunity (Hafner-Burton and Pollack, 2002). The nature of mainstreaming and crosscutting policies is at the heart of these challenges. Rebecca Tiessen writes that mainstreaming strategies suffer because they are ―everyone‘s and no one‘s‖ responsibility (Tiessen, 2005) and the findings of this study support this perception. This dissertation argues that thematic experts are being asked to operationalize an issue outside of their area of interest, which is the root of the problem. The worst case scenario is they will actively oppose the agenda if they feel it does not align with their personal views or normative commitments. If the thematic expert is uninterested in the crosscutting issue, then the most that can be realistically hoped for is that he/she will not promote but neither block implementation. Examples of other mainstreaming policies bear out this prediction as well. 37  Examining HIV/AIDS mainstreaming policy in international organization contributes further evidence towards the shared nature of mainstreaming policies. HIV/AIDS mainstreaming was established within the UN in 2005 when UNDP, UNAIDS and The World Bank published joint guidelines for ―Mainstreaming HIV and AIDS in Sectors and Programmes‖ (UN Guidelines, 2005). This document set out the principles for mainstreaming HIV/AIDS issues in sector programming as well as national planning and development processes.  Like gender inequality, HIV/AIDS is also recognized as an issue that has a social and economic impact across individuals, communities and sectors. The joint guidelines recognize the linkages and lessons learned from other mainstreaming strategies. It states that while its focus is on HIV and AIDS ―the approaches and methods offered can easily be adapted for mainstreaming other cross-cutting concerns, particularly gender and poverty‖ (UN Guidelines, 2005:10).  The definition of HIV mainstreaming agreed between the three agencies reads: ―Mainstreaming AIDS is a process that enables development actors to address the causes and effects of AIDS in an effective and sustained manner, both through their usual work and within their workplace.‖ So like gender mainstreaming, AIDS mainstreaming argues that a critical component is promoting transformative change within the workplace. The goal of mainstreaming AIDS is to address both the ―direct and indirect aspects of HIV and AIDS within the context of the normal functions of an organization or community.‖ The document lays out lessons of mainstreaming learned from previous experience and practice and recognizes that it is a ―continuous process 38  that requires ―commitment to long-term institutional transformation that changes norms, values and systems (UN Guidelines, 2005: 25). The question of norms and systems as laid out by this document aligns with the conclusions of this study that point to both the importance of normative commitment to the goal to be mainstreamed and the standard operating procedures (or ―systems‖ as the HIV document terms it) to implementing mainstreaming policies. Despite this recognition, in subsequent text the HIV policy does not list ―normative commitment‖ as one of the ―key capacities‖ for staff. It does however mention the importance of the ―motivation‖ of staff in order to be able to fulfill their roles and tasks (UN Guidelines, 2005: 29). However the authors do not question what the basis is for ―motivation.‖ This dissertation argues that normative values are the motivation to action for implementers and practitioners. In addition, in the annex of ―indicators of mainstreaming‖ two of the measurements are ―attitudes and behavior of staff regarding HIV and AIDS and sexual behavior‖ and ―attitudes and behavior of staff regarding people living with HIV and AIDS and stigma‖ (UN Guidelines, 2005: 92). Although not explained or developed further at any point, these two indicators make the case that the issue underpinning the implementation of this strategy is the attitudes of staff towards HIV and people living with HIV. Yet at no point in the UN‘s gender mainstreaming policy is there a similar question of measuring the ―attitude of staff towards women‘ rights or gender equality‖.  Both issues (HIV/AIDS and gender) touch on deeply ingrained emotional responses and internalized values related to issues such as women‘s rights, feminism, sexual behavior, homosexuality, and illicit 39  drug use. In addition HIV infection in many regions of the world is closely related to the gendered power imbalance between men and women when it comes to negotiating sexual relations. The indicators are indirectly recognizing that a programme manager‘s attitude towards these background issues will affect the zeal with which she/he implements a mainstreaming/crosscutting policy.  A 2010 evaluation of the UN‘s HIV mainstreaming strategy reached conclusions similar to UNDP‘s 2006 evaluation of gender mainstreaming mentioned earlier in the study. The HIV mainstreaming evaluation found that there was ―weak understanding of programmer‘s responsibilities and accountability" (Butcher, 2010: 18) and that the positive efforts that had been made were the result of a handful of ―committed‖ staff (Butcher, 2010; 5). This happened because the programme staff lost interest over time due to lack of perceived ownership and at times lack of agreement that HIV mainstreaming was related to their thematic development work, which was usually fighting poverty (Butcher, 2010: 20). While the evaluation, like the UNDP 2005 gender evaluation, did not go in-depth into looking at the factors behind the achievements and gaps, it is clear that many of the general factors resonate between the two documents.  Other issues that the UN is attempting to mainstream include disabilities and environment. The disabilities background note E/CN.5/2008/6 was prepared by ECOSOC in 2007 with the goal of replicating the gender and HIV mainstreaming discourse to deal with the rights of people with disabilities. The environment mainstreaming discourse was formalized in 2004 in the UNDP Environmental Mainstreaming Strategy, which is described as a strategy for 40  enhanced environmental soundness and sustainability in UNDP policies, programmes, and operational processes. The 70-page strategy outlines the manner in which environmental issues are to be mainstreamed defining it as ―the integration of environmental considerations into UNDP‘s policies, programming and operations to ensure the coherence and sustainability of our mission and practices‖ (UNDP Environmental Mainstreaming Strategy, 2004:9). Similar to the above examples, the environment document also outlines what it perceives as the lessons learned from gender mainstreaming that can be applied in this case, one of which is the requirement of having staff with expertise and that ―care‖ about the issue (UNDP, 2004:18). This question of what ―caring‖ means is not elaborated upon, although the authors must have had reason to include this concept of ―caring‖, or value-based behavior, in their document. Environmental issues can also be an emotional question for many and in this respect is not unlike gender or HIV/AIDS. The issue of climate change for example has a similar discourse between ―believers and non-believers‖ as does gender equality or AIDS. Despite the overwhelming scientific data on climate change, survey data continues to show that significant numbers of people do not ―believe‖ the science (Diethelm and McKee, 2009).  The Gendered Nature of the International System, Organizations and Policy Another critical aspect to understanding the policy process and building a holistic explanation of adoption and implementation is integrating an understanding of the gendered nature of the international system, IOs and the discrete causal mechanisms that affect policy adoption and policy 41  implementation. Feminist political theory can be a useful tool in building this understanding and has made an important contribution over the years to analyzing how the structures, organizations, policies and politics that are the subjects of research in international relations and comparative politics are in fact gendered. ―Gendered‖ refers to the disproportionate reflection of the experiences and priorities of either men or women.  Feminist theories of international relations (IR) examine the gendered nature of the field and expose the hidden and implicit power balances. Feminist scholars of IR argue that in the past state-centric security and conflict were prioritized in international relations research due to the narrow worldview and experiences of the scholars that have dominated this field over the past century (Jahan, 1995; Mohanty, 1988). Increasing attention to more diverse (including women‘s, minorities, non-Western) priorities and experiences of security contributed to this paradigm shift over the years from traditional security to non-state based security concepts (Tickner, 1997; Enloe, 2000). International organizations reflect the gendered balance of power identified by feminist political theory. Within the UN, the top echelons of decision making are dominated by men: the Secretary General, the Assistant Secretary Generals, the Special Representatives and the Director levels. Over the past three years only 25 per cent of senior level management positions (D1 and above) have been held by women, 30 per cent of middle and lower management (P3-P5) and 55 per cent of support staff (Secretary General‘s Report on the Improvement of the Status of Women in the UN system, 2010). The lack of 42  qualified women is an easy excuse that does not withstand scrutiny; when a special effort was made in 2010 to actively recruit women to the rank of ―Special Representative of the Secretary General‖ candidates were easily recognized and promoted (SG Report on Mainstreaming, 2011).  The gender distribution becomes even more unbalanced when we examine the UN‘s development work thematically – women employees are more prevalent in social inclusion and human rights work, while they are much less prevalent in security sector reform, justice work or peacekeeping (SG Report on Status of Women, 2010). The funding structure of the UN is also gendered. Research conducted by UNIFEM based on data from 2010 and again in 2012 showed that consistently less than 8% of the money contributed towards Multi-Partner Trust Funds in post-conflict countries was gender mainstreamed (mentioned gender in its activities and indicators). 45% of these mainstreamed funds were spent in the human rights and social protection sector, which makes up the smallest proportion of overall post-conflict funding (only 9% of all post-conflict funds), while the least amount of these mainstreamed funds (only 2%) was in the economic empowerment and employment sector which makes up 58% of all post-conflict funding (Cueva Beteta, 2012: 5-8). This shows that women's issues are still considered from a "protection" perspective (which tends to deny women their agency) and not an equal participation and equal opportunity to employment perspective. 43  Scholars of gender and organizational change such as Kardam (1991), Rao et al (1999), Staudt (1998) and Goetz (1992) argue that an organization that is gender biased in its operations is incapable of producing gender equality outcomes in its development work because it is itself perpetuating a ―deep structure‖ of inequality. The gender biased nature of international development organizations (such as the UN) means that not only is the organization unable to respond to the needs and priorities of women employees, it also cannot respond to the development needs and priorities of women beneficiaries. Without addressing the discrimination of these deep structures of inequality in IOs, as well as in the international system as a whole, these scholars argue gender mainstreaming cannot be successful:  Trying to ‗add gender‘ into the structure and work of organizations is like trying to add the idea that the world is round to the idea that the world is flat. Rather we must rethink the whole matter. We need first to re-conceptualize what an organization is, then we can re-invent it… in line with our vision of gender justice and racial equality integrated with sustainable development. (Rao et al, 1999: 3) Organizations within the private sector are also gendered. Kantor (1993) studied men and women in corporate organizations and argued that the careers of women are determined by the distribution of power and powerlessness within the organization and that success depends on opportunities, access to power and numbers of women or minorities. The growth of a ―masculine‖ work ethic and the ―feminization‖ of clerical work was a function of the prevalent patriarchal 44  consciousness and behaviour in the corporate organization, which was itself a function of the patriarchal consciousness of society. Despite their contributions to the fields of IR and comparative politics some feminist scholars (Butler, 2006; Gilligan, 1982) have disengaged from debates due to their critical view of the value of empirics and positivism. This disengagement was challenged by Amy Mazur (2002) who called on feminists to re-engage with policy studies using an empirical, comparative, feminist lens (Mazur, 2002: vi-1). The goal of Mazur‘s Feminist Comparative Policy is to reveal the gendered nature of policy processes and use this information to improve predictions of outcomes. This dissertation agrees that understanding the gendered nature of organizations and their staffing structures is critical to the study of non-linear policy processes in international development organizations. A gender perspective contributes to building a holistic understanding of why certain policies are implemented while others are not. The prediction that individual values have an important effect on policy implementation should not be surprising; it seems intuitive that a traditional society with strongly gendered norms regarding acceptable behaviour and attitudes constitutes a different policy context for implementation of gender mainstreaming policy than a society with less stringent norms regarding behaviour. One of the reasons that gender mainstreaming has been so poorly implemented is very likely linked to the individual norms of the practitioner responsible for implementation. This study argues that the gendered approach to policy processes will reveal more information on the policy context and mechanisms than an approach 45  that does not take into account a gender perspective of policy adoption, policy context and police implementation.  This same approach is taken by Mazur who argues that any study of public policy must take into account feminist theory in order to understand the path of policy formulation, implementation and effectiveness (2002). In other words, when studying public policy from a gender perspective, we must identify whether the policy has recognized that women and men have different needs, whether the policy affects women and men differently, whether different actors, such as the women‘s movement, have influenced the policy process from the outside, and whether feminist policy makers within the organization (i.e. ―femocrats12‖) can promote feminist goals (Mazur, 2002: 14).  One important result of Mazur‘s approach has been the recognition that a policy addressing gender issues may in fact not be a feminist policy, i.e. it may not be having a positive effect on gender equality and women‘s empowerment. In order for a policy to have a positive impact it must be directly addressing most of the following issues: the improvement of women‘s status in parity with men‘s; the reduction of patriarchy; issues typically seen as the ―private‖ sphere; both men‘s and women‘s needs; and be attributed to a recognizable feminist group/movement/actor (Mazur, 2002: 31). This dissertation reveals that these                                             12 Anna Yeatman's work on femocrats describes in detail what a ―femocrat‖ is and the tensions she faces in her daily work: The femocrat is a civil servant who faces a fundamental contradiction between his/her position and ideals. The maintenance of their position depends upon their ability to participate appropriately in the male-dominated techno-bureaucratic culture of the public administration and at the same time to remain accountable to the women's movement. Thus they must maintain a believable fiction of neutrality while simultaneously working in the interests of a politicized women‘s movement. (Yeatman, 1990: 118-119)  46  criteria can shed important light on gender mainstreaming‘s lack of impact; over the past 15 years mainstreaming policy has been overwhelmingly interpreted as the need to add women participants to a particular programme or increase women in senior management. It has not permeated the mind of the average development practitioner that at the core of gender mainstreaming is challenging patriarchy and addressing inequality in both public and private spheres (Tiessen, 2005; Lang, 2009; Njenga et al, 2011; Zalewski, 2010). The puzzle remains of how do the key individuals – the practitioners implementing mainstreaming policies – begin to understand this and integrate it into their work.  Ultimately Mazur‘s argument, which this dissertation agrees with, is that we cannot understand policy processes in general without understanding their gendered nature. This means that understanding gender mainstreaming is impossible without understanding the gendered nature of the policy context within which it exists and this gendered order is present both in the system in which organizations exist (the state-centric system), as well as within organizations themselves - their intra-organizational structure. Understanding the gendered nature of policy processes makes the research puzzle of why gender mainstreaming as a policy continues despite its ineffectiveness to achieve results even more interesting. It leads to other interesting questions as well such as why a gender policy would be adopted by a manager who is not personally committed to gender equality and women‘s empowerment or who has no incentive to adopt the policy. Given the connections between international development organizations and the international 47  professional staff that work and rotate among agencies world polity theory stands out as particularly useful in explaining this phenomenon of policy adoption – as will be discussed further below.  Gender Mainstreaming There are a great deal more academic studies examining gender mainstreaming in national contexts rather than in international development organizations. In part this is due to the multiplicity of gender mainstreaming initiatives in the European Union and its Member States (such as Woodward, 2008; Bacchi and Eveline, 2009; Kantola, 2010; Bendl and Schmidt, 2013; David and Guerrina, 2013; Lang, 2009; Lavena and Riccucci, 2012). The focus of these studies is largely on policy formulation and adoption. Studies of gender mainstreaming in international development organizations tend to appear in the form of practitioner reports, donor agency guidelines, and policy papers, which focus on implementation lessons. Few studies bring together policy adoption and policy implementation in an attempt to understand interaction between the two or conduct academic theory building and testing on how a crosscutting policy such as gender mainstreaming is adopted and implemented.  For example, the UN‘s most recent practitioner report on gender mainstreaming is the 2011 Report of the Secretary General (E/2011/114) on ―Mainstreaming a gender perspective into all policies and programmes in the United Nations system‖ and was an opportunity for the UN to reflect on the difficulties it had encountered over the past 15 years.  The report stated in its opening paragraph that  48  the recent evaluation of gender mainstreaming in the Secretariat (see A/65/266) found that the shortfalls were in practice rather than policy and that the lack of alignment between policy and practice posed a risk to the commitment of the United Nations to promoting gender mainstreaming as a strategy for achieving gender equality goals. While the report was one of the first examples within the UN to recognize with strong language that implementation was not happening, it did not venture any  predictions about what mechanisms were preventing implementation or why practitioners were ignoring policy.  Reflections by the report on the gaps and challenges reached these conclusions (UNSG Report on Mainstreaming, 2011: 15): 1. Lack of consistent integration of gender perspectives in strategic plans;  2. Lack of staff with gender expertise and their strategic placement within institutions;  3. Limited application and/or access to data disaggregated by sex and gender-sensitive indicators; 4. Limited funding; and  5. Lack of accountability mechanisms The report also discusses the continuing low proportion of women in the UN staff structure at all levels and the lack of progress in increasing the number of women: 49  Achieving gender balance in staffing across the United Nations system remains elusive. Based on statistics provided by 29 of 31 entities of the United Nations system…the average annual increase was less than one percentage point per year for all levels (other than Assistant Secretary General level). The average annual increase over the two-year period ranged from a high of 0.95 percentage points (P-2 level) to a low of 0.6 percentage points (P-3). The D-2 and D-1 levels registered an increase of only 0.5 and 0.6 percentage points, respectively. With regard to Resident Coordinators, the representation of women increased 6.3 percentage points to 37.1 per cent during the two-year period (UN SG Report on mainstreaming, 7). Scholars that have examined how international development organizations develop gender related policy priorities and how these priorities subsequently translate into action have looked at several different aspects of the debate. These include the potential of gender mainstreaming to be transformative and agenda-setting within international development and the proliferation of international development professionals (Jahan, 1995); the disconnect between policy and programming due to lack of technical knowledge and an over focus on ―checklists‖ (Rathgeber, 2005; Phillips, 2005); and an overreliance on gender mainstreaming as a technical approach at the expense of its feminist normative underpinnings (Tiessen, 2005; Lang, 2009; Njenga et al, 2011; Zalewski, 2010). This study argues that there are missing pieces to the existing research that need to be filled in – in particular understanding the impact and role of international development professionals in more detail as the disconnect 50  described by Rathgeber emanates from a lack of normative commitment and not a lack of technical knowledge. The major themes of existing studies on gender mainstreaming have focused on examining the accountability of development to women, institutionalizing gender equality into national, regional and international governance mechanisms, and examining the roles played by individual agents as well as the women‘s movement (Goetz, 1997; Jackson and Pearson, 1998). These studies originate from a development perspective as opposed to an international organizations policy process perspective (for examples of the former see Beneria, 2001; Jackson and Pearson, 1998). In these studies gender mainstreaming is put in the context of economic development theory, as opposed to international organizations theory. This compartmental focus has prevented theory building on the impact of the policy context and policy process for crosscutting mainstreaming policies in international organizations.  The use of gender mainstreaming as a concept became prominent following the 1995 Fourth World Conference on Women in Beijing. The Beijing conference shifted the discourse away from women as a focal concept and towards gender. This shift reflected the idea that equality could not be reached only through improving services to women, but that fundamental changes must occur in the relationship between men, women and the social structure (Rathgeber, 2005: 582).  51  The new approach of ―gender mainstreaming‖ was supposed to be in stark contrast to the past approaches of ―Women in Development‖ and ―Gender in Development‖. Gender issues in the development field can be traced back more than forty years starting with Boserup‘s groundbreaking study that revealed the extent to which women had been left out of development projects (Boserup, 1970). Scholars and practitioners began to increasingly discuss the differences between women‘s and men‘s lives in developing countries in the 1970s, culminating in the first UN World Conference on Women in Mexico in 1975 (Rathgeber, 2005). These earliest concerns with women‘s lives in developing countries focused on issues of education and employment opportunities, and were pushed mostly by policymakers within the existing development institutions who recognized the problem on a personal level. ―Women in Development‖ (WiD) refers to this first concern with improving women‘s material conditions through development programs (Goetz, 1997). The WiD approach continued until the mid-1980s, when concerns grew with the recognition that the material situation of women was improving very slowly despite all the efforts (Steans, 2006). A reassessment of WiD policies revealed that they focused on access of women to development programs, but did not follow up to assess the actual impact of programming on the social, cultural or economic obstacles and constraints that women faced.  In addition the WiD approach was criticized by feminist scholars from developing countries who argued that it only represented the views of Western feminists who were trying to build an ―international feminist movement‖ and were 52  not interested in the lived experiences of women in the ―third world.‖ They argued that liberal western feminists saw women from developing countries as passive and homogenous victims, without according them their due agency or understanding their differences (Kabeer, 1994; Mohanty, 1988; Rao et al, 1999). In response to this problem the ―Gender in Development‖ (GiD) approach arose. GiD argued that the material position of women would not change as long as the uneven power distribution of social relations persisted (Goetz, 1997). Consequently, this approach stressed that development programs must not only focus on the economic impact, but also have social and cultural dimensions. These types of development projects took the form of standalone projects, not integrated with other development programming, which focused on increasing women's empowerment in the political, economic or social spheres. It soon became clear however, that similar obstacles hampered both the WiD and GiD approaches to improving the lives of women in developing countries. Programs were not designed for ease of access by women and the institutional environment itself was not knowledgeable or committed.  The gendered nature of development organizations was a huge obstacle to programming (Goetz, 1997). Given their own patriarchal natures these development organizations were unclear on how they should be formulating and implementing development programmes that would address the patriarchal nature of the societies they were working in. It was in response to this awareness that Gender Mainstreaming emerged as a ―new‖ approach and included an understanding that the operations of IOs should also be a target for 53  transformative change and not only programming. Operations refers to the internal processes of international organizations, including human resources, finances, logistics, etc., while programming refers to the external work an organization does on development or other issues as per its mandate. The idea of ―mainstreaming‖ first made an appearance in international development discourse in the Forward-Looking Strategies for the Advancement of Women adopted at the Third World Conference on Women in Nairobi in 1985. One article of the document called for the ―effective participation of women in development [to] be integrated in the formulation and implementation of mainstream programs and projects‖ (United Nations, 1985). This concept slowly spread in the policy discourse over the next decade until the Fourth World Conference for Women in Beijing in 1995. Gender mainstreaming called for a holistic approach; arguing that a gender analysis should be incorporated systematically in all policy considerations, across all programming clusters (not just those fields traditionally associated with women in development such as micro-finance, education or health) and throughout the internal structure of all international organizations. Advocates argued that women‘s outcomes would improve only by holistically dealing with the deep-rooted structural causes of women‘s inequality and the gendered nature of societal power distributions.  The resulting Beijing Declaration and Platform for Action stated that  54  Governments and other actors should promote an active and visible policy of mainstreaming a gender perspective into all policies and programmes, so that, before decisions are taken, an analysis is made of the effects on women and men, respectively (Chp IV, B81). The Beijing Conference was seen as a watershed moment in the international development sphere. Since then the very word Beijing has taken on a sine qua non characteristic for international development professionals, despite the interesting fact that in the author's experience within this same network few of these actors seem in fact to be able to explain the term gender mainstreaming or the contents of the Beijing Platform for Action.  Gender mainstreaming in the UN was cemented shortly after, in 1997, through a UN Economic and Social Council Agreed Conclusions (E/1997/L.30) that gave the definition of gender mainstreaming provided earlier in Chapter One.  After the Beijing conference both government and non-government actors began to experiment with ways they could apply the novel concept of gender mainstreaming (GM) in practice. At first, GM was hailed as a revolutionary and innovative concept in the struggle for gender equality (True and Mintrom 2001; Woodward, 2008). To some it represented the evolution from liberal feminist strategies, such as positive discrimination and legal equality (Beveridge et al, 2000), towards a more critical feminist perspective advocating changes to underlying social relationships, attitudes and power structures. However, over time it became clear that both the theory and the concept of GM were under developed (Woodward, 2003; Bendl and Schmidt, 2013 ) and that resistance 55  existed, particularly towards the aspects of this approach that dealt with changing power structures within organizations, and not just promoting the needs of women in developing countries (Tiessen, 2005).  UN development agencies were among the most important actors to adopt the mainstreaming policy during this period (Winslow, 2009). In theory it meant that all programming should be tested against women‘s needs and priorities. The first mainstreaming strategies in UNDP were developed in the late 1990s, while the first ―toolkit‖ for staff was published in 2001 (UNDP, Gender Mainstreaming in Practice: A Toolkit, 2001).  Studies evaluating gender mainstreaming have looked at implementation in both development (Rathgeber, 2005; Jahan, 1995, Moser and Moser, 2005; Eyben, 2010; Njenga et al, 2011) and non-development contexts (Woodward, 2008; Bacchi and Eveline, 2009; Page, 2011; Meier and Lombardo, 2013; Bendl and Schmidt, 2013). Although both contexts are characterized by gender related power imbalances in their organizational, institutional and contextual environment, gender mainstreaming in the development context may face additional challenges. The lack of sex disaggregated data in developing countries alone impedes conducting the comprehensive gender analysis necessary for developing informed policy that takes into account women‘s and men‘s different needs and priorities.  And although western countries still have significant patriarchal influences in their societies, they do have a higher rate of acceptance of gender equality as a universal value in both the expression of direct and 56  indirect attitudes13 (Pew Research Center, 2010). The Scandinavian countries in particular have always placed highest in opinion polls on support for gender equality in society (Sainsbury and Bergqvist, 2009) and western European countries overall are highest in the prevalence of gender egalitarian attitudes among a survey of 61 countries in the late 1990s (Inglehart and Norris, 2003: 33). Lee-Gosselin et al argue in their 2013 article on gender mainstreaming that ―very few publications based on empirical data provide an in-depth study of the terms of implementation and resistances from actual experiences aiming at institutionalizing gender mainstreaming within organizations. The existing literature does not answer all the research questions (Lee-Gosselin, 2013; 472). The biggest gaps in the research on gender mainstreaming identified by Caglar (2013) include understanding the process of ―institutional knowledge production‖ and whether this happens through normative processes, scientific knowledge or day to day knowledge. Caglar finds that identifying the knowledge and values that underlie gender mainstreaming is critical to understanding what influences the discourse and implementation of gender mainstreaming. This dissertation is one step in understanding how knowledge and values may influence the adoption and implementation of different types of mainstreaming policies.                                               13 In a survey of citizens in 22 states the Pew Research Center found that in European countries more respondents felt that women should have equal rights (95 per cent and above) compared to sub-Saharan African countries (73 per cent and below). The survey also asked whether men should be given priority to jobs in economically difficult times and only 20 per cent of respondents in European countries said yes, while for other countries approximately 70 per cent of respondents felt  yes.  57  Lavena and Riccucci also point out a major gap in the gender mainstreaming agenda when they argue that the theoretical lenses with the most usefulness in explaining gender mainstreaming are organizational theory, social movement theory and feminist theory (Lavena and Riccucci, 2012: 123). This dissertation argues that part of the reason there has been so much research on gender mainstreaming but with such little to show in terms of progress is due to the failure to integrate theories from the international relations and public policy fields. The theory this dissertation builds is thus drawn from the international relations literature. Building a Theory of Mainstreaming within the International Policy Process This dissertation argues that discrete causal mechanisms are responsible for policy adoption and policy implementation and that no existing theory from the fields of IR, feminism or organizational change can on its own explain the policy process of mainstreaming in international organizations. Within the context of international development organizations (where few senior managers seem to stray from the discourse despite the lack of sanctions or incentives actually tied to gender mainstreaming and with day to day implementation firmly in the hands of middle management) none of these theories have the explanatory power to describe the different dynamics at play in adoption and implementation. Based on the author‘s own observations over recent years as a member of the network of international development professionals within international organizations, the proliferation of gender mainstreaming policies cannot be 58  explained by a sudden appearance of many managers with wholehearted commitment to women‘s empowerment. Nor has there been any clear incentive tied to these policies to encourage managers to adopt them – despite fifteen years of existence in the UN system gender mainstreaming even today remains overwhelmingly lacking in accountability or incentive. Instead this dissertation argues that the crystallization of gender mainstreaming policies is due to the influence of the world polity over the past six decades and the increase in value placed by world culture on women‘s education and status (Berkovitch 1999). However this can only explain the first stage in the policy process – adoption. Investigating why a mainstreaming policy is actively implemented is a different casual mechanism and world polity‘s main tenets – decoupling between words and action, isomorphism of policies and structures and diffusion of values through networks – cannot explain why some individuals do indeed chose to implement mainstreaming.  The theory proposed and supported by the evidence gathered is that while policy adoption in international organizations is influenced by the diffusion of world polity values that are perpetuated through non-rational mimicry of attitudes, actual policy implementation depends on the normative values of the middle management practitioners actually responsible for implementation. These two causal mechanisms will be examined in detail in the sections below. Policy Adoption The World Polity approach hypothesizes that IO policy choices are a result of world cultural influences manifested through an isomorphism of discourse and 59  structures perpetuated and diffused through non-rational mimicry. This approach is operationalized through an analysis of the definitions, principles, purposes and modes of action that constitute and motivate actors (including IOs). In their 1999 volume Constructing World Culture Boli and Thomas describe the effects and influences on states and IOs of world culture through the rationalization of intellectual, technical and economic goals. They argue that goals can be diffuse, such as world peace or human rights, or can be highly specific and technical in nature.  The increase in international organizations and the diffusion of world culture have been statistically linked to values such as human rights (Bonacker et al, 2011), women‘s rights (Berkovitch and Bradley, 1999), ending child labour (Clark, 2011) and environmental protection (Shandra, 2007). Boli and Thomas (2005) argue a world polity has successively crystallized since the nineteenth century, with the post-World War II period marking the strongest intensity of its shaping. To support these claims, qualitative as well as quantitative longitudinal analyses have been conducted of global institutions for environmental protection (Meyer et al., 1997), science and culture (Finnemore, 1993), women‘s rights (Berkovitch and Bradley, 1999), and human rights (Elliot, 2007; Koenig, 2008). This dissertation builds on the agreement among scholars that gender mainstreaming has become ―a symbol of modernity‖ and ―part of accepted wisdom‖ (Daly, 2005, p. 440) and actually delves deeper in order to identify what is the causal mechanism behind this accepted view and what its implications are. This leads to the argument presented in this research that the rhetoric and 60  attitudes related to gender mainstreaming will be similar in staff with experience around the globe in international organizations, while staff without international experience who have not had this exposure to the world culture or ―world polity‖ will not display this same isomorphism in their views. The dissertation also goes one important step further and argues for an individual mechanism of diffusion, as the world polity approach is often criticized for its lack of an argument on how values are transmitted.   Isomorphism World polity theory argues that world culture shapes the policy choices and characteristics of IOs (but also of other actors such as nation-states, domestic and international NGOs, transnational corporations, etc.) and that the identities, goals, purposes, means and causal logics of all these actors display a striking isomorphism  (Boli and Thomas, 1999: 3. Beckfield 2008) in terms of identical underlying commitments to the principles of universalism, rationalism and individualism (Meyer et al, 1997; O‘Riain, 2000; Beckfield, 2003; Drori et al, 2006; Beckfield, 2008). This isomorphism  of identities, goals and purposes operates through a ―flattening‖ process of diffusion (Beckfield, 2008) and IOs and INGOs spread this world culture through their staff not as a result of rational choices, but rather through a process of non-rational mimicry (Boli and Thomas, 1999: 9; Bonacker et al, 2011).  These diffuse world culture values are absorbed by actors through reiterative discourse and expressed as similar policies across organizations (Meyer et al, 1997; O Riain, 2000; Beckfield, 2003; Drori et al, 2006; Beckfield, 2008). Breaking down the policy process allows the research to 61  consider that this diffusion has an impact on policy adoption, but given the ―decoupling‖ that world polity identifies it seems clear that a different mechanism must be found to explain the implementation of mainstreaming where it does happen successfully by individuals.  As mentioned above Boli and Thomas identify a structural isomorphism in the central elements of the world culture. These elements are focused around the principles of  universalism, individualism and rationality (Boli and Thomas, 1999: 242-245). Like other cultures, the world culture is embedded in the organizations operating within its sphere (in this case at the global level) and transmitted through the individuals who work and live in the world polity, described by Boli and Thomas as a certain elite: politicians, diplomats, businesspeople, travelers, and international development professionals, but most explicitly through the discourse of intellectuals, policy analysts, and academics (Boli and Thomas, 1997). Boli argues that after World War II a recognizable global community emerged that came from international study programs funded by large philanthropies with a strong liberal perspective; a global culture that has a pro-market perspective while also propagating concern for the poor and restrictions on the state (Boli and Thomas, 1999: 243; Beckfield, 2003: 418; O Riain, 2000). It was organizations such as the Ford, Carnegie and Rockefeller Foundations that established and promoted this dominant liberal discourse of global culture through educational programs. Boli and Thomas explain further: 62  Since the 1960s, this combination of schooling and work experience has produced a specialized cadre of international-development professionals who spend the bulk of their working lives in a series of assignments in global metropolises and low-income country capitals. Development professionals often become somewhat detached from their countries of origin; their perspectives and concerns may be quite different from those of legislators and ordinary citizens who fund international-development organizations. Given their distance from the funders, they face significant opportunities and temptations to exercise personal and professional prerogatives. The policies and actions of international-development organizations run by international-development professionals therefore should not be assumed to represent the intentions of their nation-state or individual funders, as demonstrated in the annual struggles between the U.S. Senate's Foreign Affairs Committee and the population office of the U.S. Agency for International Development (Boli and Thomas, 1999: 243). In his 2005 book, The Globalizers: Development Workers in Action, Jeffrey Jackson paints a similar picture of expatriate international development workers who despite their differences actually share important commonalities, including specific international professional experience and training.  This isomorphism is reflected in the policy choices of the international development professionals (in this case the expatriate senior management of UN development agencies) as they respond to world culture by further mimicking and emulating its values. Since IOs exist in a neo-liberal global culture this theory 63  would expect that the gender mainstreaming options implemented by IOs would be those that fit into this neo-liberal paradigm of universalism, individualism and rationalism. Some feminist scholars (Tiessen, 2005; Rathgeber, 2005) have argued that a liberal interpretation of gender mainstreaming is the fundamental reason for the policy‘s failure. These scholars argue that mainstreaming lost its ―transformative‖ edge during its diffusion through networks of international organizations and no longer has transforming gendered power relations at its core, rather becoming more about a liberal ―business efficiency model‖ (Page, 2011). If international development professionals are indeed influenced by global isomorphic values then we would expect to see discourse and understanding of gender mainstreaming from a liberal approach and not from a transformative feminist approach. Decoupling In the author's experience and research through participant observation there are a significant number of individuals within the community of senior international development professionals who adopt a mainstreaming policy with relatively little understanding of what it means or interest in how/whether it will be implemented. This ―decoupling‖ of policy from practice is a hallmark of world polity theory (Oh, 2011).  Decoupling refers to the ―disconnect‖ between an organization‘s formal goals and its actual activities; between its intentions and results. In other words organizations adopt the norms of world culture and translate them into policy but do not implement them or follow through (Meyer et al, 1997: 152).  For example, 64  IOs may formally support values of universal human rights, while they concretely take no steps towards implementing them, instead falling back on actions that appear to more realistically fall under the rubric of cultural relativism (Oh, 2011). Why there is this lack of follow-through, or decoupling, is explained in this dissertation by employing the non-linear policy process outlined earlier and using it to frame the investigation of causal mechanisms surrounding policy implementation as opposed to adoption.  There is an interesting paradox at work in gender mainstreaming policy at the UN. Even as we see unease at the operational level with implementing mainstreaming, the idioms of gender mainstreaming are increasingly being repeated. In the Security Council gender mainstreaming of peace and security issues has achieved traction on the policy front and mandates of peacekeeping missions and SC resolutions increasingly include gender mainstreaming language14, and SC debates and consultations include statements on the importance of women‘s participation.  However when it comes to concrete implementation, such as mandating that UN Chief Mediators MUST meet with women‘s civil society or that the Secretary-General must finally appoint a female chief mediator, no action is taken. The values of the world polity clearly influence the discourse, but not the practice, which is what Meyer, Boli and other world polity theorists would expect under ―decoupling‖ policy from action.                                              14 Of the 66 resolutions adopted by the Council in 2011, 33 (50 per cent) included references to women and peace and security issues – marking an increase from 37 per cent in 2010. (SG Report on Women, Peace and Security, 2012). 65  Another study that also looks at the intersection of the world polity values and strong local norms is Oh‘s research on dog meat consumption in South Korea in the lead up to the Olympics in 1988 and the World Cup in 2002. She finds that the different interactions between transnational civil society, politicians and the general public meant that there could be both an increase in world polity rhetoric as well as a continuing adherence to local practices – what she describes as an interesting kind of decoupling (Oh, 2011).  In regards to mainstreaming in international organizations world polity theory posits that we should see a degree of decoupling of values from action – actors should support the discourse of gender mainstreaming while at the same time putting very little action into implementation of the policy. World Polity theory does not expect this lack of follow-through to be attributable to power contests or conflict within the organization or among individuals (which is what we would expect to see between administrators in bureaucratic politics theory or among norm entrepreneurs in constructivist theory). Constructivists might also argue that the decoupling of actions from words is just an early stage in the norm socialization cycle on its way towards internalization, such as in Risse, Ropp and Sikkink's "Spiral Model" (1999). The evidence from this study however draws the distinction between the "thick" culture that results from norm socialization at an individual and local level and the "thin" culture that is created and disseminated by the world polity among elites. The normative commitment of the elites stemming from the world polity is of a different nature than the normative commitment resulting from an individual and local socialization process.  This 66  argument has also been put forward by Sending (2011) in his article on the "thin" culture produced by global diplomatic interactions, including the development sphere, vs. the "thick" culture that results in humanitarian actors due to a comprehensive set of specific underlying values. Sending argues that "thin" culture results in an attentiveness to "form," while "thick" culture results in an attentiveness to action (Sending, 2011: 643-646) and that thin culture is not just a phase on the path towards thick culture.     In world polity theory the lack of follow through also comes from a scarcity of understanding or interest in action towards the stated goals. Swiss‘ (2009) study of the decoupling phenomenon looked at what factors influenced the gap between granting women the right to stand for election and the eventual election of the first woman to parliament in 92 countries of the developing world from 1945 to 1990. His conclusion focused on the density of interactions between actors at the national and global level and found that the decoupling period was decreased when interactions were increased (Swiss, 2009: 87). In addition, gender mainstreaming strategies should be very similar across international organizations and while there may be a vanguard at the very early stages, as with the example of the ICRC and human rights, as it spreads no specific person or administration should be attributable as the motivator (Berkovitch, 1999: 496). Non-Rational Mimicry and Mechanisms of Diffusion World polity theory‘s weak spot has long been its attempt to explain the mechanisms through which world culture values are diffused. Boli and Thomas 67  attempted to do this by describing the combination of schooling and work experience that produced a specialized cadre of international-development professionals who spread world polity values in their postings and travels around the globe, interactions with other similar cadre and the resulting discursive reiterations.  Other scholars of world polity have focused on the effect of international organizations on nation states (as Beckfield, 2010) at the expense of developing the actual mechanisms by which these values diffuse and perpetuate. In more recent directions in the world polity approach Bonacker et al argue that IOs are not the actors themselves but rather that they are the ones providing ―national, organizational and individual actors with depictions of their roles and identities, and recipes for activity routines that make sense in terms of other actors‘ expectations‖ (Bonacker et al 2011:117). This is the mechanism missing from many earlier studies utilizing the world polity approach and which this dissertation believes is important to explaining the adoption of mainstreaming policies (but not their implementation).  Bonacker et al use a case study of Cambodia‘s transitional justice process to demonstrate how the world polity provided influential international and national civil society leaders with the scripts and networks to share with national leaders. These leaders then repeated the phrases not because they seemed to understand or believe them (and certainly not because they benefited from it; in fact the norm of ―victim participation‖ brought the government many complications) but rather because they were slowly internalized due to repetition 68  by this cadre with transnational links (Bonacker et al, 2011: 126).  In their conclusion Bonacker et al argue that the world polity provides a ―global political opportunity structure‖ that expresses itself in its isomorphism of values, and while this can lead to policy adoption it also empowers local actors who want to implement. This is where this dissertation departs from Bonacker et al; instead it argues that their research is missing a key characteristic of those actors that chose to take world polity values and implement them. In a recent study of the European External Action Service, David and Guerrina described the diffusion of gender mainstreaming through three primary mechanisms of diffusion: contagion (leading by example); physical interactions and reiterations of concepts by individuals; and cultural filters (construction of knowledge through shared learning) (David and Guerrina, 2013: 56). These mechanisms provide a useful point of departure for this dissertation. In their book Bankrupt: Global Lawmaking and Systemic Financial Crisis, Halliday and Caruthers (2009) describe a similar mechanism of the diffusion of values from global to local. Using the cases of China, Indonesia and Republic of Korea in the wake of the Asian financial crisis they argue that what was critical to the diffusion of international finance norms into adopted national policies was the degree of connections between senior government individuals from judiciary, politics and academia to transnational banking organizations. Most importantly Halliday and Caruthers use detailed historical analysis throughout their book of the development of transnational banking norms to argue that the values of the world polity emerge through the recursive process of repeated interactions 69  among individual national and international actors. This view of ―action neither as purely instrumental, nor exclusively norm-oriented, but rather as a dramaturgical display of adherence to institutionalized scripts‖ is also supported by the research of Koenig and Dierkes (2011) in their study of world polity and global conflict dynamics. It is intriguing that even in cases where governments act in non-participatory, non-inclusive ways they still repeat the idioms of gender mainstreaming. Countries such as Russia, China, Pakistan or India have no incentive for using these concepts in their statements before the Security Council, and yet they repeat the same discourse as all others; an indication of the power of non-rational mimicry of the world polity values. Meyer et al explain this by arguing that while national states may think they are acting out of rational self-interest they are in actuality simply acting out their role within broader ―meaning systems‖ (Meyer et al, 1997: 168). Shandra (2007) also finds that the behaviour of nation states is shaped by pressures from the wide array of international organizations and the interactive and relationship character of the world polity. Diffusion of values is thus driven by the reiteration of isomorphic frames and concepts by international development professionals or other individuals experiencing similar international interactions (Meyer et al, 1997; O‘Riain, 2000; Beckfield, 2003; Drori et al, 2006; Beckfield, 2008).  In a quantitative analysis of individuals‘ views on environmental issues Givens and Jorgenson find that those with increased contact with international environment organizations are more likely to frame their attitudes according to the script provided by these organizations. They conclude that the world polity is 70  composed of both organizations and individuals and values are perpetuated and diffused by international organizations shaping the frames of mind of individuals in their network who perpetuate these same values through recursive repetition (Givens and Jorgenson, 2013). But this ―frame‖ does not necessarily reflect in concrete actions taken on environmental protection or implementation of environmental policy by either government actors or individuals. Given that ―environmental mainstreaming‖ is also a type of cross-cutting policy within UN agencies, it seems useful to consider Givens‘ and Jorgenson's results also in this context of why an environmental mainstreaming policy may be adopted but not implemented.   While this dissertation agrees that policy adoption is NOT motivated by norms, it does argue that policy implementation IS motivated by norms, as will be discussed below.  Other scholars have examined the perceived characteristics of the world polity with a focus on whether organizations and membership are unequally distributed around the globe and whether the world polity is in fact only western ―modernity‖ (Beckfield, 2003; Beckfield, 2008; Beckfield, 2010; O Riain 2000).  While membership in IGOs by states is equally distributed across the world and the interconnections of international organizations is constantly increasing, there has been a concurrent increase in regionalization of intergovernmental organizations. However international development organizations, in particular the UN agencies, have endeavored to ensure geographic representation of their member nationalities among staff and avoid domination of individuals from the 71  west. World polity would expect to see the same non rational mimicry by all international development professionals regardless of their nationality, as long as they have had multiple international postings and exposure to world culture and reiterative discourse. Using quantitative data analysis Beckfield finds that although IGO membership is dense in terms of the quantity of states, IOs are actually very sparsely interconnected among themselves and they are not dominated by developed western states. He concludes from his analysis that although developed western states may be increasing their influence through international NGOs, membership in inter-governmental organizations is in fact on a much more equal basis than was earlier posited by scholars of all traditions (Beckfield, 2003; Beckfield, 2008; Beckfield, 2010).   Policy Implementation While policy adoption is influenced by the world polity, this dissertation argues that policy implementation is influenced by the norms, values, identities and ideas within the organization and the surrounding policy context.  Constructivism as an approach argues that international relations, organizations and policy processes are affected by the norms of agents, structures and social relationships. Identities and values are socially constructed through the processes of social interaction, shared ideas and meanings of behavior and discursive interactions and are not determined through human nature (Katzenstein, 1996; Finnemore and Sikkink, 1998; Klotz, 1995; Wendt, 1999; Debardeleben, 2012; Adler, 1997: 325; Zehfuss, 2001: 57-58). 72  Although these socially constructed norms are resistant to change, most constructivists resist assumptions of determinism; they argue norms and values change over time through the inter-subjective discursive interactions constantly taking place between actors (Meyer, 2012). Legro argues that ideas and values ―interact with other factors in specific ways to cause outcomes‖ and that a synthesis of factors typically treated as logical alternatives‖ is the most persuasive approach (Legro, 2005: 13–14).   When taking into account the gendered nature of international organizations the expectation is that ineffective policy in international organizations is caused by gender norms that do not facilitate implementation and that are not conducive towards gender equality. In addition there is the influence of actors who hold values against potentially more effective or alternative types of gender mainstreaming strategies, making policy implementation difficult or impossible (Tickner, 1997).  The important features of the constructivist explanation of policy implementation is that the middle management of practitioners in international development organizations will only implement mainstreaming when they are motivated by normative commitments to women‘s empowerment or gender equality. This means ultimately the success of any mainstreaming policy depends on norm socialization (Barnett and Finnemore, 2004; Checkel 2005; Clegg, 2010). 73  This phenomenon among UN development agency actors is described by Grabska who found that the ―personal standpoint and biases‖ (Grabska, 2011: 87) of programme staff was the single most important factor leading to programme implementation deviating from corporate strategies. Programme managers clearly have a critical role in organizational change that has been grossly under-appreciated as this dissertation argues.  Other scholars (Weick 1976, Kardam 2004) have described the phenomenon in which orders from the top of organizations or bureaucracies fail to be implemented down the chain. Beer‘s (1989) study of organizations‘ policies finds that strong support from senior management and accompanying financial resources is still not sufficient to lead to implementation, while Goetz (1996) finds that the discretion of field level staff in development projects explains why program and behavior can differ so greatly from policy. Another study of change management in a Canadian organization found that senior management could not implement their social justice policies due to ―structural chaos, individual incompetence and staff contempt‖ for the social justice cause (Trip 1999). In addition it is interesting to note that adoption and implementation differ also in so far as middle management does not have the same exposure to the world polity as senior management, and subsequently does not display the same non-rational mimicry of the discourse and are open in expressing their norms through their behaviour. This variation in rhetoric and behavior between an IO's senior political leadership and its middle management operational leadership is a critical factor to understanding the policy process. No other study has used world 74  polity theory to explain the widespread adoption of gender mainstreaming, or taken the step of disaggregating the policy process and applying a constructivist theory to understanding policy implementation.  Norm Life Cycle The characteristics and mechanisms constructivism describes and that were also examined in the framework of policy implementation include primarily norm socialization and internalization.  The question of ―principled ideas‖ or norms is explored by Finnemore and Sikkink (1998). For these scholars a crucial exercise is identifying where changes in norms come from (and thus the corresponding political changes that norms cause). They argue that ―norms evolve in a patterned life cycle and different behavioral logics dominate different segments of the life cycle‖ (Finnemore and Sikkink, 1998: 888). The Norm Life Cycle is composed of three stages. First, ―Norm Emergence‖ in which ―norm entrepreneurs‖ begin to advocate for norms out of motives such as altruism, empathy or ideational commitment using persuasion. Second, ―Norm Acceptance,‖ in which norms cascade and a dynamic of imitation is present. The motivation is a combination of pressure to conform, enhancing legitimacy and self-esteem. The mechanisms of this cascade are socialization, institutionalization and demonstration. The third stage is ―Norm Internalization‖ in which the norm is no longer a matter of debate and is incorporated legally, professionally and bureaucratically. The motive for actors is conformity and the mechanisms are habit and institutionalization.  75  Norm Entrepreneurs and Norm Socialization Norm entrepreneurs have been identified in constructivist literature for their influence on driving new agendas and actions. Keck and Sikkink examine the role of individual champions within ―transnational advocacy networks‖ and their impact on IO policy change. These transnational advocacy networks are described as ―densely connected collections of individuals and groups united across national boundaries around the pursuit of some particular set of principled ideas.‖ These individuals and the networks they operate within undertake norm socialization through issue creation and agenda-setting; and by influencing discursive positions, institutional procedures and policy change for IOs (Keck and Sikkink, 1998). The influence of norm entrepreneurs in international relations has been widely documented through a number of high profile cases. For example Finnemore discusses the International Committee of the Red Cross (ICRC) and the man who almost single-handedly brought states to bind themselves, at times unilaterally, into the Geneva Conventions, Mr. Henri Dunant (Finnemore, 1999). Johnstone describes the post-Cold War Secretary-Generals, especially Kofi Annan, who demonstrated a keen capacity to promote values and norms such as human rights and development (Johnstone, 2007: 123). He explains that the Secretary-Generals were ―actors with a cause who mobilize support for that cause and seek to have it crystallized as an accepted standard of behaviour‖ (Johnstone, 2007: 126). Haas‘ (1992) (thought not entirely a constructivist scholar) describes the importance of studying ―epistemic communities‖ in order to 76  understand the human agency at the center of systemic conditions. Networks of knowledge-based experts, i.e. ―epistemic communities,‖ propose specific policies and interests to international organizations and national administrators. The epistemic community shares a set of normative and principled beliefs which provide a valued-based rationale for action. These epistemic communities potentially have power over decision-makers as they provide interpreted knowledge and an interpretation on what the important issues are.  Keck and Sikkink point out that if the norms, values and ideas being promoted do not ―fit or resonate with the larger belief systems and real life contexts‖ of the target societies then socialization will not result (Keck and Sikkink, 1998: 20). Other constructivists such as Nadelmann (1990), Lumsdaine (1993), Finnemore (1999) and Sikkink (1998) also argue that individuals from ―shared ethical traditions‖ should be fairly susceptible to the process of norm socialization, persuasion and framing.  This dissertation argues that understanding the gendered patterns of norms and ideas held by middle management implementers provides an excellent explanation of the policy context and the causal mechanism influencing policy implementation but not policy adoption. It also argues that individuals who hold norms resistant to gender equality will not be susceptible to norm socialization and that norms on gender equality held by individuals from both within and without the organization can affect the policy choices and strategies the organization ultimately chooses for implementation. 77  The variation in attitudes towards the norms underpinning gender equality (commitment, disinterest, resistance) is critical to implementation. The norms of international development professionals can institutionalize and crystallize change in individuals and structures, which means that understanding the norms underpinning the meanings of behavior are critical to understanding the policy process. In fact some feminist studies (Taylor, 2005, Sandler, 2012) have examined how feminist advocacy coalitions external to international organizations have influenced norms and changed policy. This dissertation looked for evidence of the socialization of norms on effective policy implementation. However the mechanism of ―norm persuasion‖ was not present in the results. This is in line with the constructivist expectation described earlier that if society and individuals still hold a belief system at odds with the new norms being promoted then norm socialization will fail. Colleagues with active norm resistance were indeed impervious to norm socialization.  Alternative Explanations  The explanation proposed above by this dissertation focuses on isomorphism and norm socialization. However, there are alternative potential explanations suggested by existing IR literature that the dissertation considers in evaluating the evidence. These include the bureaucratic politics theory and organizational behavior theory.  Bureaucratic Politics Theory The bureaucratic politics approach hypothesizes that ineffective policy choices and implementation are caused by the personalities and parochial 78  preferences of ambitious actors within organizations trying to protect their power, position and mandates (Niskanen, 1971; Allison and Zelikow, 1999; Pollack, 2003; Peters, 2001; Farazmand, 2010; Marsh, 2012, Lee, 2013). While the dissertation did not find evidence that UN officials were primarily motivated by bureaucratic competition in their efforts to adopt or implement policy, it did find limited evidence that where competition between personalities and preferences existed it could act as a complicating factor on policy adoption.  The bureaucratic politics model argues that organizations are not unitary actors, but rather are divided over options, goals, interests and policy preferences. Policy preferences of different actors are influenced by their own parochial concerns, including the perceived effects of gender mainstreaming policy on their position and the essence of the organization‘s mission from their perspective. Actors engage in bargaining through ―action-channels‖ in order to secure their policy preference and final decisions are the product of competition and conflict and ―not an objective, unified cost-benefit analysis‖ (Marsh, 2012).  Scholars have described the effect of bureaucratic politics in the United States government (Niskanen, 1971; Wood, 1994; Guy, 1993; Allison, 1971), within the EU context (Pollack, 1997; Alter, 2009) and within organizations such as the UN (Lewis, 2005). Bureaucrats have a set of choices framed by their personal ambitions and their position in the bureaucratic map and they choose the option that they most prefer unless they face major constraints from their environment and masters (Niskanen, 1971: 11). In the second edition of his seminal work Allison argues that decisions are made based on bargaining among 79  officials and can lead to outcomes that are not primary for any individual in the organization but rather are due to personalities and disagreements. These officials are influenced by conceptions of organizational, personal and national goals that can be in flux (Allison and Zelikow, 1999: 255-257).  Pollack argues that this exercise of bureaucratic power is the norm with IOs (Pollack, 2003). Changes that would put power back into the hands of member states are unlikely due to the lack of unanimity usually required to dissolve or reform IOs, which means they can function with relative impunity once they have inertia. In addition, while authority is often concentrated at the top, actual expertise is located at the bottom, which creates an expectation of conflict between policymakers and policy implementers (Peters, 1989: 65).  Bowornwathana and Poocharoen (2010) point out in their study of bureaucratic politics theory that actors engage in bargaining and competition at a variety of levels – between senior managers, between senior and middle managers, and between middle managers. Without understanding the complex nature of relationships and competition a theory cannot reveal the true mechanisms influencing the policy process. Bowornwathana and Poocharoen critique the Principal – Agent approach as too simplistic in this regard.  More recent research (Lee, 2013) has focused on how bureaucrats and middle managers take advantage of asymmetrical access to information to achieve their goals. Lee argues that with changes in technology access to information has become easier and that bureaucracies no longer have a ―monopoly‖ over information. Instead the significant source of asymmetry now is 80  not in accessing information, as the principal-agent theory argues, but rather in influencing the source of information, meaning that the principal is often actually less effective at influencing the agent than external lobbying and advocacy groups (Lee, 2013: 691-693).   Stephen Lewis gives an interesting example of bureaucratic politics and its impact on gender mainstreaming in his 2005 book Race against Time. In his indictment of the ineffectiveness of gender mainstreaming Lewis argues that the incessant competition over turf and funding between the former agencies of the UN Development Fund for Women (UNIFEM) and the Division for the Advancement of Women (DAW)15 was a major factor in their policy failures.  Although UNIFEM was supposed to be the UN‘s ―operational‖ organ on gender equality and DAW the ―conceptual‖ organ, in their activities each organization consistently tried to increase its budget and prominence at the expense of the other‘s mandate (Lewis, 2005: 127). A gender analysis of bureaucratic politics leads to the expectation that if senior managers have no preference for gender equality then we would not expect them to support gender mainstreaming as it is likely to be perceived as a threat to the status quo. As Guy (1993) argues bureaucracies often exhibit ingrained biases, intentional or unintentional, against change and diversity (which includes women). Feminist scholars have also questioned whether bureaucratic politics is by its very nature oppressive to women as the ―institutional arm of male dominance‖ (Calas and Smircich, 1999; Ashcraft, 2001).                                             15 As discussed earlier these two agencies were among the four joined to form UN Women in January 2011.  81  This dissertation synthesizes all these bodies of work and hypothesizes that the senior and middle management of an organization influence policy choices based on personal preferences and against the goals or preferences of the organization‘s ―principals‖, whether these be member states or senior management. This theory assumes these individuals are rational actors that make their choices based on their own motivation of protecting the organization and their turf and career possibilities, and so policy choices on gender mainstreaming are implemented in such a way as does not threaten the IO‘s status quo or ―equilibrium‖ or in fact enhances status. If bureaucrats perceive that certain gender mainstreaming strategies are a threat to this status quo then they are less likely to be implemented.  Organizational Behaviour Theory The organizational behavior approach argues that policy adoption and policy implementation in international organizations are limited by the pathologies of international organizations that come about as a result of lack of capacity (time and resources), inefficient organizational culture and stringent adherence to rules and procedures (Allison and Zelikow, 1999: 145-6; Barnett and Finnemore, 1999; Nelson, 1995; Wade, 1996; Haas, 1990; Dijkzeul and Beigbeder, 2003). Pathologies include ―inefficiency, waste of funds on petty causes, deviation from mandate, duplication, overstaffing, slow decision-making processes, patronage and fraud‖ (Dijkzeul and Beigbeder, 2003: 2).  The main feature of the approach is that international organizations behave autonomously, although the pathologies of their behavior patterns often 82  lead to dysfunctional and inefficient behavior that hampers policy implementation. Barnett and Finnemore identify specific factors that breed resistant pathologies in IOs, some of which this research also recognizes as having a secondary influence on policy implementation. For example, the manner in which gender mainstreaming policy uses a universal approach in dealing with local contexts often leads to the flattening of diversity (and presumably ill-suited policies for local conditions). The ―normalization of deviance‖ refers to the high staff turnover in international organizations that leads to a lack of institutional knowledge and thematic expertise; a situation which becomes normalized instead of being understood as an exception.  ―Organizational insulation‖ refers to the process by which IOs develop an internal culture that is insulated from feedback; for example professional training that is rarely updated or modernized.  These mechanisms lead to inefficient implementation of policies that are trying to change the status quo – including those such as gender mainstreaming. In their edited volume on UN entities Dijkzeul and Beigbeder (2003) identify several specific pathologies of international organizations including challenges they face related to human resources, corruption and misguided programming. However they also argue that pathologies are only symptoms of underlying problems within the overall context in which international organizations exist. As such understanding policy context is a critical aspect of understanding overall why international organizations may fail at implementing certain policies.  83  Organizational insulation means that the international organization may dialogue with partners, but implementation and change is consistently defeated by a discourse of the ―rules‖. These rules may often take the form of standard operating procedures (SOPs) that limit ―information, options and action‖ (Allison and Zelikow, 1999: 392). This issue of SOPs gets to the heart of the difference between the ―logic of consequences‖ used in bureaucratic politics models and the ―logic of appropriateness‖ in the organizational behaviour approach (Allison and Zelikow, 1999: 146).  It seems plausible that SOPs, organizational inertia and transaction costs of change become an insurmountable hurdle and lead to large organizations avoiding the introduction of new plans or policies (Allison and Zelikow, 1999: 148). This ―cumulative‖ logic of appropriateness, where each subsequent decision deviates a little more from the center can lead organizations over to time to deviate substantially from effective policy, behaviour and outcomes (Allison and Zelikow, 1999: 162). If organizational culture unintentionally leads to dysfunctional behavior (Barnett and Finnemore, 1999) this can cause slow policy formation, incomplete strategy development and poor implementation. The gender mainstreaming policy within the organization should not focus on the concept and goals but only the logistics of implementation and how that fits with the ―rules,‖ and if a gender mainstreaming directive comes from above we expect to see it implemented in a boilerplate/cookie-cutter fashion without consideration for the local context and diversity.  84  Conclusion Gender equality as a goal of international development policy has evolved significantly since the 1970s. Where earlier it was seen as a specialized focus of economic development or education policy, it is now seen as a holistic approach to dealing with the inequality in men‘s and women‘s social, political and economic power relations. This holistic approach has focused on mainstreaming gender into programming and operations of international organizations with the goal of transforming power relations at all levels. In practice however the literature is in agreement that the policy has no results to show almost twenty years on. Which factors are responsible for this varies according to scholars from different theoretical approaches: from a lack of senior management interest, to lack of norms, problems of turf and mandate competition among managers or lack of a supportive organizational culture.  This dissertation has proposed a theoretical framework that breaks down the policy process into a non-linear approach where different causal mechanisms interact between the policy context, policy adoption and policy implementation.  The proposed explanation of separating policy adoption from policy implementation comes from observations throughout mainstreaming evaluations that there is a gap between policy and practice and yet no theoretical attempts to explain the significance of this phenomenon exist. This theoretical framework proposes a model of interaction for two specific causal mechanisms: on the one hand the power of isomorphism and non-rational mimicry on the interaction between the policy context and policy adoption and on the other hand the power 85  of norms and values on the interaction between the policy context and policy implementation. Alternatives have also been proposed in this theoretical framework as they can provide additional explanatory power to parts of the policy process. The findings of this research on mainstreaming policy in development organizations has significant implications for how the field of international relations explains the functioning of policy processes in international organizations. There cannot be a full understanding without considering how policy context affects actors at different levels of the organization; how policy context affects different stages of the policy process; and how different actors within organizations interact with each other.  Given the persistence of a failed policy (gender mainstreaming), it is striking that discussions of alternative policy options have not arisen within international development organizations.  If senior management were committed to gender equality we would expect to see them experimenting with alternative methods to find something more effective in achieving this goal.  Yet they continue to parrot the discourse of gender mainstreaming when there are no sanctions or rewards, no successes, and no personal gain. The last chapter of this dissertation will examine both theoretical and policy implications of these findings.  86  3. Methodology Research Design This study reviews and compares the experiences of UN development agencies in three countries in a region with similar political cultures and societal attitudes towards gender equality and women‘s rights: Albania, Bosnia and Herzegovina (BiH) and Kosovo. These three case studies were selected because they hold constant the relative general conditions of the cases (social, political, economic) while at the same time having good variation on the independent variable of the strength and presence of UNIFEM/UN Women, which represents the availability of technical expertise and leadership on gender mainstreaming to the UN system and was expected to be a factor in strength of gender mainstreaming. The outcomes of the research showed this in fact not to be the case, which has important policy implications discussed in the conclusion.  The case study approach is particularly effective for this dissertation‘s goals as it is useful in identifying new or previously omitted variables as well as examining intervening variables in order to make inferences on the causal mechanisms that may be at work (Bennet, 2004; 27). The dangers of selection bias, inherent to the case study method, are discussed later in this chapter as well as the efforts to avoid them. While the qualitative case study method is an important tool for building new theories and identifying new causal mechanisms, a future research agenda to complement this work is to use a quantitative research design to test how far the theory proposed in this study can be generalized. In particular for example it would be interesting to collect data for a 87  longitudinal survey comparing the values and norms of an individual at various stages over the course of progression of their career from local to global arena in order to examine whether norms and values change and whether a shallow world culture persists over time or eventually moves deeper into a true norm socialization process.  Sources of evidence for this dissertation include content analysis of policy documents, semi-structured interviews with key informants and participant observation conducted by the author over a seven-year period. The dissertation uses process training to analyze the evidence collected, examine the observable implications, and assess the linkages and processes operating between cause and effect and the mechanisms of interaction (Bennett, 2004; Bennett and George, 2005; King, Keohane, and Verba; 1994, 85-65; Checkel, 2004). By describing causal processes involving policy adoption and policy implementation separately, this dissertation is able to deepen understanding of policy processes in international relations and to build a theory that explains the non-linearity of the policy process.  The first stage used analysis of relevant documents (internal reports, public reports, donor reports), notes from participant observation and interview responses to establish the factors leading to initial adoption of corporate support for gender mainstreaming in each UN country team. The second stage identified used implementation reports, notes from participant observation and interviews to establish what the barriers to implementation actually were. The last stage of 88  the research investigated the discourse surrounding gender mainstreaming in each country team and asked whether the approach to the policy changed over time as internal and external evaluations16 continued to indicate failure. Content Analysis A dataset was created of references to gender mainstreaming adoption and implementation in approximately two hundred policy documents and implementation reports by UNDP and UNICEF, the two largest UN agencies in each country that do not deal specifically with gender issues (such as UN Women and UNFPA). Documents and reports fall into the following categories: country level development assistance frameworks and country level situation assessments (both setting out policy); and annual global reports, annual country reports, annual evaluation reports and individual project reports (all reporting on different aspects of implementation).   The data is disaggregated by country and year in order to examine how policy adoption and implementation of gender mainstreaming have compared to each other over time and location.  The data was also used to examine the dissemination of the concept of gender in policy documents overall. The data is used to triangulate against the evidence collected from interviews and participant observation by the author.                                              16 The most important of which was UNDP‘s 2006 critical evaluation of Gender Mainstreaming as a policy with no effect.  89  Semi-Structured Interviews  More than 75 semi-structured interviews were conducted from 2010-2013 with individuals in New York, Tirana, Sarajevo and Pristina. The study draws on several sources of evidence. Three field visits were conducted during 2010 to each of the three case study countries and from 2011-2013 interviews were conducted at UN HQ in New York. During this time, over 75 key informants were interviewed. Consent was given for approximately 80% of interviews to be taped and these were transcribed verbatim. For the remaining interviews detailed notes were compiled by the interviewer during and immediately following the interview. Transcripts were coded by the author to analyze patterns, themes, trends, categories and relationships. Responses were compared within countries as well as across countries, across agencies and by management level. The comparison of interview data provided valuable insights into the mechanisms of policy adoption and policy implementation and how both are affected by policy context. Many of the key informants asked for either all or parts of their interview to be partially anonymous (indication of only individual‘s position allowed) or completely anonymous (neither name nor position to be indicated). This is partly due to the fact that the circle of senior international UN staff in the region is relatively small and some felt more comfortable providing a critique of their agencies‘ or UN system work if their statements were not attributable. For this reason, not all statements are fully referenced with name and position and date. The information and quotations that are not fully referenced come from anonymous interviews and participant observation notes by the author. 90  Data and observations were collected through open-ended semi-structured interviews with key informants from three different sectors: major international organizations present in the country (UN country team agencies, the OSCE mission and/or the EU mission), government, bilateral donors, civil society and UN consultants. A full table of interviewees and characteristics is available in Appendix A. All levels of UN staff at the country office level were interviewed: the UN Resident Coordinator or Deputy Resident Coordinator, the UNDP Resident Representative or Deputy Resident Representative, the UN Women and UNFPA country directors, UNDP gender focal points17, UNDP communication and coordination officers and programme officers from across UN entities. From other international organizations the gender focal point or gender advisor was interviewed. Government partners interviewed were those that were identified as having responsibilities to collaborate with the UN on implementation of gender mainstreaming strategies as well as government representatives affiliated with gender equality mechanisms such as a Ministry for Women, Gender Ombudsperson, Office of Gender Equality, etc. The third type of interviewee was national civil society partners collaborating with the UN country team on the implementation of gender mainstreaming strategies. This includes not only representatives of the women‘s movement and women‘s NGOs but a sample of                                             17 Gender focal point is a term used in many organizations to denote a staff member who has been assigned a coordinating role on ―gender issues‖ on top of their regular duties. This role is added to their portfolio as an addition to existing duties.  91  other organizations that receive funding on gender related development programmes.  The interview questionnaire (see Appendix B) was designed to investigate aspects of each of the four theoretical approaches. Each interview started with preliminary questions; however the open-ended approach allowed each interviewee to also autonomously identify key issues and points as related to policy adoption and implementation. Follow-up questions by the interviewer also occurred in many of the interviews. Participant Observation  The author also engaged in participant observation of the UN country teams in BiH, Kosovo and Albania during employment in UNIFEM in Southeast Europe and UN Women in New York. Information on events, phenomena and observations were collected and recorded during this time period as private notes for the purposes of the dissertation. Responsibilities of the author centered on both policy development and programme management at the middle management level, allowing the opportunity to observe in detail the processes and mechanisms at work in the context.  Debates on the ethics of participant observation in social science research have focused on the continua between overt and covert research and the insider/outsider role of the researcher (Uldam and McCurdy, 2013; Vinten, 1994). Access as an insider can provide important insights and while there are no ethical concerns about purely overt research as an insider, it is unlikely that every single individual the author came into contact with during the course of the study 92  was aware of the research agenda. Individuals with whom the researcher had contact with over multiple occasions were aware of the employee's simultaneous role as researcher and many also provided informed consent as interviewees. Individuals that had only limited contact with the researcher may not have been aware of the research. However as Uldam and McCurdy note, covert research does not all face the same ethical dilemmas - the important distinction in their perspective is whether the research is conducted through deception (misrepresentation of intentions), as opposed to lack of disclosure (just not telling subjects they are being studied) (Uldam and McCurdy, 2013). The insights and dynamics that can be gained from insider participant observation, when conducted with sensitivity for the ethical dilemmas that may be encountered, are significant and should not be discounted (Johnson et al, 2006). The participant observation was undertaken in an overt manner, with many colleagues also providing informed consent to be interviewees for the research.  The literature on participant observation recognizes the specific role of "research as employee" (Vinten, 1994) and highlights the ethical dilemmas that are often present when the research is undertaken in a covert manner. However in this particular case the authors' role as a PhD candidate researching the topic of this dissertation while simultaneously being totally immersed in the role of employee did not lead to any ethical dilemmas.  While some traditions in political science discount research undertaken through participant observation, others have increasingly made arguments that this methodology has been dangerously undervalued in the discipline (Gillespie, 93  2011; Mazur, 2002; Burt, 1995). In the field of anthropology researchers embed themselves in a community for long periods of time to observe, collect notes and track systematic patterns to make inferences about causal mechanisms of social phenomena (see Burawoy et al. 1991; Emerson, Fretz, and Shaw 1995). Gillespie argues that political scientists need to understand better that participant observation can be a rigorous and labor intensive data collection process that can provide insight into causal mechanisms in a unique manner and is not just the recollection of anecdotes to ―color prose‖ (Gillespie, 2011:261). In addition the real world application of theory building and testing requires researchers to reach out and interact with politicians, organizations, governments and others.  The most common accusation against participant observation is that subjects change their behaviour when they know they are being observed and that there is a lack of objectivity on the part of the researcher. Gillespie addresses the first point by arguing that moving from pure observer to also being an active participant in the community under study (as the author was in this particular subject of research) allays this concern since the more participatory and the less purely observational a researcher behaves the more likely it is that the behaviour of the group they are embedded in is unadulterated (Gillespie, 2011).   Concerning the second issue – the objectivity of the researcher – feminist theorists have long argued against the expectation of objectivity by researchers in general. In their seminal works prominent feminist scholars Sandra Burt (1995) and Amy Mazur (2002) both argue that all researchers bring their own subjective assumptions to research and no researcher is objective. Feminist political theory 94  argues that context-sensitivity is a critical aspect of political science and that participatory methodologies are an important tool for research that wants to build or test theories.   Observable Implications The observable implications outlined in the previous chapter were examined against the evidence collected during the research. According to the theory proposed in the previous chapter, the research expects to uncover evidence of a degree of decoupling of values from action that cannot be explained through specific incentives to actors for lack of implementation. Actors should support the rhetoric of gender mainstreaming while at the same time putting very little action into it and without any motivation tied to conflict with one or more individuals that reflects the power hierarchy in the organization or personal incentive. This type of conflict would be an expected observable implication among managers in the bureaucratic politics approach or among norm entrepreneurs trying to get their norms institutionalized in the constructivist approach. In addition an observable implication of the theoretical framework is that no clear career or financial incentive should be identified for adopting a gender mainstreaming approach and there should be no clearly attributable source to the values on gender held by international development professionals. Gender mainstreaming strategies should all look alike (ie. appear isomorphic) and not focus on any values not part of world culture, such as social transformation or dismantling of patriarchy. It would be expected that even if local 95  implementing partners provide advice to international development professionals it will be ignored if it does not fit their isomorphic model of gender mainstreaming.   To examine the constructivist explanation, interview questions delved into the norms and values of the respondent towards gender equality, identifying their source and history and precedence to policy adoption or implementation. Questions were framed in such a way as to understand whether and how officials were socialized into new norms (Wallace et al, 2005: 24; Hooghe, 2002; Hooghe, 2005). If officials hold neo-liberal values and norms of development they would be expected to be interested in gender mainstreaming only as far as it facilitates accountability mechanisms within the neo-liberal paradigm (Phillips, 2005). Or if they hold liberal feminist norms they will perceive gender mainstreaming only in terms of legal mechanisms and quotas. However if they hold critical feminist norms then they would be expected to support the full implementation of gender mainstreaming in a transformative manner throughout the whole policy and programming cycle. To examine the bureaucratic politics explanation interviewees were asked questions targeting the nature of manager‘s preferences and knowledge on gender equality; their view on how gender equality as a goal has an impact on the organization; whether they exhibited reluctance to addressing and discussing gender as an important concept; their views on budgeting issues and decision-making processes; how comfortably they feel they can insulate themselves from outside pressures regarding gender mainstreaming; how well-organized and smoothly the international organization operates; whether it is 96  disorganized/overloaded; the relationships between the ―bottom‖ and the ―top‖ of the international organization; and conflict within the organization and with other organizations working on gender equality. International development professionals might express concern that certain aspects of gender mainstreaming strategies might take a larger budget share than is feasible, or may negatively impact the international organization‘s ability to fulfill its mandate. We might see that the international organization has an independent perspective on gender mainstreaming from national NGOs or local government partners and has no problem insulating itself from its partners‘ preferences. Wood‘s (1994) research also leads us to expect the presence of strong personal leadership in the bureaucracy/administration that influences the decision-making processes according to their preferences. Individual preferences in relation to gender mainstreaming should come across particularly strongly.  To examine the organizational behaviour explanation the evidence should point to unintentional effects of the bureaucracy such as organizational norms that flatten the diversity of gender mainstreaming strategies (i.e. taking a ―cookie cutter‖ approach) without any interest in impact by managers. Gender mainstreaming policy is chosen according to how it fits in with the organization‘s standard operating procedures as opposed to an assessment of what could be effective. And even if advocates have allies in the organization to push through their policy choices, the policy will still not be implemented without the organizational ―rules‖ and SOPs to support them. 97  Case Selection The three case study contexts share similar political, social and economic influences; all are post-communist states in southeast Europe, have a moderate Muslim religious and cultural background, and have experienced violent political instability since the collapse of communism in the 1990s. By holding constant socio-cultural factors this project will ensure that any variation in the efforts to implement gender mainstreaming is not related to wider societal factors or the commitment and will of States. In other words the differences should come from factors other than the country context in which international organizations are operating. In addition gender mainstreaming has been relatively uninvestigated in these case studies – none of the studies cited in the review of gender mainstreaming in the previous chapter touched on these three cases. These three countries share similar baselines and benchmarks of development and gender equality. Levels of economic development are low as compared to Europe (GDP per capita as measured by the World Bank for 2010 in USD was: Albania 3,677, Bosnia 4,409, and Kosovo 3,059. All three have a relatively significant per capita inflow of foreign remittances contributing towards GDP (Albania 11 per cent of overall GDP, Bosnia 12 per cent, and Kosovo 17.5 per cent) (World Bank, 2011). Unemployment is high in particular in Kosovo (Albania 12 per cent of total labor force unemployed, Bosnia and Herzegovina 23 per cent, Kosovo 45 per cent) (World Bank, 2011: 51-52) however this may not reflect the reality that in Kosovo many workers, in particular women, are employed in the informal economy.  98  General political conditions are also similar. The political systems of the three countries have the same broad characteristics (multiparty, proportional representation, history of power-sharing arrangements and political deadlock) (Jarvis, 2000; Cousens and Cater, 2001; Judah, 2002). The position of women in the three countries shows similar levels of participation in politics (parliament, ministerial posts) and all three have electoral gender quotas (IPU, 2014; QuotaProject, 2014). In addition, attitudes expressed towards women in public and family life are similar between Bosnia and Herzegovina and Albania in the last two waves of the World Values Survey. Unfortunately because of Kosovo's continued disputed status as an independent country it was neither included in the waves of the survey conducted in Serbia, nor was it surveyed independently.18 Similar percentages of sampled individuals in Albania and Bosnia agreed that men make better political leaders (in 1998 52% and 59% respectively and in 2001 50% and 33%) or that university is more important for a boy than a girl (in 1998 30% and 31% and in 200115% and 18% respectively) or that both a husband and wife should contribute to income (1998 97% and 92% respectively and in 2001 96% and 94% respectively).19 It is not expected that Kosovo would present variation from these trends even if there were data available given its shared nationality and culture with Albania and its shared geographic and political historical ties with Bosnia during Yugoslavia.                                              18 World Values Survey. Technical Specifications by Country. http://www.wvsevsdb.com/wvs/WVSTechnical.jsp?Idioma=I 19 World Values Survey. Online Data Analysis. http://www.wvsevsdb.com/wvs/WVSAnalize.jsp?Idioma=I 99  Because the contexts are similar this means we can examine the mechanisms linked to policy processes without concern for whether the results are attributable to other case-specific/contextual factors. Albania UN development agencies arrived in Albania in 1991, after the collapse of communism and amid severe economic, political and social instability. During this transition chaos and instability grew, fuelled by rampant corruption and financial instability and regular incidents of political violence and bloodshed. Instability culminated between 1996 and1998 with widespread violence and assassinations of leading political figures (Jarvis, 2000). In March 1997 rioters attacked government offices and military barracks and seized weapons ranging from Kalashnikovs to tanks. The government dissolved, the police and military did not act to restore order and insurgents were able to quickly take over most major cities in the south of the country. The insurrection quickly spread to the north, including the capital Tirana, and Albania found itself in a state of anarchy and conflict with other countries scrambling to evacuate their nationals to safety (Jarvis, 2000).  Given the institutional breakdown the country was experiencing the President quickly agreed to a power-sharing arrangement with elements of the opposition and the new government immediately requested the UN‘s assistance in re-establishing its jurisdiction and peace and security. The Security Council voted on Resolution 1101 to send a multinational peacekeeping force to Albania for a mandated period of three months to re-establish security, distribute 100  humanitarian aid and assist the government‘s establishment of jurisdiction in all regions of the country (UNSCR 1101 (1997)). The mission deployed in April 1997, however incidents of political violence, assassinations and social chaos continued throughout 1997 and 1998. Approximately 2000 people were killed during this period of conflict. In addition to the internal instability, from 1998 to 1999 increasing numbers of Kosovar Albanian refugees arrived in the north of the country, fleeing the campaign of violence of the Serbian police and paramilitaries in Kosovo. The pressing problem of what to do with 450 000 refugees from Kosovo when the conflict there peaked in 1999 (UNHCR Albania, 2011) also destabilized Albania. UNHCR established its presence in the country in 1992 when it first began assisting the government to deal with asylum seekers, refugees and migrants deported and returned from Italy and other European countries and internally displaced persons due to the domestic conflict (UN Albania, 2011). Every Albanian election since the 1997 conflict has resulted in violence between the main political rivals. Campaigns for local elections in 2011 resulted in riots, shootings and deaths and analysts agree that each election brings violence due to the depth of political divides (International Crisis Group, 2011) and the 2013 national elections witnessed the shooting of one candidate (Reuters, 2013). It is in this context that UNDP has been operating as the largest UN development agency since 1991 and has been joined over the years by the 13 101  other UN development agencies in residence as of 2012.20 Since 1991 the development objectives of UNDP included pillars identical to the priorities of UNDP in Bosnia and Kosovo: Democratic Governance, Poverty Reduction, Crisis Prevention and Recovery, and Energy and Environment. As will become clear below, the projects and programs falling under each pillar are also strikingly similar.  Albania is one of the UNCTs that have undergone the UN agency ―Delivery As One‖ pilot program21 and the UNIFEM/UN Women office is considered a strong and knowledgeable leader within the UNCT while the UN gender theme group is the most collaborative and coordinated of the three.  Bosnia and Herzegovina Bosnia‘s conflict lasted from 1992-1995, during which time more than 100000 people were killed, 30- 50000 women systematically raped, and up to 2.2 million individuals displaced (UNHCR, 2012). The war ended in a peace deal, the Dayton Accords, which established an elaborate power sharing government                                             20 Agencies present in Albania as of July 2012 are: the United Nations Development Programme (UNDP), the United Nations Children‘s Fund (UNICEF), the United Nations Industrial Development Organization (UNIDO), the United Nations Population Fund (UNFPA), UNAIDS, the United Nations Volunteers (UNV), the United Nations Fund for Gender Equality (UN Women), the UN High Commissioner for Refugees (UNHCR), International Fund for Agricultural Development (IFAD), the International Labour Organization (ILO), the World Health Organization (WHO) and non-resident agencies the UN Environment Programme (UNEP), the UN Educational, Scientific and Cultural Organization (UNESCO) and the Food and Agricultural Organization (FAO).    21 Delivering as One means the UN agencies jointly fundraise, have one UN Resident Coordinator leader, reside in one combined office space and jointly plan and align their programming. While eight countries volunteered to pilot the projects in 2007 the results so far have shown a small change in the programme planning process but not in the office structure, staffing or management; a full evaluation of the pilots is now pending and a decision on further rollout is expected soon. http://www.undg.org/?P=7  102  among the three parties to the conflict and an international high representative that would oversee political decision-making (Cousens and Cater, 2001).  In terms of humanitarian and development programming, UNHCR worked in Bosnia from 1992 onwards with a mandate of providing humanitarian aid and assistance to displaced persons. The UN peacekeeping mission, UN Protection Force (UNPROFOR)22,  arrived in February 1992 from neighboring conflict-affected Croatia to protect humanitarian convoys arriving and leaving the besieged city of Sarajevo. The mission‘s mandate evolved over time to include other humanitarian issues; protection of civilians; monitoring of zones of separation; and weapons checkpoints (DPKO, 1996. The mission was troubled by unclear rules of engagement and poor operational capacity, which left it paralyzed in the face of the atrocities occurring around it and led to a highly critical review of that decade‘s peacekeeping operations in the 2000 Secretary General‘s Report on Peace Operations (Jenkins, 2013). It is important to note that UN peacekeeping missions and UN development agencies operate quite separately on the ground, usually without significant coordination. The peacekeeping missions are not considered part of the UN country team.  UNDP opened its country office in 1996 after hostilities had come to an end. Over the years the number of UN development agencies in Bosnia grew to                                             22 UNPROFOR was deployed first to Croatia and Bosnia in February1992 with a mandate to establish conditions for peace talks and subsequently monitor a cease fire. For more information see DPKO‘s Former Yugoslavia UNPROFOR Profile webpage http://www.un.org/Depts/DPKO/Missions/unprof_p.htm  103  ten23. UNDP‘s pillars of work in Bosnia are identical to Albania (Democratic Governance, Poverty Reduction, Crisis Prevention and Recovery, and Energy and Environment), with one additional pillar - HIV/AIDS.  Bosnia and Herzegovina (BiH) until recently had no UNIFEM/UN Women presence. Its UNCT was thus the least coordinated of the three cases, with the least expertise on gender mainstreaming available to the UN country team. The office opened only in 2006 with a skeleton staff and has suffered consistently from problems of retaining staff.  Kosovo The conflict in Kosovo began in earnest in 1998 and 1999 as Serbian police, army and paramilitary began increasing their presence in the province and through violence (including sexual violence), arrests and assassinations orchestrated a campaign of intimidation of the ethnic Albanian population (Judah, 2002). In 1999 almost one million people were displaced in the space of a few months and fled to the borders with Albania and Macedonia. Spurred by shame over inaction in the face of a long string of ethnic cleansing campaigns in the Balkans conflicts, in March 1999 NATO voted to begin an aerial bombing campaign against Serbia, which lasted until June 1999 when the Serbian authorities capitulated and began to remove their security forces from Kosovo. Violence flared again in March 2004 when Albanians rioted and killed members of the Serb community (Judah, 2002).                                             23 Agencies present in Bosnia as of July 2012 included: ILO, IOM, UNHCR, UNICEF, UNFPA, UNESCO, UNDP, UNV, UN Women, WHO. 104  The UNHCR office in Kosovo was originally opened in 1992 to assist refugees fleeing from Croatia and Bosnia to Kosovo to seek protection and humanitarian aid. In the wake of the 1998-99 Kosovo conflict, UNHCR‘s mandate was broadened to include the provision of humanitarian supplies (UNHCR Kosovo, 2012).  The UN Mission in Kosovo (UNMIK) (a peacekeeping mission entirely separate from UNKT) was mandated by UN Security Council resolution 1244 (1999) and established a provisional administration in June 1999, while NATO leads the military peacekeeping force in the province, called Kosovo Force (KFOR) (UNMIK, 2013).   UNDP, as the main development agency, opened its office in Kosovo in 1999. UNDP is the largest development agency in Kosovo today, with the same five programming pillars as Bosnia: Democratic Governance, Poverty Reduction, Crisis Prevention and Recovery, Environment and Energy and HIV/AIDS.  As of July 2012 there were 14 UN agencies with offices and programming in Kosovo.24  Kosovo has the longest presence of a UNIFEM/UN Women office, since 1999, which has to some extent endeavored to coordinate gender mainstreaming in the UNCT, however is not widely recognized as a strong leader, partly due to a lack of international staff that could position the agency better within the UNKT hierarchy.                                             24 UNDP, UNICEF, UNFPA, FAO, ILO, IOM, OHCHR, UN Women, UN HABITAT, UNHCR, UNOPS, UNESCO, UNEP, WHO. 105  Conclusion The evidence collected in these three case studies provides rich insight into the causal mechanisms that affect the adoption and implementation of mainstreaming policy.   The next two chapters outline the evidence pointing to a combination of causal mechanisms at the adoption stage and at the implementation stage of the mainstreaming policy process, while Chapter Six examines the implications for theory and practice of these findings.   106  4. Policy making: World Polity leads to Policy Adoption when Bureaucratic Competition Lines Up    “Ten years ago when you tried to talk to senior management of government and UNDP about gender they ignored you. But somehow an agreement that gender equality is a goal just morphed into existence.”25 – UN Official  This chapter outlines the causal mechanisms that led to policy adoption in the three different cases. It supports the proposed theoretical framework that the policy adoption occurred at senior levels of management and displayed the characteristics and observable implications of the world polity approach; however it also depended to some extent on positive relationships as the bureaucratic politics approach hypothesizes. Senior management was mimicking the language of gender mainstreaming without displaying a clear understanding of its principles or concern about its follow through. If competition between senior managers over turf or personality got in the way, then the policy adoption stalled.  The chapter will discuss the main recurring themes, trends, categories and relationship coded from the interviews. The main findings include patterns and evidence of declarations of senior management support; lack of displayed accompanying background knowledge or commitment; the similarity in rhetoric and attitudes across the three case studies and agencies; the decoupling                                             25 UNIFEM Official, Interview by Ana Lukatela, 2010.  107  between rhetoric and financing; and the occasional competition in personalities and mandates that served as obstacles to policy adoption. Declarations of Senior Management Support Over the years since the discussion of women‘s status was taken up by international organizations, Berkovitch and Bradley (1999: 483) identify a world polity that has grown around the concepts of gender equality and gender mainstreaming and has become structured over time through a myriad of policies and programmes, which although separate, share isomorphic justifications and structures. It is this "world polity" that this study is arguing has gradually diffused into UNDP country offices through international development professionals circulating between organizations and exposing themselves and their colleagues to the isomorphic attitudes of the different international stakeholders they interact with on a regular basis. They argue that during the period 1945 to 1995 international organizations undertook activities and produced publications that led to a dramatic increase in attention to women's status in the framework of universalism, individualism and rationalism - the underpinning structures of the world polity according to (Meyer et al, 1997).   UN mid-level staff respondents generally described noticing the use of gender mainstreaming in the development discourse sometime during the mid 2000s. This is largely supported by the examination of the spread of UN agency mainstreaming policies which began to appear in larger numbers in the second half of that decade (see Table 2 in Chapter One). They described gender mainstreaming as entering the discourse without any particular discussion of its 108  meaning or value to the organization. A staff member of UNFPA in one of the three countries said that gender mainstreaming gradually appeared as a topic mentioned by senior management approximately five years ago. Prior to that time she could not recall it ever being mentioned in any documents, strategies or policies or acted upon in any programming.26  One anonymous UN official described how ten years ago when she would try to talk to senior management of the UN or the host government about gender equality they would dismiss it as irrelevant. However she then observed over time what she described as a slow ―creeping in‖ of the gender mainstreaming narrative among international development professionals where the term would begin to be used more and more often without ever being explained or justified. She felt it was based on signals of a changing discourse at the global level of UN development agencies as well as other international organizations with which these professionals were regularly interacting.27  Senior management‘s support of gender mainstreaming played an important role in adoption of policies and directives.  Respondents from each of the countries highlighted a gradual change in UN senior managers' statements and attitudes towards both gender equality as a goal and gender mainstreaming as a strategy. Importantly senior management of these UNCTs were not all from western states, refuting the argument that world polity reflects only western culture or western values or that only western nationalities mimic world values.                                             26 UNFPA Official, Interview by Ana Lukatela, 2010.  27 UN Official, Interview by Ana Lukatela,  2010. 109  Nationalities represented in senior management in these three cases since 2009 included for example Russian, Israeli, Dutch, Japanese, Turkish, Tunisian, Indian and Finnish. All of these individuals had held a minimum of 4 international postings in the past.28  More than 90% of the respondents said that senior staff within their organization did not exhibit negative views on gender equality and mainstreaming, and there was never a lack of positive statements. Upon probing, the respondents seemed to feel gender mainstreaming was part of the discourse but without being able to pinpoint a specific motivation for its appearance or an incentive for supporting the discourse. These are strong indications of the presence of isomorphic values on the issue. One respondent in Albania felt the UN Resident Coordinator seemed to regard her declarations of support for gender equality as ―a star in her hat‖29 and an indication that this was indeed a change from the past fifteen years of senior leadership regimes. Eight out of eleven (73%) public keynote speeches delivered by the RC during 2012 for example included a reference to the importance of gender equality or gender mainstreaming all UN projects (UNDP Albania, 2012). However as Figure 4 below shows, the annual Albania RC reports from 2005 to 2012 included only one mainstreamed project and only one instance of mainstreamed operations over the course of more than 7 years. This is not different from the situation in the other three countries. Figure 5 shows the same data for Bosnia and Figure 6 for Kosovo. The content of RC reports are particularly important as it is the                                             28 Data collected through a number of interviews and notes from participant observation.  29 UN Official, Interview by Ana Lukatela, 2010. 110  mechanism through which the UN Development Group holds the Resident Coordinators accountable for results for overall UNCT performance.   Figure 4: References to Standalone Women's Empowerment Projects, Mainstreamed Programmes and Mainstreamed Operations in UN Albania Resident Coordinator Reports from 2005-2012.       Figure 5: References to Standalone Women's Empowerment Projects, Mainstreamed Programmes and Mainstreamed Operations in UN Bosnia and Herzegovina Resident Coordinator Reports from 2005-2012.    111  Figure 6: References to Standalone Women's Empowerment Projects, Mainstreamed Programmes and Mainstreamed Operations in UN Kosovo Resident Coordinator Reports from 2005-2012.  UNDP Albania Country Director, Vibeke Risa, worked over the past 10 years in the UN system in five different UN country offices and before her UN career was in the world of international development NGOs. She noted that when she started her career ―gender‖ was a concept met with complete disinterest and inattention. However over the course of many years she feels that this has changed; now she sees other senior managers and senior colleagues openly and without reservation giving full declarative support for gender mainstreaming, without requiring any explanations or justifications of its worth or what it actually entails. Upon being pressed she was unable to identify a specific reason for why these attitudes had changed. She did, however, note that despite always being prepared to give strong declarative support many of her senior colleagues still displayed a mild disinterest if they were asked to act on their declarative support and for this reason she felt that the support ultimately hinged only on ―words and not deeds.‖30 This is a good example of the declarative support that world polity theory expects of gender mainstreaming adoption.                                              30 Vibeke Risa, Interview by Ana Lukatela, Tirana, March 4, 2010.  112  Upon reflection Risa stated: ―Our project here doesn‘t really mainstream. Our project here is a very gender equality specific project targeting women.‖ Risa is recognizing the fact that while UNDP Albania is in full declarative support of gender mainstreaming, they are not actually doing it in any of their thematic areas or their own structure or organization. An examination of UNDP Albania‘s country reports vs. its UN Development Assistance Framework (UNDAF) commitments confirms this observation. While the UNDAF sets out gender mainstreaming as a priority objective, the Resident Coordinator‘s annual report is able to provide only two cases of this being implemented in a seven year period, from 2005 to 2012, as demonstrated in Figure 4 above. So UNDP Albania may have adopted a policy of gender mainstreaming but in reality are still implementing only women in development projects. Despite a declarative commitment to a policy of mainstreaming in the UNDAF, the strategy has been implicitly abandoned in practice although not in theory. So clearly declarations of support by senior management have not been enough for UNDP Albania to move from policy adoption to implementation. Even though the Resident Coordinator present on the ground mentioned gender equality in her statements and public appearances31 there was no follow up within the organization. In addition the reports of the RC focused very much on women in development projects while still using the term gender in reference to these projects; which increases the suspicion that women in development and standalone gender equality projects were perceived by the RC and UNCT as                                             31 Based on notes from participant observation while working in Albania from 2006-2010. 113  being "gender mainstreaming" when in fact they are not. Figures 7, 8 and 9 below demonstrate the results of a content analysis of the RC reports for the frequency of the use of the word "gender" and "women." Even though no gender mainstreaming projects were being reported, the phrases gender, gender equality, gender perspective, gender mainstreaming, etc. were being used in relation to standalone women's empowerment and women in development projects.   Figure 7: Frequency of  occurrence of words "gender" and "women" in UN Albania Resident Coordinator Reports from 2005 - 2012.   The RC's reports from Bosnia and Herzegovina and from Kosovo show that the word gender is not being used more over time in comparison to the word women. In fact the trend seems to be leaning more in the other direction, with the word women being used more often. This is also likely related to the predominant reporting in these reports on standalone women's empowerment projects, as opposed to the gender mainstreaming of other programmes.    114  Figure 8: Frequency of  occurrence of words "gender" and "women" in UN Bosnia and Herzegovina Resident Coordinator Reports from 2005 - 2012.   Figure 9: Frequency of  occurrence of words "gender" and "women" in UN Kosovo Resident Coordinator Reports from 2005 - 2012.  The gender focal points of different UN agencies from each of the case studies agreed that gender mainstreaming policies are no longer a point of contention in international development organizations and in all cases have been adopted as outcome-level policy goals in UNDAFs or Country Strategies, which indicate at least a tacit acceptance of gender equality. However they also agreed 115  that declarative support was not enough for implementation. Given the widespread agreement in both the practitioner and academic literature that gender mainstreaming has not been successful32 this tacit consent has clearly not been enough to lead policy adoption towards policy implementation. One example of a gradual progression towards mimicking world polity values by a senior level policy maker was  given by an anonymous UN official describing a colleague‘s move from middle to senior management in the international development community. This official described how initially the national staff mid-level programme manager joined the UN and openly questioned the very premise of gender equality as ―against the natural order created by God.‖33 After several years within the UN this individual was offered a senior position with another international development organization based on his programme management skills. At this new position he was exposed to continuous contact with international actors such as EU officials in Brussels, other international development professionals and sent for training on human rights and gender (among other topics) to international locations. When his former national level UN colleagues observed him now mimicking language on gender mainstreaming they were stunned. In their minds there was no explanation or incentive for him to be repeating the discourse since he had gotten the job with no requirement to be gender sensitive and his employment was secure. The non-rational nature of his mimicry was a fascination to them all                                             32 See the conclusion of the 2006 UNDP Gender Mainstreaming Evaluation or the 2010 Secretary-General‘s report on Gender Mainstreaming  in The United Nations. 33 UN Official, Interview by Ana Lukatela, 2010. 116  and the only difference between his position with the UN and his new position was the exposure he was getting to the world of international development professionals. They felt it unlikely that he had undergone such a fundamental shift in his values and norm socialization just on the basis of his new job, especially since his rhetoric was not backed up with any substantive actions.  Proponents of other approaches, such as the bureaucratic politics theory or the organizational behavior theory, might argue that discourses change because actors fear a credible threat of punishment if they do not conform or else expect personal gain from conforming. However as will be discussed below and in the next chapter, it was clear to everyone in the UNCTs that no sanctions were ever imposed by senior management on mid-level staff that did not implement gender mainstreaming in their programmes. It was not even discussed.  In addition, a bureaucratic politics approach might expect that individuals would mimic mainstreaming discourse only when meeting with international development professionals in order to conform and as a strategy for personal gain; however he mimicked the discourse even when speaking with old national staff UN colleagues that had not had exposure to the world polity.  All the UNDP Kosovo staff interviewed for the study mentioned the change within the office in the gender mainstreaming discourse over the past six or seven years. While they partly ascribed this to a change in management they witnessed also a gradual change in the senior management that stayed on. Prior to 2008 few efforts were made to contextualize the corporate mainstreaming strategy to the country situation or examine the extent to which gender was 117  addressed in UNDP‘s local level programming or operations. The Resident Representative during this time  had only held one international posting and prior to the UN has been in international banking, not international development. He had thus had much less exposure to the world polity on international development issues. One UNDP project manager felt that during this period "When I approached anyone [about gender] (anyone!), they did not want to listen to me, they did not hear me. Now it is a different story.‖34 The new senior management of Kosovo‘s UNDP arrived in mid-2008 and consisted of two new senior managers each of whom had a minimum of five previous international development postings. The arrival of new management triggered a change in UNDP Kosovo‘s atmosphere and resulted in clear declarations by senior management on gender equality as a development goal and the preparation of an internal gender mainstreaming policy. All UNDP Kosovo staff interviewed confirmed that the creation of a country level gender mainstreaming policy had never been discussed prior to the arrival of new management. One explained: ―Gender was never a priority for UNDP since its beginning here in 2000. It was non-existent.‖35  This project manager went on to describe how the new management conducted an evaluation and began drafting a policy on how to mainstream gender in projects and operations.                                              34 Virgjina Dumnica, Interview by Ana Lukatela, Pristina, May 4, 2010. 35 UN Official, Interview by Ana Lukatela, Pristina, May 2010. 118  Osnat Lubrani, an ex-UNIFEM director and UNDP Kosovo Resident Representative, describes her own views as less about gender mainstreaming and more about women‘s empowerment. In an interview she explained that in her experience gender mainstreaming cannot be implemented if staff have not internalized women‘s empowerment, which most have not. She felt that it is one thing to mimic the values you hear around you (the world polity) but entirely another to act on values that may not be deeply internalized.36 Ultimately she felt that the lack of implementation by senior managers that she had worked with in the UN system over the past 7 years was an indication of how the commitment was not accompanied by any results or sanctions. The lack of sanctions or incentives attached to gender in the UN system means that middle management has no strategic reason to mimic values they don‘t necessarily understand or agree with. Since gender mainstreaming rarely appears in evaluations or implementation reports there is no concern that not mimicking will have negative consequences for a staff member. Figures 10 and 11 below chart the number of standalone women's empowerment projects reported against the number of mainstreamed programmes and mainstreamed operations in annual global implementation reports by UNDP and UNICEF.  These annual reports provide an overview of the organizations' programming and operations.  In a content analysis of the last 12 years of reporting, since a standard and longitudinally comparable reporting format was adopted, it is clear that there has not been a marked increase in the number of                                             36 Osnat Lubrani, Interview by Ana Lukatela, Pristina, May 10, 2010 119  mainstreamed projects being reported in comparison to standalone women's empowerment programmes, and certainly there has not been in an increase in reporting on progress on mainstreaming in the organization's operations. The reporting on operations that is included in this graph remained limited over the years to reporting on efforts to improve recruitment policies to promote gender balance among staff.   Figure 10: References to Standalone Women's Empowerment Projects, Mainstreamed Programmes and Mainstreamed Operations in UNDP Annual Reports from 2001-2013.    120  Figure 11: References to Standalone Women's Empowerment Projects, Mainstreamed Programmes and Mainstreamed Operations in UNICEF Annual Reports from 1999-2012.  A comparison of the frequency of the use of the word "gender" vs "women" in the annual reports (Figures 13 and 14) indicates that the use of the latter still dominates, however with a slow increase over time in the use of the term gender despite no concurrent increase in actual gender mainstreaming being implemented and reported. The slow spread of the term in global reporting may in fact be another effect of the world polity  on senior managers.  121  Figure 12: Frequency of  occurrence of words "gender" and "women" in UNDP Annual  Reports from 2001 - 2013.    Figure 13: Frequency of  occurrence of words "gender" and "women" in UNICEF Annual  Reports from 1999 - 2012.  In addition to the global implementation reports there are also UNDP's annual global evaluation reports, which evaluate the work of the organization on a yearly basis. A content analysis of these reports, from 1993 to the present, finds no noticeable increase in concrete recommendations for increased gender mainstreaming over the last 20 years although there are a large number of occurrences of both the words gender and women (see Figure 14 below). In the years when the ECOSOC resolution on mainstreaming was passed (2007) and 122  when UNDP published its large evaluation of gender mainstreaming (2005) we see an unusually high occurrence of references to mainstreaming (emphasized in darker blue in Figure 14). In the year after the publishing of the gender mainstreaming evaluation we see zero recommendations related to mainstreaming and the use of the word gender drastically falls. The use of the word women varies over the years, however appears to remain in a stable range. The term gender in development (not reflected in this graph) was used multiple times in the 1994 and 1995 reports, prior to the Beijing Platform for Action and the introduction of the concept of mainstreaming.    Figure 14: Frequency of references to "Gender", "Women" and or Mainstreaming in UNDP Annual Evaluation Reports from 1993-2012.   If the spread of gender mainstreaming was a stage in the norm socialization process then we would expect to see an increasing number of actors do more than just mimic the discourse; we would expect to see them take 123  active measures to implement as they become socialized in the norm and internalize it. The fact that these active measures don‘t occur was identified by Lubrani as significant evidence of the lack of internalization of norms or values on gender. When we consider this in relation to the observable implications outlined in the theoretical framework, it is the lack of progress along the norm socialization spiral that is the strongest indication of the existence of the world polity‘s non-rational mimicry as opposed to the norm life cycle.  A number of UN agencies have commissioned independent evaluations of their gender mainstreaming results (or rather lack thereof) over the past decade. These evaluations came to similar conclusions across the cases – there were no results to show for mainstreaming efforts. Figure 1 in Chapter One lists these evaluations and disaggregates their results by the main factors identified as responsible for the gap. A wide range of factors are identified but common ones are the lack of senior management commitment and lack of accountability. The ―gender marker‖37 approach was one of the methods used to evaluate programmes and will be discussed further below. In UNDP in Albania one respondent described how immediately after the poor results of the evaluation were announced the senior management made strong declarations that they intended to remedy the gap. However, other than the declarative statements made in team meetings no other steps were taken to follow up on the                                             37 The gender marker is employed in several UN entities, including the Office for the Coordination of Humanitarian Affairs (OCHA), UNDP, UNICEF, ILO and UNV, to track gender-related resource allocations and expenditures. It assigns a score of 0-3 to programmes in terms of the extent to which gender is mainstreamed into the program. 124  recommendations of the evaluation and middle management was given no motivation, incentive or sanctions. UNICEF in Kosovo underwent a similar internal evaluation directed by their Headquarters in 2008, the results of which were obtained by the author. The report found that there were missed opportunities to implement the corporate mainstreaming guidelines and that while senior management was supportive, they needed perhaps to explore other mechanisms such as sanctions on programme implementers in order to move progress forward.  The evaluation found that senior management was not resistant to gender mainstreaming and that they were supporting ―the narrative of gender mainstreaming‖, however this had not led to movement from paper to practice. The internal evaluation suggested that management formalize its statements on mainstreaming and send clear signals on expectations by putting gender knowledge into all staff‗s core competencies in their official Terms of Reference. When asked whether management was changing its approach to gender mainstreaming and becoming actively seized of the issue the gender focal point continued to be skeptical. She felt that programme staff was continuing to provide excuses to senior management about why they had been unable to mainstream their programmes and that with no sanctions imposed by senior management this meant there was no motivation to change behaviour.38                                             38 Arbena Kuriu, Interview by Ana Lukatela, Pristina, May 13, 2010.  125  Senior leadership of other international organizations and structures such as the Organization for Security and Cooperation in Europe (OSCE), European Union Rule of Law Mission in Kosovo (EULEX) and the European Union Military Operation in Bosnia and Herzegovina (EUFOR) were also recognized by respondents in this study as critical actors in adopting or not adopting a gender mainstreaming policy. For example, the outgoing Head of the Human Rights and Gender Unit of EULEX39 described the institution‘s approach to gender issues as ―very complex and very secretive.‖40 Senior leadership was composed of military commanders from EU countries who had no interest in discussing gender mainstreaming and were very derisive of her mandate and priorities. The military commanders that made up the top ranks of EULEX were not international development professionals like her and had not experienced multiple international development postings where they would have been exposed to the values of the world polity. Her attempts to begin a process of adopting gender mainstreaming policy were stopped by senior management.  Not only was there a brake put on gender mainstreaming policy, but Rautio also described that she was being approached with allegations of sexual discrimination, harassment and gender based violence in the mission that were going unaddressed by senior military commanders. She felt there was enough evidence to warrant an investigation into these allegations and that it was                                             39 The Human Rights and Gender Unit has an advisory and coordination mandate to ensure that all EULEX Kosovo activities respect principles of human rights and gender equality. The Office had two positions for international gender experts and one position for a national gender expert, along with a number of staff dealing with human rights . 40 Sirpa Rautio, Interview by Ana Lukatela, Pristina, May 13, 2010. 126  useless to talk about gender mainstreaming if fundamental women‘s human rights were still going unaddressed without transparent recourse to action within the organization. It is certainly interesting that she did not perceive gender mainstreaming as able to make a contribution towards human rights, which is different from what its theoretical roots would argue. The difference in the attitudes and behaviour of national level leadership compared to international development professionals also lends evidence to the existence of a world polity. Both UN officials and civil society partners described how those national leaders that had intensive contact with the international development community began to mimic the same rhetoric and declarations of support for gender mainstreaming, without questioning the concept, its value or its meaning. On the other hand, they also described how those that had previously had little contact with the international community and their discourse, such as local government councilors or mayors, initially did not mention or discuss gender equality in any way or show any awareness of the concept of gender mainstreaming. However they gradually began to imitate the discourse used by the international development actors and mimic the attitudes they were hearing.  Bureaucratic politics theory would argue that these officials expect financial incentives from mimicking gender mainstreaming concepts, however the support and partnerships being offered to them by the UN agencies and other actors were overwhelmingly technical in nature, as described below, and not about the provision of development funds, meaning a financial incentive was not 127  in fact present. The constructivist approach would argue that declarative support, or "talking the talk" (Risse, Rop and Sikkink, 1999: 16) is only one step in the norm socialization model leading towards internalization and action. However there was no doubt in the minds of senior managers interviewed that they believed gender equality to be important and the author observed countless instances during the years of participant observation where senior managers would request policy advisors to insert "something" about gender equality in their statement or speech "because it is so important for what we are trying to achieve." The belief was genuine, rather the obstacle seemed to be that they understood their role only in terms of form and process. Perhaps the world polity, composed after all of individuals belonging to an elite,  lacks the imagination or connection with the represented community (marginalized women) to understand their role as one of action. The focus on form and process over action displayed by senior UN managers resonates with Sending's (2011) account of thin and thick culture and supports the theory that members of the world polity internalize norms in a very different way than constructivists typically anticipate. However the constructivist expectations of socialization and internalization at the middle management level hold, as will be discussed below in Chapter Five. This has interesting implications for constructivist theory that will be discussed further in the conclusion. For example, the Austrian Development Agency (ADA) had for some years been engaging local level municipalities in Albania in modernizing their operations, including mainstreaming gender into their planning and budget 128  processes. Initially ADA‘s project manager described how mayors were uninterested in the gender equality aspect of the programming. However, after a certain amount of engagement and visits from international development professionals they began to mimic the same phrases and discourse, even among themselves after the project was over, as noted the project officer. She commented that they still did not understand the concepts even though they had accepted the discourse as the valid one and were happily ―checking off the gender boxes‖ provided by ADA‘s technical experts regarding municipal  planning and budgeting processes. She pointedly noted that this never went beyond checking boxes and making declarative statements of support. So although the mayors gradually began to imitate the language that ADA‘s international experts were using they did not necessarily act on implementing the recommendations.41 This is non-rational mimicry in so far as there was no incentive for the municipal officials involved in the provision of this technical expertise – they were not directly receiving funds from ADA.   A UNIFEM staff member also working at the municipality level had the same impressions of ADA‘s provision of gender mainstreaming expertise. She described similar experiences with municipal actors and argued that as exposure to international development actors increased the local level leaders began to mimic the international discourse. However even while using the phrases about gender mainstreaming the mayors and councilors remained passive; neither                                             41 Dolly Wittberger, Interview by Ana Lukatela, Tirana, March 6, 2010. 129  resisting nor actively supporting implementation efforts. 42 This fits with world polity approach; as exposure to international development professionals increases we see actors being to mimic the same attitudes and values, however those values never go through the norm cycle in terms of socialization or internalization by any of the actors, which goes against constructivist expectations. These actors are characterized by their unfailingly positive yet completely passive approach to gender mainstreaming.  This passivity in implementing gender mainstreaming is corroborated by the author‘s experiences in over five years of implementing programmes in this region. Leaders at the national level in Albania, who were identified by almost all respondents interviewed from UN, government or civil society, as "on board" and supportive of gender mainstreaming policy, were the Prime Minister, the Minister of European Integration and the Minister of Labour. Unlike the local leaders these are individuals who had been in politics and working with international development actors for many years and their attitudes and rhetoric had aligned with the values of the world polity over the years – unlike with the local level leaders where a period of exposure was clearly necessary before an isomorphism of values crystallized. The difference between national level leaders and local leaders can be explained through the extent of exposure to international development professionals and the world polity.  This same situation was also described in Kosovo by a respondent from within the government. She felt that working with national level gender focal                                             42 Emira Lubani, Interview by Ana Lukatela, Tirana, March 8, 2010.  130  points was completely different from working with the municipal level gender focal points who had not had any substantive or direct contact with international development professionals and agencies. She felt the latter were a ―curiosity‖ to her in terms of not automatically agreeing  that gender mainstreaming has inherent value, unlike colleagues at the national level who were working with international development agencies on a regular basis.43  Similarly, within the UN country teams, while senior UN management (always international ex-patriot staff) made declarations mimicking the isomorphic values of the world polity, this was not necessarily the case among mid-level programme managers. These mid-level staff were more likely to be national (as opposed to internationals who had already travelled the world in different postings with the UN), or if they were international staff they were more likely to be lower in rank and thus newer to the UN system and only just gaining exposure to the world polity through their networks and international postings.  In general, the three countries all displayed examples of senior management contributing to the adoption of policies on gender mainstreaming through their statements of support to programme staff. However, it is important to note here that support in its declarative form exclusively focused on mainstreaming programming, and never on operations. This is the difference between liberal gender mainstreaming (only programmes) and transformative gender mainstreaming (operations, structures and social relations) that was identified in the theoretical framework‘s review of gender mainstreaming (Jahan,                                             43 Schqipe Krasniqi, Interview by Ana Lukatela, Pristina, May 12, 2010.  131  1995). The content analysis of policy and implementation reports in the Figures above also confirmed that mainstreaming operations was rarely reported on in the three cases studies or at the global level. It is important to note that the type of gender mainstreaming strategy supported in senior management declarations entails a liberal approach to mainstreaming, which aligns with the world polity values on gender mainstreaming, as described in the theoretical framework. Liberalism is identified as a key characteristic of the world polity – it is grounded in ―universalism, individualism and rationalism‖ (Meyer et al, 1997) – and the types of women‘s rights supported by the world culture fall strongly within a liberal framework  - the right to vote and the right to work, for example (Berkovitch and Bradley, 1999). Only on one occasion was support for transformative gender mainstreaming expressed by any of the senior UN officials, and in that case it was by a feminist senior gender advisor placed on a pilot project basis in the UNDP Country Team in Kosovo.44 Her perspectives and impact on implementation of gender mainstreaming will be discussed in Chapter Five.  Ultimately, changing the fundamental structures and operations of international organizations is a radical transformative approach, certainly not a liberal one, and is another indication of why gender mainstreaming policy can be adopted, but not necessarily have any substantive effects in its implementation. If implementation focuses on increasing the number of women participants in a programme, but not on addressing underlying gender inequalities of the subject                                             44 Jocelyne Talbot, Interview by Ana Lukatela, Pristina, May 5, 2012.  132  matter a programme is dealing with, then long term results of improved gender equality are unrealistic.  Lack of Accompanying Background Knowledge  The world polity‘s influence occurs through a diffusion mechanism of non-rational mimicry. This mechanism is described as the process through which individuals in the world polity provide ―depictions of their roles and identities, and recipes for activity routines that make sense in terms of other actors‘ expectations‖ (Bonacker et al, 2011:117). One of the observable implications is that subjects do not have a lot of background knowledge of the narrative that has influenced them and which they are mimicking. In the cases in this study, a lack of background knowledge was evident across the different types of actors. They had relatively little understanding of what gender mainstreaming meant, how it was implemented or what the root causes and manifestations of gender inequality are in their context. This was clear from interviews with senior UN management, as well as programme staff (and their descriptions of other colleagues), and national government officials and civil society members. The exceptions to this observation came from those individuals who, for reasons discussed below, identified as personally committed to issues of women‘s rights.  UNIFEM staff had some of the best insights into the problematic lack of background knowledge on gender issues in the UN system. One UNIFEM staff member described how she had previously been interviewed by UNDP senior management for a gender mainstreaming position in a large community development program in a post-conflict country. When asked in the interview how 133  she would design the programme she explained that the existing programme should be reviewed to make sure women were participants in two specific outcomes in the logistical framework and that technical expertise by a gender advisor should be provided for all the other outcomes. The UNDP country director on the interview panel was surprised at the strategy and explained their expectation is a separate program and budget with some separate activities for women. When she explained to them that this approach is not in fact then mainstreaming gender into a multi-million dollar programme, but rather a ―Women in Development‖ project, the panel remained unconvinced.45 She took this as a not uncommon example of how the words gender mainstreaming are used by development officials without any corresponding knowledge of what it actually entails.  Another senior UNIFEM staff member described the problem as going all the way up the UN chain. She gave the example of the hiring process for the position of Special Adviser to the Secretary General on Gender Issues and Advancement of Women. The successful candidate was included in the pool of candidates with expertise in gender mainstreaming or women‘s rights because her background in Human Resources included a good track record of recruiting women. While appointments at high levels in the UN are certainly also prone to political pressure and geographic quotas the fact that she was even included in the pool of candidates was an indication to this UNIFEM staff member of the UN system‘s misunderstanding of what gender mainstreaming entails and an                                             45 UNIFEM Official, Interview by Ana Lukatela, 2010.  134  ignorance of how to deal with the underlying structural problems that cause inequality.  Many of the gender focal points responded that their senior management seemed to suffer from a persistent lack of understanding of how gender mainstreaming ―is done.‖ The World Health Organization‘s gender focal point in Albania found that when discussions of gender mainstreaming were raised with senior management they became slightly evasive. She felt they did not always understand what was being asked of them and were perhaps too embarrassed to ask questions. Although they had verbally committed to it, the policy was never taken into action because no one wanted to admit that they did not know what exactly that would entail. She described the problem created by a lack of background knowledge on gender mainstreaming: ―People become nervous because it requires being present in meetings and contributing‖ so they rather evade meetings than have to participate and make their lack of knowledge obvious to everyone.46  Another common misunderstanding stemming from this lack of background knowledge was that it requires a significant amount of time and resources to undertake gender mainstreaming. In the ADA programme mentioned earlier that worked with local government leaders in Albania, the local mayors were clear that they supported gender mainstreaming, however they felt that they would not be able to fulfill expectations because they did not have the                                             46 Ledia Lazari, Interview by Ana Lukatela, Tirana, March 10, 2010.  135  money ―to pay‖ for mainstreaming.47 The ADA programme officer commented that she had difficulty in making the local government leaders understand that gender mainstreaming did not necessarily require finances, that it was more about changing their perspective and approach; changing their planning process from top-down to bottom-up and ensuring that local women were present in planning and budget hearings.  She explained to the local leaders that in their particular case no extra finances were required and they should just ensure that they consulted and met with local women‘s NGOs to hear their priorities and needs and then integrate them into local planning and budgeting processes.48  The support for gender mainstreaming by these local level leaders despite their complete lack of knowledge of what it means is an interesting example of the non- rational behavior and mimicry of the world polity discourse. It is hard to imagine leaders taking up other policy goals in such a whole-hearted and robust manner without any understanding of its implications and with no financial incentive from a donor considering that the programme consisted only of technical expertise.  UNICEF Kosovo‘s gender focal point Arbena Kuriu described the same problem. Speaking of the 2008 evaluation that was done by headquarters of their programming she commented: Everyone in the evaluation said that if I had the time I would do it [gender mainstreaming]. But no one realizes that it is not something you do in                                             47 Dolly Wittberger, Interview by Ana Lukatela, Tirana, March 6, 2010. 48 Dolly Wittberger, Interview by Ana Lukatela, Tirana, March 6, 2010. 136  addition – it is something you do while writing the programme and the Terms of Reference for consultants and experts. Management accepts the excuse that GM is somehow extra work and not a perspective, or an understanding.49 This statement points to the lack of background knowledge among senior management on what gender mainstreaming means and the resulting failure to communicate this knowledge or approach to staff.  There was no understanding that gender mainstreaming a programme meant changing perspective and including women‘s views and women beneficiaries in the process; or looking for a consultant who was a thematic expert but ALSO gender sensitive, for example. Instead the question of gender mainstreaming remained a mysterious process that seemed to require extra and undefined ―resources and time‖. Risa described UNDP Albania‘s evaluation as resulting in very similar conclusions: senior management‘s low awareness of what exactly mainstreaming meant even though everyone clearly felt committed to engaging on the issue.50 In Bosnia similar concerns were expressed about the lack of background knowledge on gender and how local expertise could be easily integrated into their programming in order to gender mainstream it.  One informant anonymously complained that the UN agencies‘ leadership (past and present) had adopted gender mainstreaming without any idea of what it meant in practice and thus sidelining local women‘s perspectives instead of including them:  No one who came in of [the] internationals knew that we had great practices in socialism on                                             49 Arbena Kuriu, Interview by Ana Lukatela, Pristina, May 13, 2010. 50 Vibeke Risa, Interview by Ana Lukatela, Tirana, March 4, 2010. 137  gender equality and that women were organized. They sidelined us and worked only with male elites that had fought and all our knowledge was thrown away.51  By not including local women the UN system and other international organizations were losing valuable knowledge on how to gender mainstream their programming. Again, as the theoretical argument put forward in this dissertation expects senior management was supportive of gender mainstreaming to the extent it required no other effort from them than declarative commitment and they were certainly not interested in thinking through how to formulate it more effectively. A UNFPA official in Bosnia, Faris Hadrovic, echoed these thoughts when he mused that women had actually enjoyed better social protection, child-care and health benefits under socialism, so why was the UN trying to teach the government how to gender mainstream new legislation instead of finding better ways to support women‘s civil society to advocate for the return of old laws and benefits. He felt the UN was going about gender mainstreaming by ―reinventing the wheel‖, since not enough individuals had made an effort to collect the background knowledge necessary to understand how to gender mainstream effectively in the context.52 In conclusion, while the verbal commitment to gender mainstreaming had been expressed by senior leaders it tended to not be accompanied by an understanding of what the policy actually entailed. As world polity theory expects,                                             51 Anonymous, Interview with Ana Lukatela, Sarajevo, June 2010.  52 Faris Hadrovic, Interview with Ana Lukatela, Sarajevo, June 16, 2012. 138  individuals were mimicking what they were hearing, without understanding it, and this did not seem to represent a concern for them. In addition there are no accountability mechanisms or incentives for senior management on gender mainstreaming, so there is no strategic personal gain from ―talking the talk‖ but not ―walking the walk‖ and no clear costs associated with undertaking action. Resident Coordinators for example are assessed through several criteria laid out in the 2009 UNDG Management and Accountability Framework of the Resident Coordinator System. In this 15 page document gender is not mentioned even once. Rather the criteria against which RCs are assessed is related to implementing the UNDAF and other joint programmes and multi-donor trust funds; maintaining good political relations with the government; maintaining good cooperation and morale within the UNCT as a whole; successful resource mobilization; and improving the strategic position of the UN (UNDG; 2009). A positive record on implementing gender mainstreaming is not a factor in performance assessments or promotion of senior management. Vibeke Risa, the Deputy Country Director of UNDP Albania was aware of the existence of UNDP‘s Global Gender Mainstreaming Strategy for 2008-2011, however ventured that not much about its contents is known by the management of the country team. She pondered that perhaps the strategy might be ―a tool that we need to start using really; however given the magnitude of tasks we have, it is hard to do it in a systematic way.‖53                                              53 Vibeke Risa, Interview by Ana Lukatela, Tirana, March 4, 2010. 139  This concern points to a continuing sense of uncertainty about where to begin with implementing gender mainstreaming despite the robust support for it as a policy. Similarity in Rhetoric and Attitudes  As already discussed in the theoretical framework Boli and Thomas describe the non-rational mimicry of similar attitudes and values as a key characteristic of world polity theory. They identify the mechanic repetition of principles related to universalism, individualism, and progress and world citizenship by individuals translating the global cultural reality into specific forms and actions and in the process revealing broad homologies (1997: 180). Respondents across the cases displayed this same pattern; statements, rhetoric and behavior and attitudes ascribed to colleagues all were similar. Importantly, it was also non-rational; actors did not question the origin, validity or logic behind their views, they were implicitly accepted.  Gender mainstreaming policies and strategies should be very similar across international organizations and while there may be a vanguard at the very early stages, as with the example of the ICRC and human rights, as it spreads no specific person or administration should be attributable as the motivation. In the content analysis of policy documents the isomorphism is visible. Table 2 in Chapter One described the trend over time of a convergence in structure and content of UN agencies mainstreaming policies  140  In relation to the ―progress‖, ―world citizenship‖ and ―universalism‖ aspects of world polity, a repeating line of rhetoric was the wish to be ―modern‖ and the assumption that supporting gender equality rhetoric was an inherent part of this process. Eglantina Gjermeni, a former UN consultant and at the time of interview a Member of Parliament in Albania, identified this as a repeating discourse among senior government and civil society actors engaged with the international development efforts in the country.54  UN staff in Albania also described their interaction with senior government and how the concept of ―European‖ and ―modern‖ were consistently a reference point as though they had some inherent meaning that never needed to be explained. The passage of the gender equality law, which included a section on gender mainstreaming legislation, was perceived as an expression of these aspirations. However as Ingrid Baken, a UN staff member, pointed out: ―all these countries are passing Gender Equality Laws, but it is not because they internally support the goals but rather they see it as ‗European‘ and therefore desirable.‖55 This legislation remains largely on the level of words on paper, as again there is little implementation beyond the support in words. The funding attached to gender equality initiatives, as explained earlier in the dissertation, is so small in comparison to other development priorities and funds that a financial incentive is not a credible argument.                                              54 Eglantina Gjermeni, Interview by Ana Lukatela, Tirana, March 11, 2010. 55 Ingrid Baken, Interview by Ana Lukatela, Tirana, March 8, 2011. 141  Fiorela Shalsi, another UN staff member in Albania, noted that while working on the draft gender equality law with political parties and members of parliament she experienced a backlash from some senior government officials, however it was interesting to her that even this backlash was framed within the universalism discourse and the ―language of rights.‖  Opponents were careful not to argue that they were against equal rights for women; rather they framed their argument by saying that women in Albania had already achieved full rights and equality, so the law was unnecessary and moot.56  In Kosovo UNIFEM staff member Kaoru Yamagiwa described the same situation concerning ―the rhetoric of meeting European standards‖ and its power. She heard senior government officials talk extensively about meeting ―European standards‖, but never about substantively what those standards were and what happens once ―standards are met‖ and how things in the country would actually be substantively different.57 This was corroborated by the statement of a Kosovo government official who, when asked what gender mainstreaming meant to her, replied: ―Gender mainstreaming is just about being European I think - there was interest within our organization and we felt there were standards we had to fulfill.‖58   Another UN official in Kosovo who spoke anonymously felt that while both UN and senior government repeated the same rhetoric on gender equality there was little follow through on implementation:                                             56 Fiorela Shalsi, Interview by Ana Lukatela, Tirana, March 4, 2010.  57 Kaoru Yamagiwa, Interview by Ana Lukatela, Pristina, May 14, 2010. 58 Hadije Binaku, Interview by Ana Lukatela, Pristina, May 11, 2010. 142  They say it because they perceive it as one of the boxes they have to check if they want to be modern Europeans. They understand that they have to tick boxes but without real commitment. Making declarations on gender equality is about making the right noises and ticking enough boxes... it is collateral damage of the European vision.59 Arbena Kuriu of UNICEF in Kosovo described her own ―checklist‖ experience.  After the HQ evaluation and new rounds of gender training eventually all the programming staff  were able to ―recite by memory the gender checklist, but none of them would be able to explain what any of it means in their own words if you asked them.‖60 So while they did begin to repeat the rhetoric and attitudes that they heard around them, they were not clear on what kind of commitment it entailed. Respondents were asked to describe what gender mainstreaming means and almost all answered that not many people actually did know how to ―do gender mainstreaming.‖ In general a description of gender mainstreaming entailed some combination of the following phrases: it has to do with ―taking into account gender and women‘s issues,‖ or, it is about being ―gender sensitive.‖ ―Gender sensitive‖, ―gender responsive‖ and ―gender issues‖ where all terms used over and over by respondents without any clear indication that they had a fixed idea in their minds of what this meant exactly. When pressed to explain what these phrases meant, or why they supported gender mainstreaming if they were unsure how it should be implemented, respondents referred up the                                             59 UN Official, Interview by Ana Lukatela, Pristina, May 2010.  60 Arbena Kuriu, Interview by Ana Lukatela, Pristina, May 13, 2010. 143  hierarchy or explained it was ―the right thing to do‖. Emira Lubani of UNIFEM felt that for many UN officials and senior government officials it was not unusual that they could give no specifics; they heard everyone talking about gender mainstreaming, however ―instructions themselves were never specific; no one did anything specific or followed up.‖61  One NGO staff member described how the concept of gender mainstreaming appeared in Albania in 2000 through international development agencies from the Netherlands and Sweden. She recalls that first the local NGOs working with these agencies had no background or knowledge on women‘s rights or gender equality however after working with the international agencies for some time they started to use the ―right words‖ on gender. The expectation of rational mimicry would be that other NGOs would quickly catch on and start mimicking the language in order to partner with donors. However she did not describe this happening – it was only those NGOs that were exposed to the international agencies that gradually started to use the rhetoric on gender mainstreaming.  The continual reference to ―results‖ and ―indicators‖ of gender mainstreaming, without a clear understanding of what either actually is, was another example of the prevailing rhetoric in the world polity related to progress and liberal rationalism. Both senior management, programme management, government partners and civil society all expressed latent anxiety in being measured and assessed against indicators for progress and results.62                                              61 Emira Lubani, Interview by Ana Lukatela, Tirana, March 8, 2010. 62 Safete Beqiri, Interview by Ana Lukatela, Tirana, March 6, 2010. 144  In conclusion, the similarity of the rhetoric being used by actors across the cases points to an interesting mimicry of attitudes on gender mainstreaming without any clear incentives that would rationalize it. As one UN staff member said anonymously: gender mainstreaming has become only about the process, not the results. As long as everyone seems to be following the same process, in other words talking about including a gender perspective, no one is very worried about effectiveness. She sees this as a grave concern considering the investment and commitment the UN system has put into the concept. There is no ―critical thinking‖ she says, ―everyone is happy to check off the box and that is it.‖ Those respondents that displayed an understanding of what effective gender mainstreaming should look like felt frustrated by the mimicry they were witnessing. However the majority did not find it puzzling, but rather the similarity of rhetoric they were hearing seemed more like a vindication in their minds that they must be doing something right. Decoupling between Values and Action World polity theory outlines the non-rational mimicry of isomorphic values as the mechanism of diffusion of policies and it also predicts a ―decoupling‖ or a ―disconnect‖ between values and action (Meyer et al, 1997). The UN, NGOs and government all support gender equality and gender mainstreaming as a value, as evidenced by their rhetoric. However, they simultaneously put few resources into acting on these values and implementing policies.                                                                                                                                     Brunilda Dervishaj, Interview by Ana Lukatela, Tirana, March 9, 2010. Jelena Bjelica, Interview by Ana Lukatela, Pristina, May 11, 2010.  145  As described in the earlier section, senior officials and implementing staff all exhibited values in support of gender mainstreaming policies. However what became clear in the research was that an accompanying behavior was missing – the values were not paired with action neither in the form of technical assistance nor financial allocations. As World Polity theory hypothesizes, the lack of follow through on values is in part due to the disinterest in committing either human or financial resources to carrying out the policy. As one high level international official in Albania described, the sectoral gender working group convened by international and national actors was supported by both the UN and the government and chaired by UNIFEM, however, no one was sure if there was a budget or which actors had committed funds– neither from the UN side nor the government side.63  At the working level one programme staff member described the situation: ―When discussing it [gender mainstreaming] everyone nods their heads, and then when I come for the money they turn their heads… the justification is their budget is already allocated, but I am not 100 per cent sure that the funds are the main reason because it is not a lot of money. They are just not aware of the importance.‖64  So although everyone has committed themselves to the strategy of gender mainstreaming there is still a lack of understanding on what it entails in terms of budgetary allocation and its implementation.  And in reality the budget allocation is not large if a programme is properly gender mainstreamed in its design. For example it may mean bringing in a thematic gender sensitive expert                                             63 Anonymous, Interview by Ana Lukatela, March 2010. 64 Anonymous Albanian Government Official, Interview by Ana Lukatela, March 2010. 146  to speak at a seminar on the specific theme under discussion. In comparison to the overall value of projects the allocation to mainstreaming gender is not a burden and yet those individuals that do not understand how gender mainstreaming is done in development programming assume it requires significant resources.  All national female staff in UNDP Kosovo interviewed for this study pointed to the fact that when a successful women‘s empowerment project lacked only three months of bridging funds to move into its second phase with fresh donor funds the former senior management took no action and the project ended due to lack of investment in the gap time period. Respondents attributed this not to direct sabotage, but rather to lack of interest in action on issues of gender equality. Declarative statements were assumed by senior management to be sufficient to move the policy along. Officials are aware that it would be ―inappropriate‖ not to express support for these projects however they continue to take no action – decoupling their declarations from investment and implementation. In the municipalities of Albania where various international development agencies worked on gender mainstreaming, Dolly Wittberger of ADA described how mayors in many municipalities ―understood it was not appropriate to say no to us, so they said yes,‖ but despite the expressed values ―they never took any initiative themselves.65                                             65 Dolly Wittberger, Interview by Ana Lukatela, Tirana, March 6, 2010. 147  The tactic of decoupling - saying yes and then taking no action - was present in other cases as well. In one of the case studies a committed programme staff member described how she had requested the setting up of an internal Gender Mainstreaming Task Force so that they could start monitoring concrete implementation steps. Senior management agreed to the idea, however despite her many inquiries on next steps the days turned to weeks and months and the task force was never established. This gave her the perception that management would never say no to anything to do with gender mainstreaming, however they could not be expected to take action on investing financial or human resources into her ideas.66 The gender focal point of WHO in Albania, Ledia Lazeri, described a similar situation of a ―disconnect‖ between expressed values and action in the form of human resources commitments. When she wanted to do a training on gender mainstreaming for the national implementing partner, the Ministry of Health, the ministry‘s senior management agreed and expressed full and eager commitment. However, on the day of the training the senior managers sent two very inexperienced and junior staff instead of participating themselves. WHO‘s assumption in the design of the programme had been that national participants would be the middle and senior management and not the administrative pool of staff. However the managers and directors had given the invitation to their assistants and administrators and instructed them to attend in their place.67 This demonstrates that decoupling is not only from financial resources but also from                                             66 UN Official, Interview by Ana Lukatela, 2010. 67 Ledia Lazari, Interview by Ana Lukatela, Tirana, March 10, 2010. 148  investment of human resources. Staff in at least five out of seven agencies in each country noted this disconnect between the rhetoric of support expressed by government partners and UN senior management and the actual actions taken, or human and financial resources allocated.  OSCE‘s gender focal point in Kosovo described her experience with decoupling within the organization. When she raised to senior management the idea of opening a kindergarten within the OSCE office as a way of mainstreaming gender issues into the organization‘s operations, the management gave her full verbal support for the idea; however every time she brought it up again they delayed and did not take any action. On the other hand when she suggested formation of a network of gender focal points placed in each programmatic pillar management again agreed, but then actually took action by nominating focal points and including the network in a new organizational organogram under preparation at the time.  She felt that only when minimal action was required was there any chance that the decoupling could be bridged and action taken.68 The gender focal point of a UN agency in Bosnia described a similar experience that left her with ―a bitter taste.‖69 Even though senior management was verbally supportive of her gender work, whenever she requested funds to travel to regional ―community of practice‖ seminars on gender mainstreaming her request was ignored or denied. She felt that these seminars and trainings would have been beneficial for her knowledge of how to implement gender                                             68 OSCE Official, Interview by Ana Lukatela, Pristina, May 2010. 69 UN Official, Interview by Ana Lukatela, Sarajevo, June 2010. 149  mainstreaming policy and she felt that at the same time there were huge amounts of money being spent on trivial things within the organization. Another clear disconnect between value and action can be characterized by the marginalization of the Gender Equality Mechanisms (GEMs)70 in each country by both national and international actors. While all development stakeholders expressed support of the gender equality mechanisms71 in each country, their funding and staffing levels where miniscule and the units were powerless and marginalized.  In fact the UN came to discover that partnering with the GEM, instead of working directly with a line ministry, can actually run the risk of devaluing the project in the face of other partner ministries or agencies. Several respondents described how working through the GEMs only led to bureaucratic battles of mandate and turf and to a quick decline in interest from line ministries in implementing gender mainstreaming. The GEMs marginalized the policy of gender mainstreaming and ―ghettoized‖ it as a women‘s issue. As one UNIFEM staff member said: If you let me cut my relationship with the gender mechanism and work directly with the Ministry of Justice I am going to do so much more and it will be institutionalized because the gender mechanism is not an institution in that it doesn‘t                                             70 GEMs are national structures responsible for developing and implementing gender equality policies. For example an Agency for Gender Equality in the Prime Minister‘s Office as in Kosovo. 71 In Kosovo ―Agency for Gender Equality‖ under Prime Minister‘s Office, in Bosnia ―Office for Gender Equality‖ and in Albania ―Equal Opportunities Directorate under Ministry of Labour.  150  have a budget, it doesn‘t have any staff, and it doesn‘t have teeth.72  The other major problem with the GEMs in all three countries appeared to be that their Director is politically appointed and is someone with little to no background knowledge or expertise on relevant gender equality and women‘s empowerment issues. As one interviewee sardonically quipped about GEM leadership:  ―It is always the First Lady‘s best friend‘s blah blah somebody.‖73  This is not to say that there are not capable individuals within national governments with a commitment to the norm of gender equality; however they avoid being appointed to the GEM, instead of a ministry, because they recognize the GEM is underfunded and would constitute degradation in status.  One UNIFEM staff member explained that she had worked with GEMs over six countries in her career and she found them all the same in terms of lack of effectiveness and marginalizing gender mainstreaming policy. She described it in detail:  The relationship is a very strange one that the UN and UNIFEM have with these GEMs, a very strange one. You really feel like you are the only one pushing for this little entity to exist and to be a power and we always end up misused by the mechanisms or misusing them. It‘s pushing a square peg into a round hole and we have to stop and look at the shape of the peg. Sometimes when I am convening a meeting at a very senior level with the government and I am inviting the GEM the government counterpart wants to cancel because to them it is no longer                                             72 UNIFEM Official, Interview by Ana Lukatela, 2010. 73 UN Official, Interview by Ana Lukatela, 2010. 151  a serious meeting. It‘s a serious meeting if the UN wants to talk about gender mainstreaming with them, but if you bring in these people [from the GEM] - who are they?‖ They have no legitimacy or authority because they are political appointees and not competency-based.74 In conclusion, a clear ―decoupling‖ was identified between expressed values and verbal commitments vs. taking action to allocate human or financial resources. As the evidence showed this gap is largely explained by elements of the world polity theory that expect to see declarative support, lack of understanding; and isomorphic values. However it is also related to the bureaucratic and political struggles that often accompany policy adoption in organizations, as discussed in the following section. Bureaucratic Politics  The bureaucratic politics theory argues that organizational policymaking depends in large part on officials who are influenced by their own personal goals, which are often in flux and can rest on factors such as ambition or their perspective of worthy organizational goals (Allison and Zelikow, 1999).  Two types of competitive politics emerged in the research. First, there were the personality- driven conflicts between policymakers within organizations. Second, there were also policy-driven conflicts in which policymakers tried to promote organizational goals or overcame a personality conflict because of an interest in adopting a particular policy. There were no resource-driven conflicts                                             74 UNIFEM Official, Interview by Ana Lukatela, 2010. 152  evident in the cases, which seems a likely indication of the low status that gender mainstreaming has beyond its declarative value.  In both dynamics the conflicts could impede the policy making processes regardless of the stated policy commitment. However one interesting facet of these conflicts is that the bureaucratic politics dynamics specifically between UN agencies was never about personal ambitions or personality driven; it was about organizational policy goals and mandate. To some extent perhaps this reflects the discipline with which UN agencies indoctrinate their staff to promote organization goals towards external ―others‖ and the extent to which strategic agency positioning is valued in performance appraisals. In any case an improved understanding of conflict resolution techniques, mediation and strategic management would benefit staff in all types of organizations to avoid these kinds of conflicts. So while policy adoption is influenced by the values of the world polity it is also contingent on bureaucratic dynamics and requires policy not be held hostage to personal or political conflicts. Personality Driven Politics Personality driven politics were most often identified as an obstacle to working with the national gender equality mechanisms. This is likely due to the lingering resentment between national and international stakeholders that is often present in development contexts and the result of real or imagined neo-colonialist attitudes or threats to sovereignty. 153  In Albania, a clash of ambitions led to a personality-based struggle75 between the Director of the Equal Opportunities Directorate and her direct superior, the Deputy Minister of Labour. This struggle within the Ministry of Labour, which is the major UN partner on gender mainstreaming, had brought to a standstill the UN‘s support to the process of developing a gender mainstreaming policy for national legislation. All UN and civil society respondents agreed that the feud was for status and credit, as opposed to a fight of policy goals. Both decision makers had given their declarative support to gender mainstreaming, but did not want the other to get credit for the policy‘s adoption.  In Bosnia and Herzegovina the Head of the Agency for Gender Equality was described by respondents as an individual with strong political ties that had used her strong negotiating skills to leverage relationships with key ministries and government officials to maintain her position and office resources. One                                             75 The political competition was described as a ―bureaucratic power struggle‖ between two personalities, both of whom had their own political faction. The Director and the Deputy Minister were each trying to push the other out of the political space and ensure any credit for an adopted policy belonged to them. The interviews with three different staff of the Directorate, as well as the Director herself, clearly conveyed the alienation and competition. Body language of staff betrayed the tense work environment within the Directorate. The interview with the Director resulted in nothing more than a list of her personal accomplishments and successes; with blame for the slow policy making process allocated squarely on her senior manager, the Deputy Minister. The poor relationship with the mid-level staff in the Directorate was due to their fear of reprisal. There was also an element of mistrust as the Director had indirectly taken action to remove staff members that had the institutional memory to challenge her approach and label it as destructive to the policymaking process.  Of the ten staff members that had been in the directorate since its founding, eight had left since her arrival. The remaining two were no longer permitted to contact or interact directly with their UN working level counterparts. Instead all communication was now channeled through and controlled by the Director. The Director allowed the interviews with her staff to proceed, however she insisted that an intern also be interviewed, who proceeded to laud the Director‘s substantive expertise and operational acumen. Anna Xheka, Interview by Ana Lukatela, Tirana, March 9, 2010. Etleva Sheshi, Interview by Ana Lukatela, Tirana, March 9, 2010. Brunilda Dervishaj, Interview by Ana Lukatela, Tirana, March 9, 2010. Irena Benussi, Interview by Ana Lukatela, Tirana, March 9, 2010. 154  anonymous respondent described her as successful in her work because of her methods:  She is good at pushing the government and playing bureaucratic politics. She does not push gender for the sake of gender; rather she works at keeping allies for her own sake, along with the work.  Personality conflicts also occurred between UN officials and other profiles of government officials.  UNIFEM in Albania experienced difficulties with the Ministry of Labour and moving policy development forward as they felt everything hinged on the personal relationship with the ministry‘s staff. UNIFEM had invested in its relationship with the previous Minister, Deputy Minister and Secretary of State to get the policy development process underway. Unfortunately, when the Minister and the Deputy Minister lost their seats in a cabinet reshuffle, the policy making process stalled. The new Minister did not trust his Deputy or anyone else at the policy level or the working level, to take the process forward. Respondents familiar with the situation in the Ministry of Labour described this as being in large part due to the extreme politicization of all issues in Albania and a general fear of ―making mistakes.‖ If no work is done then ―no mistakes can be made‖ was the simple explanation.76 As much as personality driven politics can stall policymaking, they can also drive it forward. In particular, civil society respondents in Albania described how they based their work on relationships with government staff at the mid and senior level of management. The high turnover in government officials means                                             76 UNIFEM Official, Interview by Ana Lukatela, Tirana, March 2010.  155  there is relatively little institutional memory and new relationships have to be built from the ground up each time.77 However once the relationships exist it is possible to move things forward quickly.  Another NGO worker described how important her personal contacts were in promoting gender mainstreaming policymaking with the UN and the government. She felt no progress was ever achieved on an institutional basis; official communication channels were useless. Instead, when she identifies an opportunity for the gender equality policy agenda she contacts her working level counterparts in the UN or the government and invites them for coffee in order to use her network to move policy adoption forward. In the case of Albania this is a more than necessary measure as all the formal communication and coordination mechanisms on developing gender mainstreaming policies were blocked for reasons described earlier. In the work of the Austrian Development Cooperation with Albanian municipalities, partners were eager to begin development of policies together with the local planner.  However, personality conflicts and a competition of ambitions between the director of the national NGO partner and several ministry representatives brought the project to a standstill. Dolly Wittberger, from ADA, described how the mistrust between the NGO and central government eventually resulted in the international development agency establishing their own NGO and working directly with the municipalities on policy development in order to sidestep and exclude both the central government and the national level NGO. This is not                                             77 NGO staff member, Interview by Ana Lukatela, Tirana, March 2010. 156  to say that the local level actors were immune from this type of competition; they themselves spent considerable time scheming against municipal rivals.78  This solution is far from ideal and in light of international development actors‘ commitments to promote the capacity development of national actors this is particularly concerning.  The UN in Albania was not immune from these personality based conflicts with government officials. All government respondents felt that a previous UN Resident Representative (RR) had made political gaffes related to a women‘s political empowerment project. One government official believed this event was partly responsible for UNDP‘s loss of interest in acting on gender equality as the whole UNCT became more risk averse afterwards, in order to avoid any further missteps. Policy Driven Politics In Kosovo the Head of the government‘s Gender Equality Agency was described by UN, government and civil society as a political appointee who lacked the qualifications for the position. However unlike in Albania and Bosnia she was not seen as a personality-driven manager and got along well with her staff and was recognized as having a sincere interest in gender equality policies and mainstreaming the work of the government in partnership with the UN and civil society.                                              78 Dolly Wittberger, Interview by Ana Lukatela, Tirana, March 6, 2010. 157  The Head of the Agency was seen as ineffective by civil society and her UN counterparts due to a certain naïveté in her attempts to navigate the corridors of power in the government. However she herself held a large dose of resentment towards the UN, and in particular UNIFEM‘s status as an intermediary between her agency and the bilateral international donor agencies. She felt like she was constantly competing for funds with UNIFEM and that this took all the attention away from the policymaking and put it on the resource mobilization process, meaning that no work was ever completed and no results could be shown.79  UNIFEM, on the other hand, felt that the Agency did not have the capacity to fulfill its policy and programme commitments and it was for this reason that bilateral donor agencies preferred to work through UNIFEM instead of directly funding the GEA.80  While there were no cases of personality driven competition within and between UN agencies, there were conflicts based on policy approaches. These tended to occur between UNIFEM, which held very specific feminist policy approaches, and UNDP and UNICEF.  Most of this conflict between UNIFEM and the other UN agencies working on gender equality and mainstreaming was evident in Albania and Kosovo. The lack of conflict in Bosnia can be attributed to the low capacity within that office at the time. The other agencies simply did not perceive UNIFEM as a threat to their mandate or policy approach; the head of UNIFEM‘s office in Bosnia at that time                                             79 Shqipe Krasniqi, Interview by Ana Lukatela, Pristina, May 12, 2010. Emin Emini, Interview by Ana Lukatela, Pristina, May 12, 2010.  80 UNIFEM Official, Interview by Ana Lukatela, Pristina, May 2010. 158  had been seconded from the Icelandic government and was described by UN staff from a number of other agencies as ―uninterested‖ in events in the country.  In Kosovo, UNIFEM and UNDP had experienced flashpoints of tension related to a policy approach on mainstreaming women‘s political participation into wider governance programming. Officials from both agencies mentioned this conflict, with the general feeling on UNIFEM‘s side that UNDP was stepping on their mandate, while UNDP felt UNIFEM was not being effective. However officials from the two agencies also described another instance in which they agreed on a policy approach and supported each other to mainstream a gender advisor into a large programme on resettlement of displaced persons and demobilized combatants. So in cases where their policy approach overlapped they were quick to collaborate. In Albania UNDP, UNICEF and UNIFEM found themselves at times proposing competing policy approaches on gender instead of coordinating due to different philosophical starting points. Although all respondents expressed the desire to work together in a more coordinated manner, they still recognized that this was a challenge due to their underlying assumptions on what gender equality means and how it can be achieved. Respondents recognized that the difficulties stemmed from the desire of each agency to protect its staff and mandate while promoting its own policy choice. Policy on combating and addressing domestic violence was one such issue where UNDP and UNICEF each proposed differing policy choices to government due to an inability to reach an agreement.  This led 159  to a period of tense relations between UNDP, UNIFEM and UNICEF and the domestic violence policy adoption process fell more than a year behind schedule.  Interestingly, in the three countries the only agency not to be a party to any of these bureaucratic politics conflicts was UNFPA. Perhaps this can be attributed to the agency‘s clear and simple mandate; UNFPA focuses narrowly on reproductive health and sexual violence policy-making and programming (which can be very contentious in conservative, traditional societies) and as such making progress in their work largely depends on the promotion of gender equality issues. In fact the Head of UNFPA in Albania described the arrival of UNIFEM as a huge relief for her. There was more coordination happening on gender mainstreaming within the UNCT and she felt that ―the presence of UNIFEM has made gender issues become more resonant within the UNCT.‖81 Within the UN agencies there were also conflicts between officials based on differing policy approaches. The most notable example was within UNDP Kosovo under the previous senior management. UNDP‘s former Resident Representative was surrounded by an inner circle of trusted advisors who monopolized his time and were not interested in gender mainstreaming. So for officials interested in promoting the adoption of a gender mainstreaming policy the most difficult aspect was finding a way to get the ―ear of the Resident Representative‖. Getting the initial meetings set up was an issue of competing with this inner power circle, which had its own policy priorities. However, once access was finally achieved and a strategy for gender mainstreaming was finally                                             81 Manuella Bello, Interview by Ana Lukatela, Tirana, March 5, 2010. 160  presented to the Resident Representative, it was clear that he would approve it, and respondents felt this was because mainstreaming was ―part of a global language no one will reject‖82 which is more evidence of the world polity‘s declarative support and isomorphic values. Other international organizations also experienced internal conflict based on differing policy approaches. In the case of the EU Mission in Kosovo (EULEX), it was clear that the Chief of the Mission had no interest in gender mainstreaming policy, perceived it as a threat to mission stability and his own priorities. He denied the Head of the Human Rights and Gender Unit access to the meetings of the senior policy management team, without any justification, even though by hierarchy she should have been included.  As a result there were absolutely no gender assessments done in the planning of mission projects and no gender mainstreaming in operations planning.  The Head of the Unit felt her work was sabotaged and paralyzed. Even though she had a substantial budget, without an organizational policy on gender mainstreaming she felt she was prevented from implementing her budget on anything other than meaningless conferences.  So even though the mission planners in Brussels had created a ―gender unit‖ (albeit folded into a human rights unit as opposed to a standalone one) the position was not given access to strategic mission policy and planning meetings.83                                              82 UN Official, Interview by Ana Lukatela, Pristina, May 2010. 83 Sirpa Rautio, Interview by Ana Lukatela, Pristina, May 13. 2010. 161  Within the government, respondents pointed to the example of Albania‘s former Deputy Minister of Labour, Marieta Zace, as an example of a femocrat and key government official who knew how to persuade others to join her policy agenda. One of her tactics to build support was to turn a positive spotlight on individuals she wanted to transform into allies: the Prime Minister, key ministers, etc. She profiled the Prime Minister in several large events related to the gender equality draft legislation where there was a large participation by the international community and ensured he came off with a positive image. These types of tactics gradually secured his full support for the adoption of the Gender Equality Law, which included a provision on gradual gender mainstreaming of all government operations and policy making processes.  In conclusion, this section outlined the types of dynamics that emerged between UN officials and national policymakers that could either slow down or speed up policymaking. Personality driven politics, in which who is involved in initiative takes precedence over what the initiative promotes, had debilitating consequences for policymaking and are difficult by nature to predict but still important to understand and identify due to their potential impact on policy processes. On the other hand, policy-driven politics, in which officials could not agree on a policy approach, could also stall policymaking. However, in the case of the latter, a united policy driven approach could also serve as a tactic to overcome personality clashes.  Despite the fact that an isomorphism of values exists among senior managers and senior national actors on gender mainstreaming, the resulting policy adoption depends also on the positive 162  alignment of personality and policy conflicts. In addition it is likely that these two factors can play a role for any type of policy process and are not necessarily specific to mainstreaming policy processes, unlike the world polity explanation of policy adoption, which is specific to the characteristics of mainstreaming described in this chapter.  Conclusion This chapter has explored the manner in which policymaking in international organizations is undertaken and has identified two mechanisms. The first mechanism is the overarching influence of world polity values on the attitudes and behaviors of actors dealing with the adoption of mainstreaming development policies. Senior international and national actors are unwilling to be perceived of as ―opponents‖ of gender mainstreaming, so they support related policies without understanding the meaning or implications. However actual policymaking is contingent on the second mechanism; the dynamics of bureaucratic politics, which can be either personality driven or policy driven, and can either block policymaking or be used as a tool to overcome differences. So while the influence of the values of the world polity may be a necessary condition for policymaking to be undertaken, it is not sufficient; policy making also requires that the politics within and between organizations is not holding policy hostage. The next chapter will examine the factors that influence whether a policy, once adopted, is implemented; how to move from paper to practice?   163   5. Policy Implementation: Normative Commitments Supported by Standard Operating Procedures   “[The UN] is not capable of delivering on gender until we sort out our house internally.”84 – UN Official   The UNDP Global Gender Mainstreaming Evaluation in 2006 highlighted a continuing gap between policy adoption and policy implementation. Identifying the factors that influence policy implementation is a key step in addressing this gap. This chapter outlines the causal mechanisms that contribute towards policy implementation based on evidence from the three case studies. It argues that policy implementation depends not only on senior management support, but rather quite significantly on the attitudes and normative commitments of middle management staff who are the ones leading programmes. In none of the cases studied was gender mainstreaming implemented systematically. Rather it was evident that on a programme by programme basis implementation was piecemeal and dependent on the programme manager. In cases where individual programme managers took an interest, irrespective of whether senior management was supportive, they would mainstream gender into their programme. In addition, the impact of normatively committed staff was multiplied in situations where they had access to standard operating procedures that could be leveraged to influence the behavior of relevant colleagues or demand                                             84 UNDP Official, Interview by Ana Lukatela, 2010. 164  accountability. However this was only effective if these colleagues did not have normative resistance to values of gender equality.  Without changing individual norms and the organizational power structures of inequality that they perpetuate and underpin, gender mainstreaming cannot be successfully implemented organization-wide. As long as the prevailing norms among staff are not supportive of gender equality, and the structure of the organization does not require and reinforce a specified pattern of values and behavior, mainstreaming policy will not be coherently implemented or lead to results.  Although there is evidence arguing that investing in women improves development outcomes (Justino, 2012; World Bank, 201285) not one single individual cited the ―evidence‖ for gender mainstreaming as the reason they supported policy adoption or policy implementation. Middle management implementers spoke about their personal formative experiences and values related to gender equality, as the constructivist approach would expect, and not instrumental arguments.  This chapter will frame the discussion through several key themes emerging from the research: first, the key actors for moving implementation forward are mid-level management and not senior management; second, an internalized normative commitment towards gender equality was the key factor for many of those who actively tried to take on implementation of gender mainstreaming in their programming area; third, smaller agencies demonstrated                                             85 The World Bank‘s 2012 World Development Report says: Gender equality is a core development objective in its own right. It is also smart economics. Greater gender equality can enhance productivity, improve development outcomes for the next generation, and make institutions more representative (World Bank, WDR, 2012). 165  a homogeneity and coherence in staff norms that led to socialization and a more concerted effort in implementing gender mainstreaming; fourth, standard operating procedures could be leveraged to influence the behaviour of colleagues who were passive towards gender equality and showed no norm resistance; and fifth, individuals with deep normative hostility to gender mainstreaming could still successfully evade standard operating procedures with no fear of sanctions. Overall, until these gaps are addressed in staffing procedures and the house is ―put in order‖, it is difficult to foresee significant progress.  Middle Management Implementers While senior management is important in terms of expressing support, allocating funds, and adopting gender mainstreaming policies, they are still not veto players; the research shows that the process of implementation generally lies beyond their attention and control. A programme manager is responsible for the design and implementation of their project and if they are personally committed to integrating a gender perspective, a racial perspective, or an environment perspective, they can do this easily without needing senior management approval. Senior management interest in programme implementation lies mostly in the area of whether fundraising is successful and delivery rate targets are met by year end and donor reports submitted in a timely manner – process issues, not substance.86 As mentioned in the previous chapter the independence of programme managers is also due to the fact the senior                                             86 Based on author‘s observations and experiences as a programme manager in the UN system. 166  management has no incentives or sanctions related to gender mainstreaming implementation. Performance appraisal systems focus on results of the programme under the manager‘s responsibility; the resources mobilized and delivered; support to corporate processes within their agency; and interpersonal relationships on the team. In their five country assessment of gender mainstreaming Hankivsky also highlighted as one of the foremost obstacles to the implementation of mainstreaming the fact that there are ―no consequences for not doing it‖ (Hankivsky, 2013).  The respondents in this study pointed to a clear divide between senior and middle management‘s roles. While senior management all mimic the same isomorphic discourse, middle management openly expresses variance in their personal beliefs and normative commitments related to gender mainstreaming. The study concludes that world values have not reached the same level of diffusion among  mid-level staff, especially national staff, as it has among senior management, all of whom are international staff and part of the community of international development professionals as described by Boli and Thomas (1999: 243). This can perhaps be explained by the lower level of exposure that mid-level national staff have to the world polity, as opposed to senior management. This is significant due to the power and influence of mid-level managers as the individuals overseeing the day to day implementation of programming.  Of more than 75 respondents across three countries, as well as Headquarters, all of the programme managers described their role as more important than senior management‘s when it comes to implementation, and more 167  than half of the senior managers interviewed also agreed middle management plays the most significant role for implementation. Programme managers can propel projects forward, purposely marginalize aspects they dislike or hand them a slow death due to disinterest.  In UNDP Albania, one respondent described how senior management reminded programme managers in no fewer than five different staff meetings that the organization had a new policy for mainstreaming gender into programmes. However no follow-up ever occurred; although senior management had adopted the policy, middle management easily ignored it or pled ignorance.87 The power of mid-level management to act on their interests and norms cannot be discounted; it proved to be substantial in these cases. The lack of an accountability structure and monitoring for follow-through by senior management contributed to this situation. The disinterest of middle management is more likely in contexts that do not have socialized or internalized norms supportive of gender equality and women‘s empowerment. Since many issues being mainstreamed, such as gender, HIV/AIDS, and disabilities, are controversial for traditional and conservative societies, the circumstances for mainstreaming are even more difficult.  The distinction between middle and senior management is not the only dichotomy that was raised by the research. Respondents also pointed to the difference between national staff and international staff at the middle level. While senior staff is always exclusively internationally-recruited, middle management is                                             87 UNDP Official, Interview by Ana Lukatela, Tirana, March 2010.  168  a mix of both nationally and internationally recruited staff. National staff begrudged international staff for receiving exponentially higher salaries for the same work and for taking up jobs that could go to their citizens; while many international staff felt uneasy regarding the lack of experience in development programming among national staff and the sometimes lack of normative commitment on issues that were perceived as foreign to the national culture, such as HIV awareness, gender equality, or environmental protection. This is not at all dissimilar from the conclusions that Jeffrey Jackson reached in his research on local and ex-pat international development workers in Honduras (2007).The theoretical framework presented in chapter two also discussed in detail the power of middle management staff in organizations and how principal-agent theory has difficulty explaining this phenomenon (Lee, 2013: 691-693). All the international staff interviewed in the three countries expressed their support of gender equality as a goal; however a number of these same interviewees felt that national staff did not understand gender issues and were not committed to gender equality as a development goal in the same manner as they were to poverty alleviation or anti-corruption. One staff member of an international organization in Albania said: ―basically it‘s the local national staff that runs the show‖ and ―most or all of them have no clue what gender means or how we apply it to men and women and their relationships.‖88 Despite these individual perceptions, the stereotype did not hold among the respondents to this study: in fact national women staff members were                                             88 Anonymous, Interview by Ana Lukatela, Tirana, March 2010.  169  more likely than any other category of staff to understand the problems of gender equality in the country and the challenges and obstacles to implementing gender mainstreaming.89 Concrete examples will follow below, however, what was clear from this research was that the lack of understanding among nationals had a more detrimental impact as they were more often the middle managers responsible for implementation.  A senior UNIFEM official recognized middle management as the key to implementation of gender mainstreaming. She felt that for a country programme as large and varied as UNDP‘s her method to tackle gender mainstreaming implementation would be to choose one cluster90 headed by a competent cluster manager who held the ―right values‖ and push gender to be mainstreamed through all the programmes in that cluster. This would serve as an example to other clusters in terms of both practical implementation but also the values the UN is trying to promote.  In other clusters, where the cluster manager has neither the normative commitment nor the interest, she felt that mainstreaming could never be successful and it was ―better to have one cluster truly mainstreamed than all the clusters only talking about it, but not doing it.‖91 This situation is only compounded by the fact that the clusters operate as ―different islands in the                                             89 A respondent‘s understanding of gender mainstreaming was measured through the questionnaire as to whether they were able to give a simple working definition of gender mainstreaming and could give an example of gender inequality in the national context they were working in.  90 Each UNDP country office has from five to six pillars as described earlier ( Democratic Governance, Poverty Reduction, Energy and Environment, etc) and each pillar has clusters within it such as a Social Justice cluster under the Democratic Governance pillar, or a Sustainable Development Cluster under the Energy and Environment pillar. 91 UNIFEM Official, Interview by Ana Lukatela, 2010. 170  same organization‖92. Each cluster keeps to itself and because they often do not physically share a space there are few chances to interact substantively.  Every single UN middle manager interviewed felt the biggest constraint on their ability to mainstream gender into their programmes was a lack of time due to overburdening of responsibilities. The result of this seems to mean that in the process of prioritizing deadlines, workloads and availability, the programmes in which managers have a personal interest are the ones that they become invested in and push forward, while others suffer. This is as true for mainstreaming as it is for any other type of policy or programming. In UNDP Albania respondents identified only two programme staff out of 40 who they felt were interested and trying to mainstream gender into their programming, while none of the remaining 38 faced sanctions for ignoring implementation of the policy. In UNDP Bosnia only three staff members were identified as interested and committed, out of over 60 programming staff. One UNDP official felt that in some cases the time burden on middle management meant that they are so overworked that they are ―terrified‖ of gender mainstreaming because they do not understand it and see it partly as a whole new assignment for which they have no time. However in other cases it was the lack of commitment to gender issues that was identified as the problem by this same individual.  One staff member of UN Volunteers93 working on a local government programme thought that the disinterest of programme managers bred                                             92 UNDP Official, Interview by Ana Lukatela, 2010. 171  misunderstandings of what a proper gender analysis entails and that this was the weak link in the implementation process. As an example she recounted the case of a program officer in a waste management project who could not conceive that his program could have a different impact on men and women; that women might access sanitation services differently from men, that there might be obstacles specific to women, etc.  She felt that a lack of interest or knowledge in an adult programme manager with fully formed attitudes could never be compensated and that the initial mistake was to hire them in the first place.94 Civil society partners interviewed in all three countries also felt that the UN‘s Achilles’ heel was its disinterest in tracking down available national gender sensitive expertise to hire as programme managers.95 Not all of the senior managers interviewed were aware of the lack of capacity among their programme managers to understand and do a gender analysis. However one in particular did recognize that the biggest problem for the implementation of the official gender mainstreaming policy was the lack of ―internalization‖ of gender analysis into middle management‘s everyday perspectives. The lack of ―buy in‖ as she phrased it, was the first part of the problem with implementation and the second part was the lack of understanding                                                                                                                                   93 The United Nations Volunteers (UNV) programme is the UN organization that contributes to peace and development through volunteerism worldwide. Based in Bonn, Germany, UNV is active in almost 130 countries and has Field Units in 86 countries.  94 Ingrid Baken, Interview by Ana Lukatela, Tirana, March 8, 2010. 95 Memnuna Zvizdic, Interview by Ana Lukatela, Sarajevo, June 16, 2010. Monika Kocaqi, Interview by Ana Lukatela, Tirana, March 5, 2010. Safete Beqiri, Interview by Ana Lukatela, Tirana, March 6, 2010. Mirela Arqimandriti, Interview by Ana Lukatela, Tirana, March 9, 2010. Luljeta Vuniqi, Interview by Ana Lukatela, Pristina, May 12, 2010. Igballe Rogova, Interview by Ana Lukatela, Pristina, May 11, 2010. 172  and expertise. She felt that ―gender equality specialists are not what we need. We need technical experts who understand how to do a gender analysis.‖96  An example that she felt proved her point was the way in which the UN chose the very first Special Representative of the Secretary General for Sexual Violence in Conflict. Advocates of the idea expected the Secretary-General to appoint an individual with a military background (sector expertise) who also had an understanding of gender issues. However to their surprise the successful candidate was Margot Wallstrom, a former Swedish politician, who had no military background or experience but was seen vaguely as a ―gender equality advocate‖ with a social justice background. The hope that the position would have a real impact on military and peacekeeping personnel was lost with the appointment because without sector specific expertise she was unable to successfully engage the military contingents she was supposed to be influencing.  In conclusion, senior management mimicked the values of the world polity; however this was not a factor at middle management level. Respondents did not identify any senior managers as openly opposed to gender mainstreaming, however they easily identified mid level programme managers who were openly opposed or derisive of gender equality as a goal and of mainstreaming as a tool. Multiple respondents independent of each other identified the same problematic individuals within their agencies and described the same incidents, giving reliability to the accounts. In addition a review of the project reports of the                                             96 UN Official, Interview by Ana Lukatela, 2010. 173  identified individuals confirmed that the programmes under their responsibility did not have any elements of gender mainstreaming.   Vibeke Risa, the Deputy Country Director of UNDP in Albania felt that this has been a typical phenomenon in the more than ten years she has been with the UN and across the different country offices she has worked in. While senior management mimic each other‘s declarations of support (and which she felt they usually did not understand) this mimicking was not visible in the subordinate staff, especially the national male staff.97 The influence of the world polity is clearly stronger on international development professionals that have lived and operated across a number of country offices.  So although all middle management programme staff has a responsibility to implement a gender mainstreaming policy it is only when a personal interest aligned that this became a reality. The quality of staff assigned to work on mainstreaming projects was also a common point emphasized by respondents. UNFPA, UNIFEM and UNDP, three agencies that work on many different types of projects with the government, complained that when it came to gender mainstreaming projects the level and capacity of the government employee assigned to the project always appeared to be lower than for other projects not dealing with mainstreaming, which were treated as more serious. UNFPA in Albania in particular described their ministry counterparts as low level political appointees who had no technical knowledge. Manuella Bello, Head of UNFPA in Albania, felt that when trainings were held on                                             97 Vibeke Risa, Interview by Ana Lukatela, Tirana, March 4, 2010. 174  a mainstreaming project the government tended to send lower level women employees to participate as opposed to policymakers or decision makers. 98  One UNDP staff described how a training session for government staff on mainstreaming started late as the only employees to trickle in were female secretaries and assistants who had been told only that morning by their superiors that they had to attend an all-day ―gender and women‘s‖ training. National counterparts voice their support for the UN‘s gender mainstreaming trainings, but due to disinterest fail to participate when the time comes.99  Ultimately, the importance of programme manager's agency in international organizations has been under recognized in organizational theory literature. The focus of organizational pathology literature and neo-institutionalism has not been on breaking down mechanisms at the political and operational levels (Barkin, 2006: 33-36). While policy may be adopted by senior management this in and of itself is not a sufficient condition to lead to policy implementation. If the implementers at the middle management level are disinterested then progress is unlikely.  Normative Commitment Middle management clearly plays an important role in policy implementation, so identifying the factors that influence an individual‘s interest and engagement is a critical part of the policy implementation question. Overall, the research found that questions of gender equality and gender mainstreaming                                             98 Manuella Bello, Interview by Ana Lukatela, Tirana, March 5, 2010. 99 UNDP Official, Interview by Ana Lukatela, Tirana, March 2010.  175  elicited emotional responses in the respondents, and that they tied the question closely to their personal belief system and set of values. The individuals who displayed the most understanding of gender mainstreaming, and were identified by others as committed to policy implementation were those who held the strongest normative commitments about gender equality and women‘s rights.  These norms were something the respondents described as having formed during childhood, either through the influence of a strong woman in the family, or after a specific life experience that they perceived as a gender injustice and that served as a catalyst in changing their consciousness. They were also very likely to express sentiments that they found the decoupling between values and action in senior management and other colleagues as highly demoralizing for their own work. Those who displayed reticent emotional responses to the policies either saw gender inequality as non-existent (i.e. women are already equal) or perceived inequality as part of the natural order. Indeed normative values proved to be the underpinnings of each manager‘s personal response to the implementation of mainstreaming policy. The middle management programme officers who described themselves as normatively committed to gender mainstreaming shared several characteristics. They tended to be female (although not exclusively) and could point to strong formative experiences in their personal backgrounds that made them sensitive to gender-based discrimination and the need to address this 176  challenge through development programming. They also described perspectives or views in their past that could be situated within the feminist body of thought. Scholars who have examined feminists within organizations have defined feminist consciousness as ―the belief that systematic gender discrimination exists and it is wrong and collective action is required to undermine it‖ (Taylor, 2005: 362). One particular study on feminist consciousness within organizations found there were three mechanisms through which individuals who applied to the position of ―gender advisor‖ in an organization came to hold strong views on gender equality. The first mechanism is by witnessing unequal treatment, the second is experiencing frustration with organizational stasis, and the third is feeling isolated, stereotyped, and discriminated against as a result of being identified by others as a ―feminist‖ (Mansbridge and Morris, 2001). Among the respondents of this study, those that identified a feminist consciousness (31 - including also three male respondents) did so either because they identified a strong woman role model figure during their formative youth, or because they described a life event related to gender injustice which shaped their belief system from that point forward.   In the case of strong female role models respondents referred to specific individuals who had shaped their belief system during their formative youth. This included figures such as grandmothers, an aunt and mothers who were specifically identified as ―strong‖ or ―formidable‖ women, who left ―lasting impressions‖.  One respondent said her strong belief in the necessity of women‘s empowerment came from her aunt: 177  I guess I got these things from her as a child and this [the respondents work on mainstreaming] is like a continuation of the tradition. So things like stereotypes and gender roles and emancipation were part of my upbringing.100  Another respondent refers to her mother:  I was raised in a family environment where there was only me and my brother and my mother tried to make sure everything was fair between us and this was unusual for the time and helped me become who I am…gender equality is something I have been thinking about all my life.101 All these individuals recognized that they were unusual in these strong beliefs as compared to their colleagues, who had little awareness or little interest in these issues or any understanding of the concept of gender.  The other specific pattern that arose to explain the presence of strong normative values was a compelling and formative experience of gender injustice during the respondent‘s youth. One reminisced how she first became aware of gender inequality issues when she had to clean the class with the other girls while the boys would get to go out and play soccer, which made her aware of how unfair society is for girls.102 In another case a respondent at a fairly young age remembered being aware of how differently the sons in her extended family were treated compared to the daughters and how she resented this in her                                             100 UNDP Official, Interview by Ana Lukatela, 2010. 101 Fiorela Shalsi, Interview by Ana Lukatela, Tirana, March 4, 2010. 102 Emira Lubani, Interview by Ana Lukatela, Tirana, March 8, 2010. 178  youth.103 Yet another respondent referred to her feminist consciousness as arising from her resentment of the unfair division of chores in the family and the everyday injustices she witnessed towards women and girls in society.104  At least five individuals recalled formative experiences with victims of domestic violence (in Albania)105 or victims of conflict related violence (Bosnia, Kosovo)106 and how these interactions formed their beliefs and values on gender relations. Two individuals pointed to particular workplace injustices; they pointed to the biased structure of promotion requirements to the disadvantage of women107 and to the fact that women who have to leave work on time to pick up children from childcare are discriminated against when senior managers organize late hour meetings and responsibilities.108  The head programme officer of UNFPA in Albania was a strong gender equality proponent. She identified prevalent constraints on women‘s security and freedom in Albanian society, partly based on the traditional kanun laws that treated women as property of their husbands. She recalled always feeling angry about this, but only becoming active after the experience of her medical                                             103 UN Official, Interview by Ana Lukatela, 2010. 104 UN Official, Interview by Ana Lukatela, 2010. 105 Manuella Bello, Interview by Ana Lukatela, Tirana, March 9, 2010. Emira Shkurti Interview by Ana Lukatela, Tirana, March 5, 2010. Monika Kocaqi Interview by Ana Lukatela, Tirana, March 5, 2010. Eglantina Gjermeni, Interview by Ana Lukatela, Tirana, March 11, 2010. Entela Lakho, Interview by Ana Lukatela, Tirana, March 10, 2010. Mirela Arqimandriti, Interview by Ana Lukatela, Tirana, March 9, 2010. 106 Memnuna Zvizdic, Interview by Ana Lukatela, Sarajevo, June 16, 2010.  Klelija Balta, Interview by Ana Lukatela, Sarajevo, June 15, 2010. Maida Cehajic, Interview by Ana Lukatela, Sarajevo, June 16, 2010. Virgjina Dumnica, Interview by Ana Lukatela, Pristina, May 4, 2010. Visare Nimani, Interview by Ana Lukatela, Pristina, May 10, 2010. 107 Alex Standish, Interview by Ana Lukatela, Pristina, May 7, 2010. 108 Arbena Kuriu, Interview by Ana Lukatela, Pristina, May 13, 2010.  179  residency, which was a turning point. During her residency she was posted in a hospital in the obstetrics ward. On multiple occasions she treated women who had undergone unsafe ―backroom abortions,‖ and after complications such as infection or uncontrolled bleeding ended up in the hospital. Tragically, many of her patients died of the consequences of these unsafe abortions. She recalled her feeling of horror that in the ―modern age‖ when contraception and safe abortions should be easily accessible to women who are in need, there was still unnecessary suffering and death. She also recalled that none of her male colleagues seemed troubled by the situation or saw it as needless suffering on the part of these women.109  In Albania, respondents from UN agency staff, bilateral international development actors and civil society representatives was an individual from the national government, the former Deputy Minister of Labour and Social Affairs, Ms. Marieta Zace. She was recognized as a tireless advocate for gender equality within the government‘s policy and planning processes and she was a critical force in the development of the National Action Plan on Gender Equality and Domestic Violence which included specific mainstreaming aspects.  Ms. Zace felt that her personal background as a young woman in sport (in her youth she had been a member of the national championship volleyball team) had given her strength and the opportunity to travel, which led to a growing awareness of discrimination in her society and gave her a deep belief in the                                             109 Manuella Bello, Interview by Ana Lukatela, Tirana, March 5, 2010. 180  importance of gender equality.110 Others described her as a deeply committed individual who would not compromise on the issue of gender equality.  In the middle of the continuum of views regarding gender equality there were individuals who could be described as holding a naïve disbelief in the existence of gender inequality. Some of the respondents made these claims themselves, and in other cases respondents described these behaviours and attitudes in their close colleagues. The typical opinions expressed were that ―women are not discriminated against anymore" and that women had been equal in the region ever since socialism and had all the same rights as men with no problem in exercising them. These individuals generally pointed to examples from their office, arguing that women and men worked together in the office on equal terms, which was proof of an egalitarian society. In cases where discrimination or equality was clear, these respondents easily dismissed it as an exception and not the rule.  On the other end of the continuum were individuals, all exclusively male national UN or government officials at the middle management level, who believed that gender inequality was natural. One respondent described encountering statements within the office such as ―God gave men more power and it is natural‖ or ―family voting is perfectly fair in the context of our culture‖111. The Deputy Resident Representative of UNDP in Bosnia summed this up when he recognized that ―gender equality elicits an emotional response‖ and                                             110 Marieta Zace, Interview by Ana Lukatela, Tirana, March 5, 2010. 111 Virgjina Dumnica, Interview by Ana Lukatela, May 4, 2010. Entela Lakho, Interview by Ana Lukatela, March 10, 2010. 181  individuals often have strong views on the topic, which can be a problem within the context of the work of middle management programme officers:  If a person does not have the personal attitudes it won‘t work, we [UNDP] have to hire and recruit based on this issue, it is the only way to do it. As a senior manager I can only do so much on implementation. You have to have senior management lead, but without the right views of people nothing will work. There are many people in the system that don‘t care at all about gender equality – how can we have these people in the UN system!? It is part of our founding framework. There is no mechanism to make people think a certain way, which means you have to hire the right people.112 This senior manager‘s statement is recognition of the primacy of middle management when it comes to implementation. He felt that senior management are important leaders, however without staff with normative commitment to gender equality this will not make a difference. This dissertation argues that given the power middle management holds over implementation senior managers may be useful as leaders; however they are not a necessary condition. If a staff member is interested and committed to applying a gender lens to their project, which is fully under their responsibility to implement, they can easily do this without consulting or needing the support of their senior managers in the agency since each programme manager is very much ―king‖ or ―queen‖ of their own island. This does not mean that senior managers‘ rhetorical and substantive support might not play an important role with other types of managers who                                             112 Peter Van Ruysseveldt, Interview by Ana Lukatela, June 14, 2010. 182  respond only to accountability sanctions, as will be discussed later in the section on how standard operating procedures may be used.  A senior UNIFEM official also raised the question of how ―to make people think a certain way.‖ She felt that there was a fundamental flaw with the assumption that it was possible to "teach formed adult personalities‖ to implement a policy if they have internalized opposing norms.113 Literature on adult learning also takes this view (Lieb,1991; Fidishun 2000; Knowles, 1984) and argues that special techniques are required to teach adults to use specific skills or knowledge that are based on a norm they have not internalized.  This view was supported by a UNV senior official who felt training was useless if people were not open to the fundamental values underpinning the issue. She felt that in all of UNDP Albania it was clear that only one person was interested in gender issues and in gender mainstreaming programmes and that no other programme manager in the office even thought about it.114  Another programme manager in a UN agency felt that in her experience working on various projects and programmes in the UN everything ―depends on the norms and commitments of the partners‖ and since this was ―a society where men decide everything‖ of course gender mainstreaming had not happened.115  In conclusion an individual‘s normative framework on gender equality plays a central role in their interest and commitment to implementing a                                             113 UNIFEM Official, Interview by Ana Lukatela, 2010. 114 Ingrid Baken, Interview by Ana Lukatela, Tirana, March 8, 2010. 115 UN Official, Interview by Ana Lukatela, 2010. 183  mainstreaming policy within the programme they manage. If a programme manager does not recognize inequality and does not believe it is wrong, they are not likely to take steps to think about how their programme can have an impact on patterns of exclusion of women. The difference between managers who tried to implement gender mainstreaming and those that did not was in relation to their belief in women‘s empowerment and a personal history related to women‘s rights or gender injustice.  However an interesting result of this study is the finding that even programme managers with an attitude of ―benevolent disinterest‖ towards gender equality are still not implementing gender mainstreaming; they did not understand it and preferred to avoid what they saw as extra effort. They were neither willing nor interested to take active steps to implement a policy to which they did not feel any particular normative attachment. The fact that the ―passive middle‖ actually falls on the side of non-implementation suggests that the challenges to mainstreaming are more significant than previously recognized. The circle of successful implementers in any given organization is significantly narrowed as ardent supporters on most policy issues are usually a smaller group than the circle of disinterested individuals plus norm resistors. Agency Dynamics Policy implementation was also influenced by the characteristics of and dynamics between UN agencies. Two factors related to the agencies proved to be important: first the mandate and size of the agency had an important impact on both the level of norm homogeneity among its staff and the ability of the 184  agency to react flexibly and quickly to opportunities as they arose; and second the level of cooperation between agencies in a joint UN gender theme group was largely contingent on whether they came to the table as committed individuals or as representatives of their organizations and mandates. In smaller groups there was a level of trust that allowed this type of cooperation to grow, while in larger groups this proved difficult due to a dilution of responsibility and interest. Size and Mandate The smaller agencies displayed a significant level of norm homogeneity among staff as compared to a larger agency such as UNDP or UNICEF. This phenomenon is in part due to the specific mandates of these smaller agencies, which are more directly linked in some way to the question of gender equality. UNIFEM of course directly works on women‘s empowerment, while WHO and UNFPA both deal with aspects of women‘s health rights.116  Respondents felt that the smaller agencies, by virtue of their smaller teams, placed greater emphasis during the recruitment process on the ―fit‖ of the candidate within the team and the maintenance of team cohesion. The gender focal point in WHO felt that the small office and the nature of the agency‘s work meant that the staff was mostly made up of ―civic- minded‖ medical doctors. She explained that when ―it was described as a matter not of gender equality but of                                             116 The World Health Organization‘s (WHO) is responsible for providing leadership on global health matters, shaping the health research agenda, setting norms and standards, articulating evidence-based policy options, providing technical support to countries and monitoring and assessing health trends. The UN Population Fund Agency, UNFPA, supports countries in using population data for policies and programmes to reduce poverty and to ensure that every pregnancy is wanted, every birth is safe, every young person is free of HIV/AIDS, and every girl and woman is treated with dignity and respect. 185  how to maximize public health‖ then the normative commitment was tied not to women‘s rights but to the Hippocratic Oath, which was accepted by all staff as the underpinning tenet of their work. In addition, there was no perceived normative controversy or emotional response to discussing evidence of increased cardiovascular disease in men and the risk of breast cancer in women; these were not seen as ―subjective‖, which is how the concept of gender discrimination was characterized in other cases. The WHO gender focal point felt that for this reason she did not encounter the ―emotional fights‖ on gender equality that the UNDP gender focal point was encountering. She also thought an important factor was that risky health behaviors are more common in men than in women, and men also have higher incidences of mental health illnesses, so the focus of their gender mainstreaming work was not necessarily on challenging social norms about women‘ position in society (as mainstreaming gender into a governance program in UNDP would have to do), but rather on improving men‘s and women‘s health based on their differing risk factors. 117 UNIFEM was another small agency that had very high norm homogeneity among its ranks, mostly due to self-selection; presumably those individuals that would be interested in working for a development fund for women have an existing deep normative commitment to gender equality. One UNIFEM official described the huge shift in attitudes towards gender equality that she experienced when she moved from a position with UNDP to UNIFEM. In UNDP she could not identify anyone among her peers and colleagues who believed that                                             117 Ledia Lazeri, Interview by Ana Lukatela, Tirana, March 10, 2010. 186  discrimination against women was a real issue affecting the country. However in UNIFEM she found her fellow staff members to be singularly committed to women‘s rights at the normative level.118 In the one case where a UNIFEM senior manager did not self-select to work for the agency, but was instead seconded to the UN by a national Ministry of Foreign Affairs as a contribution in-kind, the results were comic: all UN officials interviewed, regardless of their level, commented on the lack of action by the resident UNIFEM office and expressed surprise that the ministry‘s selection process would deliver an individual to head a fund for women‘s empowerment who had no interest in gender equality. One respondent saw UNIFEM‘s small size in relation to the other agencies as a compounding factor in its attempts to engage the right partners in gender mainstreaming work.  She perceived the focus on national women‘s machineries (i.e. an Agency for Gender Equality) as a ghettoization of issues and as the anti-thesis of gender mainstreaming. She felt the partners that UNIFEM should be engaging were the relevant thematic ministries, as opposed to the women‘s machineries of government, and she made an insightful link to internal UN structure: Who are our partners? You need to mainstream into the line ministries and ignore the national women‘s machineries. This ghettoization makes it no one‘s responsibility; it mirrors the relations within the UN where you have uninterested line programme managers and one gender advisor who is ineffective. Everyone has to do it [gender                                             118 UNIFEM Official, Interview by Ana Lukatela, Tirana, March 4, 2010. 187  mainstreaming] for their own portfolio, there is no other way.119 Not only was UNIFEM‘s relationship with line ministries problematic, so was the relationship with UNDP on occasion. One UNIFEM staff member anonymously felt that the relationship was ―difficult‖ due to the power imbalance: UNIFEM was a tiny agency as compared to UNDP, but because ―we really want them at the table, we don‘t criticize them as much as we should and we don‘t push them… we are a mouse in bed with an elephant. We let them get away with Women in Development and we all pretend it is Gender Mainstreaming.‖120 This sentiment supports the statement above, that the role of gender advisor is ineffective at pushing line partners.  However respondents also felt that smaller agencies had an easier time of trying to take concrete steps on gender mainstreaming not only because of the norm homogeneity but also because they had fewer programmes and operations to think about.  One UNIFEM staff member121 felt that the positive side of their small size was their agility and flexibility in their everyday work and in focused areas could have more of an impact. UN Habitat, another small UN agency, felt that working with UNDP was difficult because they were large and unresponsive, and that as a small agency they could take more focused action and move more quickly when opportunities arose.122 Also, respondents from the smaller agencies mentioned that due to their size cooperation was a critical value and learned                                             119 UNIFEM Official, Interview by Ana Lukatela, 2010. 120 UNIFEM Official, Interview by Ana Lukatela, 2010. 121 UNIFEM Official, Interview by Ana Lukatela, 2010. 122 Krystyna Galezia, Interview by Ana Lukatela, Pristina, May 4, 2010. 188  behavior. Instead of imposing their views and plans on civil society and government partners they had to consult and cooperate effectively. In general both other UN agencies and civil society and national government partners felt that UNDP lacked this behavior in both its procedures and staff mentalities. As one civil society activist asserted: ―UNIFEM is much more open to genuine collaboration and cooperation than UNDP.‖123 One of the issues raised by several international staff members was the impact of the non-transparent national recruitment processes within UNDP.  One respondent anonymously stated: ―People get positions here in a very strange way, they do not have the capacity or the commitment [to development] but they get the position.‖124 Recruitment processes that are not transparent and result in the hiring of candidates of dubious normative commitments is clearly a factor when it comes to implementing a norm-based policy and maintaining the overall coherence of the organization.   UNDP‘s lack of normative homogeneity was mentioned as a negative factor for gender mainstreaming. Gender focal points in UNDP across the cases felt they were often limited in their efforts to implement gender mainstreaming due to a lack of interested programme staff to collaborate with, while gender mainstreaming of operations was not even a topic they felt could be safely broached. One programme manager described how in his discussions with colleagues many admitted to not reading emails from the gender focal point with                                             123 Monika Kocaqi, Interview by Ana Lukatela, Tirana, March 5, 2010. 124 Anonymous, Interview by Ana Lukatela, 2010. 189  general gender mainstreaming tips as they were not interested in gender equality. The problem also seemed to be partly due to the wide range of portfolios held by programme managers across the pillars of UNDP‘s programming.125 The general instructions provided by the gender focal point were shared with colleagues working on such diverse issues as small arms and environmental protection and many did not seem to understand that this was applicable equally to such varied portfolios. The diversity of UNDP‘s mandate and programming seemed to make it difficult for colleagues to understand why gender equality was important for ALL their portfolios.126 Another programme manager who was sympathetic to the gender focal point‘s mandate agreed that those colleagues who were not personally invested and interested in gender equality were not reacting to the gender focal points instructions and were not facing sanctions for their behaviour. An agency‘s mandate could also have an impact on how staff perceived their approach to gender equality or gender mainstreaming.  UNICEF proved to be a special case across all three contexts; it was clear that their framing of gender relations was through a prism of the best interests of the child. As one UN official said: ―UNICEF is hands down the single most complicated agency to work with in any country.‖127                                              125 Democratic Governance, Poverty Reduction, Crisis Prevention and Recovery, HIV/AIDS  and Energy and Environment 126 UNDP Official, Interview by Ana Lukatela, Pristina, May 2010. 127 UN Official, Interview by Ana Lukatela, 2010. 190  UNICEF is in part seen as a difficult partner because of their excellent standalone public relations and fundraising capacities, which means they have better public relations and fundraising opportunities on their own, than when they work through the UN system. This is certainly tied to the ease of appealing to donors on the basis of ―children‘s welfare.‖ Unlike women or men, it is agreed that children do not have agency, and are always innocent victims of poverty, conflict, or poor governance. Partnering with UNICEF on gender mainstreaming is often seen as difficult because their staff has ―deeply ingrained and set beliefs that children ultimately take priority over women.‖128 The curious aspect seemed to be that UNICEF staff were not recognizing the gender identities of boys and girls themselves and seemed to be ascribing a gender identity only to men and women.   Their approach became contentious through joint UN programming in Albania on domestic violence. UNDP and UNIFEM supported a critical feminist approach of advocating for the transformation of societal gender relations, while UNICEF advocated for the empowerment of women within the context of their roles as caregivers in families (in essence viewing women only through their roles as mothers), without touching on underlying societal gendered power structures and inequalities. Given UNICEF‘s mandate of child protection their inability to perceive of women as anything more than mothers is perhaps not surprising.  Their insistence on working within the framework of families as the unit of intervention, instead of individuals, was rooted in their conceptual vision of                                             128 UN Official, Interview by Ana Lukatela, 2010. 191  what is ―best for children.‖ UNDP and UNIFEM felt this stress on the maintenance of the family unit came at the cost of the woman‘s/mother‘s safety. UNDP and UNIFEM criticized UNICEF for short-sightedness and a failure to recognize that once the girl-child grows up she is herself at danger of living in a vicious cycle of domestic violence with her own children unless mothers and women are empowered to stop the social and economic inequality that marginalizes them and lends to the root causes of domestic violence. In the 2013 National Survey of Domestic Violence, carried out by the Ministry of Social Welfare and Youth and UNDP, 59.4% of women in Albania reported experiencing  domestic violence.  Although the UNICEF focal point was cognizant of root causes of gender inequality, ultimately her normative framework129 meant that mothers had to sacrifice themselves for children and accept injustice.  Another UN official reflected on this debate:   It was my personal belief that was conflicting with UNICEF‘s belief. I don‘t like to deal with domestic violence issues in a way that continually and repeatedly links women and children. The responses are different and the identification is different. Yes, ultimately they will leave the family as a group, but there are a number of instances where she is hit and the children are not, or where she is hitting the children, and I don‘t know that you necessarily see girls being hit more than boys in this country. Children are hit here. But also as a feminist I don‘t want to see them coupled, I don‘t                                             129 UNICEF official‘s normative framework repeatedly referenced the ―family structure‖ as opposed to women‘s rights. 192  like even that the domestic violence strategy and the gender equality strategy are one and the same. I don‘t like it. UNICEF of course deals with domestic violence, and they talk about the mother and the girl-child so they think they are dealing with it from a gender sensitive perspective, but I do not think so.130  UNICEF's perspective to domestic violence described above also demonstrates an interesting parallel to the pathologies of IOs described by Barnett and Finnemore (1999). Joint Gender Theme Groups Mirroring the cohesion and norm homogeneity of the smaller agencies, the Joint Gender Theme Groups appeared to be more effective at promoting gender mainstreaming implementation in the UNCT when they were composed of a smaller number of representatives with a personal commitment to gender equality. In Albania a core group of programme officers from across the agencies formed the joint gender theme group, coordinated by UNIFEM. UNIFEM‘s strong and positive leadership of the group was recognized by multiple UN, government and civil society actors. The head of UNFPA felt that ―the presence of UNIFEM and the joint gender theme group has made gender issues become more resonant within the UN country team.‖ The UNIFEM country programme manager thought that the success of the group (in terms of organizing a number of high level profile advocacy campaigns                                             130 UN Official, Interview by Ana Lukatela, 2010. 193  on women's empowerment issues and not related to gender mainstreaming as such) was due to the fact that it was small and that the representative of each agency was self-selected as opposed to appointed, so they were all cohesive in their interest and commitment:  You can tell they care, despite coming from different starting points. Everyone at the table may have other jobs but this is something they really care about and at the end of the day we all have deep mutual respect for each other and for working together within the UN country team to try to promote a gender joint programme and make sure everyone contributes from their own perspective.131 In Kosovo the gender theme group itself was inactive and UNIFEM was not perceived by other agencies as a leader. UNFPA had in the past made several attempts to revitalize the joint gender theme group; however the members failed to prioritize the group‘s activities or work and it remained inert. Representatives of other UN agencies felt the overall inactivity of the group was largely due to UNIFEM's lack of an international staff head of office who would be at a sufficient position in UN hierarchy to command the necessary influence needed for this convening role to be effectively fulfilled. In Kosovo, UNIFEM convened a ―Security and Gender Coordination Group‖ specifically for joint gender and security sector work, however other agencies‘ felt the group had notable gaps: UNDP was hesitant to give the group importance and the membership of the group was too diverse with what some                                             131 UNIFEM Official, Interview by Ana Lukatela, Tirana, March 2010. 194  perceived as too many marginal actors invited to participate, which watered down its membership.  The size of the group was seen as too large and unwieldy with an absence of a critical mass of normatively committed members. One member anonymously stated that the group ―has no dynamics, no results and no one sitting there is interested or senior enough to have any real influence.‖132  Some members of the group described protracted squabbles related to insignificant details such as the phrasing of the deliverables in the group‘s terms of reference. Others mentioned that the group members constantly avoided any workload investments and that this was an example of a lack of normative commitment and interest in gender equality issues. In general, while everyone felt the concept was positive, many respondents said the group‘s problem was the wrong membership and the lack of strong leadership. One member felt that UNIFEM had failed to be a principal player in moving the UNCT agenda forward on gender mainstreaming due to its focus on sub-regional projects; which was partly a result of being forced to compensate for the lack of work being done in the weaker UNIFEM Bosnia office. Unfortunately for UNIFEM Kosovo this meant they had less time to devote to UNCT issues and working with line ministries within their own context. However they also suffered from the lack of a senior leader, the representational level of the Kosovo Project Office was a nationally-recruited officer-in-charge, and later an international middle management programme officer, neither of whom were appropriately placed in the hierarchy of the UN country team.                                             132 UN Official, Interview by Ana Lukatela, Pristina, 2010. 195   In the case of Bosnia the joint gender theme group suffered from a lack of visibility and leadership, also as a result of the lack of leadership from UNIFEM and was completely inactive. Interviewees could not recall when the last meeting had even been held.  Leveraging Standard Operating Procedures  Gender focal points or programme managers committed to gender mainstreaming were able to leverage standard operating procedures (SOPs) to their advantage under certain conditions. The following examples will show that SOPs could be effective in steering the behavior of colleagues who were passive towards gender equality as a norm. However when these champions were trying to use SOPs to steer the behavior of an individual who held negative views on gender equality then it was not a successful strategy. In other words the negative normative commitments of these middle management staff were a stronger factor in predicting their behavior than the SOPs. These norm-resisting individuals were able to find methods to evade SOPs, as will be discussed below.  One SOP that was identified as useful by gender focal points and other committed staff was the process of drafting Terms of References (essentially a job description) for proposals by civil society organizations or for expert consultants. SOPs of the UNDP corporate gender mainstreaming strategy required the inclusion of a clause detailing the requirement for either a competency in ―gender‖ or for results indicators that specifically mention sex disaggregation of gender specific impact.  Committed staff felt that where there were no normative objections by their mid-level colleagues they could use this 196  standard clause later in the project cycle to justify why a gender sensitive consultant should be hired or a gender sensitive evaluation conducted. They cautioned that this only worked where the project manager did not hold negative norms regarding gender equality but rather they had a passive approach towards the issue of gender and thus could be persuaded. In interviews with project managers who demonstrated a passive awareness and commitment to gender equality their responses revealed that they did not engage in gender mainstreaming because they felt it was too difficult but would be willing to do it if it could be made ―easier‖.  For example several programme managers in Bosnia felt that it was their time burden that had led to the failure to gender mainstream their programmes. They agreed with the general premise that their programme must in some way have a different impact on men and women but were not sure what that entailed exactly or what the consequences of that would be in terms of changing programme design or how much extra work it would create for them. In general this set of managers could be described as passive towards the norm of gender equality: they neither actively promoted and declared support, however neither did they display open opposition.  One programme manager in the environment cluster in UNDP Bosnia, wanted to hire a consultant who could help him with gender mainstreaming his programme. However in over 50 applications he did not have any candidates with expertise in mainstreaming gender into environment programmes. He casually observed that senior management had adopted a gender mainstreaming 197  policy and then left middle management to figure out what it meant without any systematization. He felt that if there was a standard and systematized way of identifying an expert in both the environment and gender he would have used it. But unfortunately everything was left to chance in his perspective and he was not prepared to delay his programme and jeopardize his delivery rate. Instead he decided to drop the gender mainstreaming aspect even though he was not in principle opposed to it.133  Without a standard operating procedure in place that could systematize rosters with legitimate, sector-specific experts the implementation of gender mainstreaming was clearly stymied. The gender focal point explained that the existing rosters focused on the amorphous category of ―gender expert.‖ Peter Van Ruysseveldt, the Deputy Resident Representative of UNDP in Bosnia, also had concerns about these gender mainstreaming experts; he felt that these consultants have had no clear expectations placed upon them in the past and more often than not have left no clear results behind them, and he was unsurprised that given the lack of a system the programme managers were not moved from inertia.134 There are also examples of how standard operating procedures do not work on the category of programme managers mentioned above – those that are in open hostility towards gender equality. A standard operating procedure for most programme managers was an induction training that included a component                                             133 Igor Palandzic, Interview by Ana Lukatela, Sarajevo, June 17, 2010. 134 Peter Van Ruysseveldt, Interview by Ana Lukatela, Sarajevo, June 14, 2010. 198  on how to do a gender analysis. Respondents that had received this training felt it was only useful for those individuals who were still non-committal or passive towards the issue of gender equality. Those who agreed that gender inequality was a societal problem were bored by it and found it too simplistic, while those opposed to the gender equality discourse were alienated and hostile in the training.  As one respondent said ―the problem with gender training is that if it is not delivered properly, basically adjusted to your target group, it can do more damage than actual benefit.‖135 Another respondent who was responsible for organizing gender mainstreaming training in her organization commented that "you always have to know ‗when‘, ‗to who,‘ and ‗in what way‘ to present gender." She felt that if a training was strategically presented to the target audience then there was a way to move the agenda forward no matter who the audience.136 Program Approval Committees (PACs) are used by almost every UN agency as the standard procedure for initiating and implementing new programmes. By sitting on the PAC gender equality supporters and norm champions can have the opportunity to influence programme content. One gender focal point commented that although there was no institutional structure for her to look at every single programme document, she would volunteer to sit on the PAC as often as possible and use that opportunity to try to gender mainstream programming. She felt that in circumstances where the programme                                             135 UN Official, Interview by Ana Lukatela, Tirana, March 2010. 136 Nikolina Marceta, Interview by Ana Lukatela, Sarajevo, June 14, 2010. 199  manager was not hostile the PAC was a useful forum for ensuring her comments were taken on board.137  In Kosovo the gender advisor took advantage of a pilot UN initiative titled the ―gender marker138‖ system where each project was scored on a scale of 0-3 for its inclusion of a gender perspective. The gender advisor saw this as her attempt to introduce a measure of accountability to project content; however the weakness continued to be that there was no system in place to deal with the outcomes of this scoring or follow-up on projects scoring a zero (there are no sanctions for a poor score). Although every country with the pilot gender marker has to report on their scores once a year to headquarters this only has to do with the mainstreaming of gender into programming and not also operations. This means the actual structure of inequality within UN organizations remains unchanged with no systematic form of identifying problems. The programme manager of a small arms project in Bosnia for example mentioned that he had recently heard of the pilot gender marker, but since there were no sanctions attached to it ―no one cares.‖139 UNV in Albania had a champion for gender equality in senior management who was successful in including the requirement for a score on mainstreaming                                             137 Virgjina Dumnica, Interview by Ana Lukatela, Pristina, May 4, 2010. 138 The gender marker is employed in several UN agencies, including the Office for the Coordination of Humanitarian Affairs (OCHA), UNDP, UNICEF, ILO and UNV, to track gender-related resource allocations and expenditures. OCHA coordinates the application of the gender marker as a tool to improve humanitarian response by ensuring that the projects fully take into account the different needs and capacities of women, men, girls and boys. In 2010, the gender marker was rolled out in nine countries. The results of the gender marker roll-out demonstrated a 300 per cent increase in the number of projects mainstreaming gender; however this still means that only 8 per cent of the projects in the nine sample countries now ―significantly‖ include gender equality targets.  SG Mainstreaming Report, 2011: 12) 139 Jasmin Porobic, Interview by Ana Lukatela, Sarajevo, June 15, 2010. 200  gender into programming in the most recent staff evaluation and assessment of her team. She hoped this would be perceived as an incentive for middle management staff who were not performing on gender mainstreaming. However she recognized that until there were clear sanctions attached to a poor staff evaluation this would not happen.140 In Albania a UN staff member felt: ―The cluster [programme] managers are the critical people –they have to be sensitive and putting it under their responsibility might be a way to tackle it – make it part of their job description and include in their Terms of Reference‘s Indicators for Performance.‖141  In Bosnia the Deputy Resident Representative echoed the thoughts of his UN colleagues from other countries:  UNDP, Bosnia included, says it is committed to gender mainstreaming yet we do no evaluation or assessment on it or include in performance appraisal of staff. If we did this it would be different… [E]very project worth more than a certain financial amount should have to have a gender advisor to make sure the whole huge project is mainstreamed. This should be systematic in some way.142 UNDP deputy country director, Vibeke Risa, felt that having SOPs on gender mainstreaming would be extremely useful. She noted: The single biggest help would be if we had systemized and institutionalized things on gender mainstreaming our organization and our programming so that we all know what to do and                                             140 Ingrid Baken, Interview by Ana Lukatela, Tirana, March 8, 2010. 141 UN Official, Interview by Ana Lukatela, Tirana, March 2010. 142 Peter Van Ruysseveldt, Interview by Ana Lukatela, Sarajevo, June 14, 2010. 201  how. We need procedures and systematic tools. Senior managers have to be held accountable for performance reviews on gender mainstreaming and we need performance indicators to be measured against.143 The WHO regional director was also a champion with similar ideas. The WHO gender focal point explained that the regional director had tried to create a Terms of Reference for a reporting and monitoring structure that included gender indicators mainstreamed throughout. She felt that the ―value added in institutionalized structures [i.e. the reporting and monitoring structure] is that they are too elaborate for those who are undecided to do battle with.‖144 In other words, the path of least resistance for an undecided programme manager is compliance with a new structure, as opposed to working against it. A gender focal point in another UN agency shared her feelings that if an institutional part of her agency had been officially tasked with monitoring gender mainstreaming then it would have happened. But since no one was officially tasked and there was no structure in place it fell off the agenda and just did not happen.  She felt that "gender mainstreaming needs a mechanism to capture it and anchor it." It has to be everyone‘s responsibility to implement, but someone specifically has to have the responsibility to monitor.  In her organization the mainstreaming stays on paper. It is "not reflected in practice or in monitoring and evaluation systems and we need standard gender indicators if we are going to                                             143 Vibeke Risa, Interview by Ana Lukatela, Tirana, March 4, 2010. 144 Ledia Lazeri, Interview by Ana Lukatela, Tirana, March 10, 2010. 202  monitor and see if anything is actually happening. Senior management needs to realize it is not happening and there needs to be accountability."145 UNDP in Kosovo has a large resettlement programme that was relatively well mainstreamed as it had a dedicated gender officer, employed full time to undertake gender analyses and collect data on gender sensitive indicators and make sure all programme indicators were sex disaggregated. The programme manager was not opposed to gender equality as a norm and the gender focal point managed to sway several other key members in the PAC arguing that the programme document should have more than just a standard paragraph on gender and should in fact employ a fulltime gender officer. She argued that resettling families was impossible without specifically targeting women, in particular mothers and elders. This argument resonated with the other members of that particular PAC, none of whom were actively opposed to gender equality and the suggestion was accepted. The programme manager himself was not openly opposed to gender equality and accepted the recommendation for his programme made by the PAC.   A UNDP programme manager in Kosovo felt that everything in UNDP was about ―ticking boxes‖ and not being substantive. He felt that most of the SOPs used by UNDP systematized negative effects on gender mainstreaming rather than having a positive impact. He felt that changing SOPs to make them more focused on outcomes rather than process would have a positive effect on the                                             145 UN Official, Interview by Ana Lukatela, 2010. 203  implementation of gender mainstreaming.146  In part this must be understood through the prism of the structures from which the SOPs originate. It is important not to forget, as discussed in the theoretical framework, that standard operating procedures are also gendered given that they are a product of the structured power relations that otherwise perpetuate gender inequality. The standard operating procedures in organizations are often ―gender blind‖ and not designed in such a way as to address underlying factors perpetuating inequality and promote women‘s access to power (Arneil, 1999; Enloe, 2000; Mazur, 2002). The literature on organizational pathology also recognizes that standard operating procedures can lead to shortcomings (Barnett and Finnemore, 1998). In the case of gender mainstreaming these shortcomings are the gendered nature of SOPs – they were created by a structure that is itself unequal and they reflect that inequality by making actors blind to diversity, imposing one-fit solutions on all cases, or making individuals feel like they are not personally responsible for poor outcomes for women.  For example, one programme manager with UNDP in Kosovo, Jelena Bjelica, felt that the UN‘s methods of gender mainstreaming were patronizing and both intentionally or unintentionally excluded national expertise and local women‘s voices, which hampered the ability of mainstreaming to be effective.  As an example she discussed the UN‘s development planning methodology, which invented a new community building approach without asking local women what community forums they were already using and how. Ignoring national expertise                                             146 Alex Standish, Interview by Ana Lukatela, Pristina, May 7, 2010. 204  and understandings of gender relations (and even worse assuming there is no local knowledge or awareness of gender relations) the UN was predetermined to fail at gender mainstreaming. In addition, Bjelica, who describes herself as a ―political feminist‖, felt that gender mainstreaming within the UN was unsuccessful because implementation in programming could never happen until the inequalities between staff are addressed and staff that oppose gender equality are removed from programming functions. Ultimately the precursor to effective gender mainstreaming is the dismantling of gendered power structures within UN country teams themselves.147  The above cases show that where SOPs have been useful to champions of gender equality it has largely been because they have found creative ways of using them to influence a targeted population; those individuals that are not opposed to gender equality but whose disinterest, passive approach and inertia has in the past prevented them from implementing gender mainstreaming in their programme. As UNDP‘s Gender Advisor in Kosovo explained another gap is that donors have problematic SOPs with regard to funding requirements. The UN‘s major bilateral donors148 do not mainstream gender into their funding criteria or results indicators. If donors had a monitoring and reporting structure requiring the UN to report back on how their money had a different impact on men and women                                             147 Jelena Bjelica, Interview by Ana Lukatela, Pristina, May 11, 2010. 148 Swedish International Development Agency, Canadian International Development Agency, Norwegian Agency for Development, SwissAid, and others. Jocelyn Talbot, Interview by Ana Lukatela, Pristina, May 5, 2010. 205  then the UN would be required to change its behaviour. The lack of an SOP related to integration of gender in reporting is another reason for the poor translation from policy to practice in the process of gender mainstreaming development programmes. It relates to the accountability issue – programme managers are only interested in their results on issues against which their performance is assessed. Since gender equality is not one of the issues against which they are formally assessed then there is no reason for them to think about integrating gender in their reporting; and if they don‘t need to report on gender then they will not need to think about implementing gender mainstreaming.  The importance of SOPs was also identified at the broader political level. Estela Bulku, a nationally recruited UN official, felt that Albanian ―society lacks a consensus on gender equality.‖  On a study visit to Spain she found that the country was just as patriarchal as Albania. However partners in Spain explained that they had made progress on gender mainstreaming ―by building structures and procedures with accountability and making sure both political parties are on board.‖149 Also at the broader political level, UNFPA‘s country manager in Albania, Manuella Bello, described the lack of SOPs as the critical problem with getting the government to implement gender mainstreaming. Although a network of gender focal points exists on paper for each line ministry and, they are not implementing it. Says Bello:                                             149 Estela Bulku, Interview by Ana Lukatela, Tirana, March 9, 2010. 206  What they need is a set of SOPs with an organogram and a Terms of Reference for each ministry, including this gender focal point position, and then it would be a clear procedure to appoint someone.  They could not act like it was so difficult and confusing to get done or say that they want to do it but it is just too difficult.150 Several respondents felt that military organizations with their standard operating procedures were more successful in implementing gender mainstreaming. One respondent who remarked that personal attitudes were NOT relevant for policy implementation was the representative of the EU military force (EUFOR) in Bosnia, who felt that in military hierarchies, under the threat of constant sanctions, orders to implement a policy are followed implicitly.  She argues that the negative effects of a lack of internalized norms on the implementation of mainstreaming are only overcome in organizations such as police and military due to their SOPs. They have rigid hierarchical structures; accountability through standard operating procedures and command lines; inevitability of sanctions for non –compliance; and recognition is largely based on merit and results measured against official policies. This combination leaves little room for individual will in regards to implementation. The EUFOR representative described the military in the following manner: ―getting senior command on board is enough. People listen to their commands and there is no space for personal attitudes.‖ For example, when senior command ordered all officers to take the gender mainstreaming training there was 100% attendance.  No one left the room for a coffee or a phone call, and certainly no one excused themselves for                                             150 Maneulla Bello, Interview by Ana Lukatela, Tirana, March 5, 2010. 207  ―urgent reasons‖ and failed to return - situations described in several UN agencies. In international development organizations, which lack the SOPs military and police hierarchies have related to command structures and disciplinary action, organizational policy can be held hostage to personal attitudes. The ease with which hierarchical organizations such as the military and police take on gender mainstreaming was recognized also by Lubrani, UNDP Kosovo Resident Representative: once an order is officially issued there is full operationalization by subordinates, non-compliance is not an option, unlike in other international organizations.   In summary, the research revealed several clear instances where gender focal points or champions for gender equality felt they could use standard operating procedures to steer the behavior of their middle management colleagues, while a number of individuals also identified improved SOPs as the critical missing factor that they felt would assist them in promoting their agenda among those colleagues that could be susceptible to their influence.  Evasion of SOPs by Norm Resistors While SOPs can be effective for persuading those who are passive towards norms of gender equality, those who are clearly opposed to the discourse easily take steps to evade SOPs and avoid implementation. Many of the respondents pointed to the problem that individuals who do not believe in the goal of gender equality pose for the implementation of gender mainstreaming. These ―norm resistors‖ ignore the gender mainstreaming policies and strategies of their organizations and find ways to evade standard operating procedures. The 208  ease of evasion is striking: in a situation where there are no sanctions for lack of implementation, while delays in project delivery are highly discouraged, evaders have nothing to fear. One respondent described the situation thus: ―Evaders can get away with it because management does nothing and there is no punishment.‖151 Norm resistance was also related to gendered power structures within the local context. Norm resisters were overwhelmingly male staff, and more likely to be nationally than internationally recruited. As one female UN national staff official pointed out,  The amount of resistance in the country comes from the social norms. Women are still not seen as equal and that is fundamental to the whole story. On a society level that is true, however there are individuals that are enlightened, but overall it is a highly traditional society and this number of norm resisters creates problems for implementation.152  Different methods of norm evasion were identified. In cases where Programme Approval Committees (PACs) required gender to be considered in the programme documents the norm resistors would ignore the requirement and allow the PAC to cut and paste the identical vague paragraph about ―taking gender into account in the project implementation‖ into the context section of their forms and the programme document would be passed as ―sufficient‖ on the gender component.  As a UNDP official said ―gender mainstreaming doesn‘t                                             151 UN Official, Interview by Ana Lukatela, 2010. 152 UN Official, Interview by Ana Lukatela, 2010. 209  really just mean that you add a little paragraph in the introduction that you will take women into consideration.‖ She mused that in every country she had worked in during her career, the ―gender issues have always been difficult just because of the cultural attitudes. Gender is so divisive.‖153  This comment provides an interesting insight into what feminist political theory argues: women‘s rights in regards to family law, citizenship, reproduction and nationality are the battleground on which national identity is built in many societies (Enloe, 2000; Tickner, 2001). Supporting women‘s rights advocates to push back against these issues can elicit extremely emotional responses. Without understanding mainstreaming policy in light of this, implementation stalls because staff do not understand the tensions and divisions they are dealing with. One UN staff member recalled that during a posting in Yemen, when the issue of gender mainstreaming was brought up within the office in relation to operational issues and the rights of women national staff vs. male national staff, the dialogue became extremely divisive in particular between the national (both men and women) and the international staff. To avoid ―discomfort‖ the senior management decided to ―shelve the topic.‖154 Another strategy, which was described by all of the gender focal points interviewed across the countries and agencies, was the late submission of programme documents for comment prior to the PAC meeting. In cases where the gender focal points were instructed to review a draft programme document                                             153 UNDP Official, Interview by Ana Lukatela, 2010. 154 UNDP Official, Interview by Ana Lukatela, 2010. 210  the responsible programme manager would share the draft by email only the evening before (or even the morning of!) the PAC meeting and ask for quick turnaround, which made substantive input impossible. Several gender focal points agreed that more often than not this was done on purpose to prevent them ―from meddling with the project and changing it.‖155 One gender focal point recalled how she was constantly being sent proposals at the very last moment to review and ―pass‖ in terms of gender, but the drafters resented any change she would try to make other than a cut/paste insertion of a standard paragraph on ―gender considerations‖ into the context/background sections. Eventually she described getting tired of inserting this minimum paragraph so that the project drafters could move on in the PAC process and she rejected a late submission and explained to the drafters that if they were genuinely going to integrate gender it required a consultation between the programme manager and local women‘s NGOs in order to identify their needs and priorities and discuss how the project would address them. The programme document in question was never resubmitted to her and much to her chagrin she later heard it had been passed through the PAC process. She was unable to discover why and how the programme was passed through PAC and ultimately she admitted with resignation that ―there are just so many ways to                                             155 Virgjina Dumnica, Interview by Ana Lukatela, Pristina, May 4, 2010. Entela Lakho, Interview by Ana Lukatela, Tirana, March 10, 2010. Klelija Balta, Interview by Ana Lukatela, Pristina, June 15, 2010. 211  cheat around it [the PAC]‖ by those who put some effort in and that she found herself often fatigued with the whole process and her role.156 Another gender focal point identified specific programme managers in her agency whom she knew as problematic; they were not interested in gender equality, and they completely ignored her emails and requests for meetings. Ultimately she put it down to a ―personality problem‖ – she suspected they did not like a younger female staff member reviewing their documents, and it was clear that their attitudes towards gender equality were hostile.  In regards to the PAC process in UNDP Kosovo, several respondents agreed that senior management could play a role in accountability and sanctions vis a vis norm evaders. One respondent stated: ―Senior management will never say: ―This [the input from the gender focal point] is mandatory, full stop. I don‘t want to see any document without a gender component. The gender focal point is here to teach you, so send all documents to her." This respondent felt that if senior management put in place accountability systems – an important kind of SOP – then there might be progress. Instead the gender focal point described running around asking each cluster to share documents with her, something which they continually failed to do in any consistent manner, and she would find that ―all of a sudden there is another PAC and we have never seen this document before and it still passes.‖157                                             156 UNDP Official, Interview by Ana Lukatela, 2010. 157 UNDP Official, Interview by Ana Lukatela, Pristina, May 2010. 212  The reasons for these gaps can be identified in the pressures within the organization as well as the intentions of the individuals. As mentioned earlier, when bureaucratic politics come into play senior management will only support gender mainstreaming when there is no conflict with other goals such as mandate, funding, building relationships, etc. If a delivery rate to a donor is threatened by delays in passing a document through PAC and getting a project started and funds transferred, senior management is not likely to look lightly upon even further delays as gender mainstreaming is considered. One respondent from UNDP felt that while senior management may say mainstreaming is important, their only interest is the delivery rate in tandem with demonstrated results which leads to securing more funds.158 As long as gender equality is not considered a key ―result‖ by donors, as mentioned earlier, then there will be no motivation to ensuring the PAC process leads to effective incorporation of gender.  The programme managers were aware of the fact that delivery rates are prioritized above all else and that the lack of a gender perspective is not seen as a legitimate factor to stop the process of programme development. Given this, it is not unusual that they do not fear any sanctions for not soliciting and including the gender focal point‘s inputs into their planning process. Consequently they ignore phone calls from the gender focal point, they send the document for comment on the morning of the PAC meeting and ask for immediate turnaround, (which is almost always impossible given other work commitments) and they                                             158 UNDP Official, Interview by Ana Lukatela, 2010. 213  never want more than a standard paragraph on gender in the introduction or context sections of the project document.  This problem is compounded by the fact that gender focal points cannot ―fail‖ a project at the PAC for not having passed through their hands first, rather their role is only to ―advise.‖ Without accountability mechanisms put in place the gender focal points expressed frustration that nothing will change.  One source from within UNDP believed that if the right structures and procedures were put in place with real sanctions then those that cared about gender mainstreaming could try to have more power over those that normatively resist.159 A tactic that was common to evaders across the case studies was the refusal to communicate.  Programme managers that were against the mandate of one gender focal point simply ignored her and refused to engage: ―They never respond and say no [she explained]; they just do not respond.‖ After complaining to senior management she concluded there was no point to escalating the situation as the programme managers seemed ―untouchable‖ on the gender issue since they were only judged on their programme delivery rates.160 Another gender focal point described how a particular male national staff member who did not take kindly to her interventions to mainstream gender into his justice programme used the strategy of not answering or returning her calls or emails. So even though there was a procedure in place by which he was supposed to incorporate her inputs he evaded it by making communication                                             159 Entela Lakho, Interview by Ana Lukatela, Tirana, March 10, 2010. 160 UNDP Official, Interview by Ana Lukatela, 2010. 214  difficult or impossible. According to the gender focal point the ultimate problem was that at a fundamental level he just did not believe in gender equality.161  Another gender focal point identified this same problem and explained that it was easy for managers to avoid input procedures by making communication difficult. When it came to collaborating on their project documents and providing her input to the programme manager, the gender focal point observed: ―if they don‘t want to tell you, they just won‘t tell you.‖ In another situation if a group meeting was scheduled for all programme managers with the gender focal point one programme manager in particular would come and then leave five minutes into the meeting due to an ―urgent issue‖ and never return. The gender focal point had complained to management; however no steps or sanctions were taken against the individual in question. She felt this sent the message that evasion could occur with impunity.162   UNIFEM staff in Kosovo appreciated the efforts of the UNDP gender advisor but speculated that her impact would be limited due to the level of normative resistance she found herself facing. One individual commented:  She [the gender advisor] is having a hard time; there is lots of resistance from program managers. She has gone into each cluster to analyze the programs but if they don‘t believe it they won‘t implement it the way she tells them to. Her problem is the programme managers.163                                             161 Virgjina Dumnica, Interview by Ana Lukatela, Pristina, May 4, 2010. 162 UNDP Official, Interview by Ana Lukatela, 2010. 163 Flora Macula, Interview by Ana Lukatela, Pristina, June 14, 2010. 215  The pilot gender marker project is one concept of a method to monitor the inclusion of gender in programmes. All the programme heads were aware of the initiative, but had difficulty understanding its purpose or how it would be applied since they did not understand what mainstreaming meant, let alone how it would be assessed within their programme document. As one programme manager in Bosnia said:  This is an attempt by headquarters to do institutionalization of gender norms and markers etc., and yet those that are not interested can still easily work around the system. The induction training for programme heads has little to no gender issues in it and so from the beginning those that are not on board feel like they can avoid the issue. Peter [the Deputy Resident Representative] insists on each project trying to get to Klelija [the gender focal point] but there are ways to avoid him and when he leaves what then? Not even all the donors insist on gender aspects to your project so how can we mainstream if we are not required by donors to do it?164  He concludes that really he sees little point in ―internal mainstreaming‖ (meaning operational mainstreaming) as no one has explained it to them and it is not clear to him what rules would need to be changed. The Programme Manager of UNDP's resettlement programme in Kosovo noted that putting the gender marker into UNDP‘s financial tracking system does not actually change the behavior of staff. He saw it simply as one further step of the ―email activism‖ that the gender advisor was doing, which has actually led to                                             164 Jasmin Porobic, Interview by Ana Lukatela, Sarajevo, June 15, 2010. 216  an increase in evasion since individuals have become more defensive and hostile to what they perceive as attacks on their work. He felt the work of the gender advisor needed to be much more strategic and targeted. The expectation that one gender advisor would be able to mainstream all programmes was clearly unrealistic; mainstreaming cannot be done without sector-specific expertise and consequently programme managers, who are thematic experts in their area, resent what they see as her ―layman interventions‖ in their work. This problem of gender experts was echoed by UNIFEM officials who noted that the greatest difficulty was making other UN agencies understand that they had to stop hiring so-called ―gender experts‖. Although these individuals were women‘s rights activists and feminists they had little of the area-specific or thematic expertise that was required to command authority and legitimacy as an expert. This is similar to the concern that advocates had when a Swedish politician was hired as the SRSG on Sexual Violence in Conflict instead of an individual with military expertise, as mentioned in an earlier chapter. UNIFEM was trying to encourage other agencies to hire technical experts who are also sympathetic to gender equality instead of broad ―gender experts‖. As an example, for gender responsive and participatory budgeting UNIFEM was pushing other agencies to hire economists and finance specialists who were sympathetic to women‘s empowerment. As one UNIFEM official explained: ―You need a gender expert who can speak the language of the sector you are trying to 217  mainstream and instead we keep sending ‗gender mainstreaming experts‘, which DON‘T exist.‖165 Some of the most blatant evasion of gender mainstreaming procedures occurred in the EU Rule of Law civilian mission to Kosovo (EULEX). Although a gender equality and human rights unit was included in the structure of EULEX, it was in fact added as an afterthought by policy planners in Brussels, when someone pointed out their gap.166 However once on the ground the unit was never fully staffed by the senior mission leadership and the Head of Unit (when there was one, as in fact the position was often left empty) found herself having to constantly defend her mandate and the Unit‘s existence to the mission leadership. Successive mission leaders displayed openly hostile attitudes towards the Unit‘s mandate, which included both gender mainstreaming and dealing with internal harassment claims. The Head of the Unit felt that senior mission leaders and most of her EULEX colleagues did not ―believe‖ gender-based discrimination was a problem in the mission despite the fact that the professional working environment within EULEX was described by both the Unit Head and all the UN agency gender focal points as appalling. Within the hierarchy of EULEX the Head of Unit was ranked high enough to be included in the senior policy committee, however she was never invited to their meetings and they resisted her constant efforts to inquire into the policy process and participate. When she would ask to be included in senior                                             165 UNIFEM Official, Interview by Ana Lukatela, 2010. 166 Sirpa Rautio, Interview by Ana Lukatela, Pristina, May 13, 2010. 218  management meetings the mission head would never directly deny her claim, but his office would schedule the meetings and not inform her. When the senior management became tired of her persistent inquiries they punished her by making recruitment for her unit interminably slow –new staff members to her Unit took more than two months to arrive in mission after the predecessor had already left, which was not the case with other units such as police or justice. A unit consisting of eight individuals on paper never actually had more than four on the ground at any given moment, a constraint that severely limited her ability to fulfill the Unit‘s mandate and thus gave her an image of being irrelevant to the organization‘s work. Policymakers in Brussels displayed little interest in the concerns of the Head of Unit, which she linked to their lack of interest in topics that were not part of their performance and accountability measurements. Another strategy for evading the SOPs was described by UNICEF‘s gender focal point in Kosovo. She described how several staff members hostile to gender equality norms undermined the process of ―standard gender analyses‖ by making their analysis all about men‘s situation and ignoring the situation of women. So for example overemphasizing men as victims of domestic violence or discussing domestic violence only from a perspective of psychological violence by wives against husbands while ignoring physical violence, which is much more likely to be perpetrated against women. She believed they did this to show their spite for the process and because they felt it was ―unfair‖ to men. The experience left her believing that someone who is truly hostile to the notion of gender mainstreaming due to their own normative commitments cannot be convinced 219  otherwise. In her experience of working with UN, government and civil society, the success of gender mainstreaming always has depended on the individual she has to work with and their openness to changing social norms on gender.167 Taken as a whole, norm resistors and SOP evaders are clearly a major obstacle to policy implementation, particularly since senior management had little interest in the dynamics playing out at the working level beyond programme funding delivery rates, which in turn is what they as senior managers have to report further up the hierarchy. Even in a case such as UNDP in Kosovo, where senior management was committed to gender there were still no sanctions for those that avoided their mainstreaming responsibility, likely due to organizational constraints on the ability of senior management to enforce accountability as well as no clear guidelines on accountability processes. This led one UNDP gender champion to conclude that "UNDP is not capable of delivering on gender until we sort out our house internally."168 A civil society activist in Kosovo also identified the limitations that UN staff attitudes place on implementing gender mainstreaming. Even though the Resident Coordinator and senior gender focal point are committed to gender equality, they are not ―magicians‖ as this respondent described, and ―they cannot change the attitudes of people. She [Osnat Lubrani] can change policy on paper, but not the actual work being done.‖169                                              167 UNICEF Official, Interview by Ana Lukatela, 2010. 168 UNDP Official, Interview by Ana Lukatela, 2010. 169 Luljeta Vuniqi, Interview by Ana Lukatela, Pristina, May 12, 2010. 220  Another individual speaking anonymously echoed these thoughts but also felt that even more difficult than trying to gender mainstream programming was trying to gender mainstream operations. The level of normative resistance in the operations staff (who are always hired locally and are thus always national staff) was even higher than in programming staff: "Gender mainstreaming makes people emotional, we need to call a spade a spade and handle this; we need a better sense of UNDP culture.‖170 Programme managers at least have some nominal interest in development work, whereas operations staff are hired for their accounting, finance, logistical or organizational skills, and their values and norms are not considered relevant to the hiring process. Staffing Issues As one senior UN official stated in the quote at the beginning of this chapter, the chaos of the UN‘s ―house‖ clearly indicates that there is a fundamental problem and inequality with how the UN recruits, retains and promotes its staff.  In all three countries both national and international staff at varying levels commented that ―people get jobs in strange ways‖171 or that the hiring structure was disadvantageous for women candidates.  One programme manager in UNDP, Jelena Bjelica, felt that gender mainstreaming within the UN was unsuccessful because implementation in programming could never happen until the inequalities between staff are                                             170 UNDP Official, Interview by Ana Lukatela, 2010. 171 UNDP Official, Interview by Ana Lukatela, Pristina, May 2010. Anonymous, Interview by Ana Lukatela, 2010. UN Official, Interview by Ana Lukatela, Tirana, 2010. 221  addressed and staff that oppose gender equality are removed from programming functions. Ultimately she felt the precursor to effective gender mainstreaming is the dismantling of gendered staffing structures within UN country teams themselves.172  UNDP, a large organization in terms of both financial and human resources, was perceived by numerous respondents as having opaque recruitment policies and not sufficiently focused on ensuring normative commitment to the principles underlying the UN‘s development work (such as gender equality). In all three case countries concerns and problems were expressed anonymously in relation to UNDP‘s recruitment: in Albania several respondents anonymously described the process as opaque and with no criteria of commitment to normative principles of diversity and equality, in Bosnia the Deputy Resident Representative himself described the difficulties in hiring candidates that espouse the appropriate social norms on gender equality if they have other strong qualifications. One international programme manager with UNDP gave a clear gender analysis of past hiring problems within UNDP Kosovo. He was sympathetic to the difficulties and discrimination faced by national women staff members in terms of recruitment, retainment and advancement.  When he brought this to the attention of the senior management the response he received was that there was a lack of ―qualified‖ internal female candidates for promotion.173 The human resources                                             172 Jelena Bjelica, Interview by Ana Lukatela, Pristina, May 11, 2010. 173 Alex Standish, Interview by Ana Lukatela, Pristina, May 7 2010. 222  department explained that advancement required an education level at the equivalent of a Master‘s degree, which local women UN staff predominantly did not hold. In response an opportunity was created for local UN staff to attend a fast track Master‘s programme at a university in England with cost-sharing offered to defray the tuition. However once the opportunity was advertised only male national staff members applied. When no women applied to attend the eight month Master‘s programme the senior management interpreted it as lack of women‘s interest in career advancement. National women staff however bitterly complained that when designing the programme, UNDP senior management had not asked them if they would be able to take advantage of such an opportunity considering that it meant leaving children, families and husbands for eight months to reside in the UK. Traditional attitudes in the local culture still made it difficult for women to leave their families to go abroad, especially for a period as long as eight months, and few husbands would be prepared to care for children on their own for so long. The resulting effect was that other women were burdened and expected to care for children for eight months while husbands took advantage of the opportunity.  Poorly conceived initiatives such as this one have not helped UNDP invest in building a cadre of women mid-level management or gender sensitive programme managers. Instead it has just perpetuated the unbalanced gender power relations within the organization.  As another example, staff from UNDP and UNIFEM complained that the standard scheme of issuing short-term contracts meant that staff was constantly 223  being ―recycled‖ as opposed to ―invested‖ in. Short term contracted staff are not eligible for professional development opportunities as are fixed term staff. However they are hired over and over on rolling contracts, even though they have no positive norms on gender equality and are never trained on gender mainstreaming. Staff felt there was no real commitment on the part of the agency to staff development, learning, evaluation or career planning.  In particular several respondents drew attention to the obstacles that national female project managers faced when competing for promotion. The ―inflation of job requirements‖ meant that the job description for a driver in UNDP might actually require a Bachelor‘s degree and all other professional staff a Masters. Respondents argued that this ―inflation‖ discriminates against women ―because they cannot take time out from family commitments to get the MA needed for promotion so immediately they have to make choice between family and developing their career.‖174  The repercussions of this discrimination go beyond employment with the UN and have an effect on employment prospects in the future: ―If we leave out women from promotion in UN service, they are out of jobs in the civil service. They will not get the jobs in the civil service once the UN leaves.‖175 Instead it is likely that these types of well paid secure jobs will all go to men who have the seniority and connections they gained from their UN service.  Beyond increasing the number of women in the organization and providing them with skills training to make them more competitive candidates, the structural                                             174 Alex Standish, Interview by Ana Lukatela, Pristina, May 7 2010. 175 Alex Standish, Interview by Ana Lukatela, Pristina, May 7 2010.  224  barriers of the organization and indirect discrimination also must be addressed. Affirmative action such as quotas provides equal opportunities; however the different gender identities, masculine vs. feminine, and their corollary attributes must be valued equally within the organization. In other words attributes such as listening, collaborating and building relationships, which are often seen as feminine, are perceived as less valuable in a manager in comparison to being competitive and aggressive.176 Without an understanding of how social structures and identities are gendered within international development organizations no analysis is complete.177 Conclusion The findings of this dissertation suggest that the gender mainstreaming agenda in UN development agencies has over-estimated the positional and agenda setting power of senior managers, relative to the types of power exercised by middle management: hidden power, power of dialogue, and power of conflict (Rao et al, 1999). The Secretary General has asserted in multiple arenas (SG‘s Report on Mainstreaming, 2011) the importance of senior management using its power to set an agenda of gender mainstreaming, however the focus on senior management, while relevant as this research affirms, has come at the expense of  understanding how middle management influences gender mainstreaming and concurrently hinders progress. As this study has demonstrated, programme managers control communication and                                             176 Rao et al conducted interviews with staff in development organizations to map the attributes that were perceived as the most valuable for an organization. ―Masculine‖ attributes were seen positively, while ―feminine‖ attributes were seen as ineffective. Rao et al, 1999 177 Adapted from Kolb and Meyerson – Four frames approach: 1. Add women; 2. Equal opportunities; 3. Masculine vs feminine;  4. Gender identity as an organizing category of society. 225  reporting, often lack gender consciousness or belief in gender equality as a core value and may be willing to build alliances and even enter into open conflicts to avoid implementing gender mainstreaming if it runs counter to their norms. Norms are a necessary, but not always sufficient factor; they must be paired with organizational behavior protocols that can support implementation among those who are passive and there must be ways to sanction programme managers who are in opposition to gender equality. It is in these cases that we then see instances of policy implementation.  The final chapter discusses how international organizations can address these gaps in implementation by developing hiring policies that place more weight on judging the degree to which a candidate has internalized the norms an organization embodies in its policies, such as the norm of gender equality. SOPs can then be used to ensure that recruitment processes take such factors into account. Given the evidence of the important role that middle management in international development organizations play in policy implementation (relative to senior management expressions of intent, standard operation procedures, and bureaucratic politics for policy adoption), the final chapter will examine what this means for the way forward and how organizations such as the UN can go about changing the way that they implement crosscutting policies such as gender mainstreaming.   226   6. Implications, Recommendations and Conclusion The persistence of gender mainstreaming as a strategy in international organizations is due to the inertia of the world polity and the values embedded in the world culture, which is perpetuated by and through senior international development professionals. This has resulted in a situation where despite an evolving academic debate on gender mainstreaming and its alternatives there has been very little change in the conceptualization and implementation of the approach at the practitioner level. This chapter will apply the lessons from the theoretical framework and the research findings to other policy processes in development organizations and will also discuss policy recommendations regarding the future of gender mainstreaming and UN agency reform. The implications of the findings are relevant not only for gender mainstreaming but for any type of crosscutting policy process in international development organizations. These mainstreaming strategies rely on the same mode of implementation – individuals with sector specific expertise and a commitment to the issue being mainstreamed. Examples mentioned earlier in the theoretical framework include HIV/AIDS mainstreaming, environment mainstreaming, disabilities mainstreaming or poverty mainstreaming.  Crosscutting and mainstreaming policies have proliferated in international development. The types of issues now referred to in the context of mainstreaming vary widely. While some such as gender and HIV/AIDS 227  mainstreaming were adopted over 10 years ago, others have only now started to enter the development discourse, as the author has observed: good governance mainstreaming, disability mainstreaming, child right‘s mainstreaming, sustainable development mainstreaming, resilience mainstreaming, etc. Often these policies are not fully defined or articulated. For example, UNDP and UNICEF have recently begun to discuss ―resilience‖ mainstreaming; a non-paper developed by the two agencies and obtained by the author outlines how UN development efforts should mainstream ―building resilience‖ within communities to respond to external shocks such as crisis, conflict or natural disaster. In this example, there is no clear definition of what resilience means in the context of a mainstreaming policy. Some informal explanations offered by stakeholders from the two agencies mention the importance of ―traditional coping mechanisms‖ within communities. However for gender equality advocates any discussion of a return to "traditions" is often concerning in contexts where traditions are conservative and typically discriminatory for women. This could possibly lead to a situation where the UN is contradicting itself in trying to promote both resilience and gender equality.  In addition to the problem of creating contradictions at the policy level, the proliferation of mainstreaming has also led to competition for the attention of programme managers. Given the results of this research, which argue that normative commitment is a necessary condition for implementation of mainstreaming policies, the expectation that programme managers will (or are able to) simultaneously mainstream a number of different issues into their 228  thematic topic could prove problematic. There is also no guidance on how programme managers are expected to prioritize when presented with a multiplicity of official mainstreaming policies, especially when working in a climate of limited time and resources. This might be particularly disadvantageous for gender mainstreaming if gender equality is less likely to be valued by programme managers as compared to the norms that underpin other mainstreaming agendas. Not only does the underpinning norm put gender mainstreaming at a disadvantage compared to the other mainstreaming agendas it is competing against, but it also has a more complex nature in that it is more than just integrating participation; which is how disabilities mainstreaming or HIV/AIDS mainstreaming seem to be perceived. The transformative aspects of gender mainstreaming require much more thought and effort be put into programming interventions than a purely participatory approach.  Authors such as Hankivsky (2008; 2013) argue that there is the potential for a different approach to mainstreaming that would take away this zero sum game in the eyes of programme managers. If mainstreaming were understood in terms of the intersectionality of diversities and inequalities then these issues could be strategically linked in one approach instead of competing as single issues. This could be a useful way of communicating mainstreaming to those programme managers who are already sensitive to inequality and diversity issues.  229  Theoretical Implications for Crosscutting Policy Processes in International Development Organizations Two general theoretical implications for mainstreaming policies in international development organizations are outlined here and revolve around the importance of context and actors. First, the theoretical implications of a non-linear policy process and the different ways that context can affect adoption and implementation; and second, the implications of understanding which actors hold the balance of power when it comes to implementation. Understanding these implications is critical to understanding why mainstreaming policies have failed in the past and what can be done to improve their general implementation.  Understand the Policy Process is Not Linear and Context Has Diverse Impacts  The findings of this study have made it clear that when it comes to crosscutting policies the process is far from linear; the local policy context does not always influence policy adoption, and adoption does not automatically lead to policy implementation. Instead different causal mechanisms affect adoption and implementation and without a holistic theory that takes into account both stages of the policy process theoretical explanations will remain partial and incomplete. In order for crosscutting issues to be formalized in policy and then implemented it is necessary to understand whether they are embedded in world values, what is the local context and values related to the issue and who would be responsible for implementation. For example it is likely that the concept of resilience mainstreaming or sustainable development mainstreaming will not be adopted as widely as gender mainstreaming given that the concepts of resilience or sustainable development seem to have no foothold currently in the world polity 230  and world values through which it would begin its dissemination into policy. However an issue such as child‘s rights mainstreaming is more likely to disseminate and be implemented given that it has justification in both the global context of the world polity as well as norms in local contexts where implementation would occur.   While the policy context at the global level is more important for policy adoption, the policy context at the local level is what influences implementation. In order to understand mainstreaming it is critical to have a thorough understanding of the local context that the mainstreaming policy is targeting. The focus of other organization theories has been on the internal context and culture of IOs, however the findings of this research show that the national staff of an IO cannot be separated from the local context that produced them. This local context and the values internalized will have a larger effect on implementation than the global context and world values that are influencing policy making and adoption processes at a higher level. This means, for example, that environmental mainstreaming might be easier to implement in an island state whose citizens readily understand the threats of climate change and flooding than in a developing state in a mountainous region that has little perception of how they are affected by climate change.  The different effects of global and local contexts also have theoretical implications for constructivism in terms of the specific kind of norm socialization that is happening in the elites who are members of the world polity. The evidence suggests that while the elite members that make up the polity may be socialized 231  to internalize and believe in certain norms, that does not appear to translate into action. The focus on form and process among senior managers may be a difference in the type of culture and norms that the world polity produces and perpetuates, as well as the nature of actors (elites) that are involved in the process. This stands in comparison to the type of culture produced through the norm cycle, the nature of actors at the middle management level (Risse, Ropp and Sikkink, 1999), and Panke's "deep impact of norm socialization" (Panke, 2012:722).  RRK recognize this to the extent that they acknowledge the state is not a black box and that once ideas become norms there still needs to be a better understanding of how these norms influence individual behaviour. For example, in the case of human rights norms what brings actors such as a member of the military to stop ordering extrajudicial killings (Risse, Ropp and Sikkink, 1999). This question of individual belief, and form over action, is an interesting area for further theoretical research as it has significant implications for expectations of how global norm-based policies will be implemented; an issue in which both world polity theorists and constructivists are heavily invested. In April 2011 the UNDP Administrator launched an ―Agenda for Organizational Change‖, which had the goal of improving the effectiveness and impact of the organization and transforming UNDP into ―an intellectual leader on global development issues… able to translate policy to solutions in different country circumstances― (UNDP Agenda, 2011: 4). The Agenda recognizes that taking a global policy and applying it to a local context is not always straightforward, hence the need for decentralization of the organization that 232  would empower staff to take more decisions at the country level and to decentralize operations in particular. However the theoretical implications of this research tell us that decentralizing a global policy such as gender mainstreaming where it does not have a foundation in normative values in the local context will be ineffective. Understand the Importance of Diverse Actors In the past scholars of policy processes in IOs have focused their attention on the influence of senior management and member states on policy adoption and policy implementation. The findings of this study indicate that a different profile controls the implementation phase of crosscutting policies– middle management – and that without understanding their values and norms mainstreaming policies cannot be successful. This is a critical contribution of the theory proposed by this dissertation; by assuming a linear policy process organizational theory literature has overstated the importance of senior managers in relation to middle managers. And while some scholars have already pointed out the power in the hands of middle management in organizations this theory has put it in the unique context of a non-linear policy process in order to accord this finding its true weight in relation to policy implementation in particular. The other reason this is particularly significant in the case of mainstreaming policy is the relative importance of senior management vis a vis middle management. For a thematic or issue-based policy, there can be no implementation without first a policy being adopted, which then still may or may not be implemented depending on dynamics within the organization. However 233  with mainstreaming policies, individual programme managers and practitioners can mainstream any issue they are interested in into their programme regardless of whether there is a formal policy on the books or not. All it takes is the personal interest of the programme manager, whether the issue being mainstreamed is gender, people with disabilities, HIV/AIDS awareness etc. The main focus of the programme does not change, all that changes is the personal approach the programme manager applies to implementation. So in the case of non-linear mainstreaming policy processes senior managers are not veto players and while they may be a necessary condition for systematic mainstreaming to happen they are far from sufficient.  The UN‘s 2011 report on gender mainstreaming only reinforces the typical approach that responsibility for the lack of implementation lies on ―senior managers‖ and their lack of commitment and promotion of the institutional mandate to mainstream gender (SG Report on mainstreaming, 2011: 9). The responsibility and role of middle implementing managers continues to be a blind spot for both practitioners and academic scholars. Mainstreaming approaches overall must take this into account if they want to achieve results. Policy Recommendations for Gender Mainstreaming The choice of case studies in this dissertation was based on variation in the strength of the UNIFEM/UN Women presence. The anticipation was that a strong gender equality agency, as a source of technical expertise and advocacy, increases systematic UNCT efforts on gender mainstreaming. However the findings show that the strength of UNIFEM/UN Women is not a significant factor. 234  This has important policy recommendations as the UN system has largely functioned on the assumption that the creation of UN Women is enough to see increased results on gender mainstreaming policies. However this research has shown that other important factors need to be considered in decision-making, and that while UN Women is important for targeted actions on gender equality and women's empowerment its effect on gender mainstreaming is negligible.    If the normative attitudes of programme managers are a considerable factor in implementing mainstreaming policies then this raises the question of whether adopting a formal mainstreaming policy has added value. In particular in the case of gender mainstreaming where there has been agreed policy failure, the question arises of whether the approach continues to have a purpose. The findings of this study on the catalytic role that SOPs can play in building accountability for mainstreaming implementation are a strong argument for the continued adoption of mainstreaming policies, in particular gender mainstreaming. Without a policy on paper, norm champions will mainstream gender in their programmes anyway, however a policy on paper coupled with strategic SOPs can be used to also ensure passive norm holders are also moved to implementation.  This section identifies and discusses four main policy recommendations for gender mainstreaming: changing recruitment policies to reflect what the lessons of the research demonstrates on normative commitments by programme managers; building accountability structures and SOPs that can support mainstreaming; returning the focus to mainstreaming gender internally in 235  organizations and not just in their externally focused programming (as a strategy to return gender mainstreaming to its socially transformative characteristics and differentiate it from ―women in development‖ approaches); reconnect gender mainstreaming with human rights discourse to return the understanding that the purpose of gender mainstreaming is gender equality and women‘s empowerment which are inextricable from fundamental human rights – a value much more likely to be widely understood and internalized.  Recruitment in International Development Organizations The UN‘s 2011 Report on gender mainstreaming, discussed in Chapter Two, reached the conclusion that one of the main problems leading to lack of implementation and policy failure was the gap in staff ―expertise‖ on gender mainstreaming. This led the report to recommend that more training is needed. However the research and the theory proposed in this dissertation argues that technical training will not change the underlying reason for failure which is the large number of programme managers that normatively oppose gender equality. Programme managers are not interested in increasing their mainstreaming expertise because essentially they are opposed to the goal behind mainstreaming. The issue of normative allegiance to the issue being mainstreamed is not discussed anywhere, as it is understandably sensitive.  The report mentions perfunctorily in one section that ―incentives for good performance‖ (SG Report on mainstreaming, 18) might prove useful; however the idea is not followed up on at any point later in the document. The mention of ―incentives‖ is important because it is a latent indication that the issue of staff 236  motivation needs to be rethought. Staff clearly lack motivation, however if the problem with motivation is actually normative opposition then the types of incentives that might be effective in eliciting good behavior are less straightforward. An assumption that financial incentives or employment tenure sanctions can be effective in overcoming deeply ingrained normative opposition may be dangerously incorrect.  Recruitment, retainment and advancement policies are at the crux of the implementation issue. What is clear from this study is that mainstreaming will not be implemented without accompanying normative commitment. In this respect the UN needs to come to terms with how it recruits its programme managers and the impact this process has on mainstreaming policies. The UN‘s development work is underpinned by its commitment to advancing peace and security, human rights, development, humanitarian affairs and international law. The normative principle of promoting women‘s empowerment and gender equality has become an integral aspect of all five of these areas since the late 1990s when UN agencies first started developing gender equality strategies. These normative commitments, and the inter-governmental framework they are grounded in, are what give the UN its comparative advantage over other international development agencies. UNCTs hold considerable convening power, impartiality and experience which enable them to be able to promote a normative agenda and not only an instrumental development agenda, which is a characteristic that other regional and international development organizations cannot always claim. 237  In order for mainstreaming policies to have an impact the UN must stop relying on generalized ―mainstreaming experts‖ as the solution and must develop recruitment procedures that attempt to highlight an individual‘s normative commitments. Recruitment of programme managers should focus on finding sector specific experts who are sympathetic to the issue that they will be required to mainstream whether it is gender or HIV/AIDS. In 2011 the Secretary-General commissioned a report from an independent expert advisory body on the UN‘s recruitment of civilian staff in post-conflict (Civilian Capacity Report, 2011). In his response to this independent report, the Secretary General agreed that there were serious concerns with the continued failure of the gender equality agenda and stated that the early engagement of thematic gender experts is critical (SG Report on Civilian Capacities, 2011). The findings of this dissertation support the need for thematic experts who are also gender sensitive. A number of officials expressed their frustration with the generalist ―gender experts‖ who have created a profession for themselves at the expense of thematic experts who are also aware of gender inequality issues and are gender sensitive and would be happy to integrate such issues in their work: The biggest problem with gender mainstreaming is we keep hiring activists to help us do it, but we should be hiring technical [thematic] experts who speak the language of government operations to help us, people that know how to use the way the government functions. This would be real mainstreaming then. One way to deal with this lack of expertise is more joint programs, so that 238  we can draw on expertise of other agencies on technical areas like environment or economics, then we help them gender mainstream. If we could institute strong procedures of joint gender work then we could do this very successfully.178  Given its underpinning normative commitments the UN cannot continue to hire programme managers only for their thematic expertise nor gender experts only for their gender expertise; candidates must be thematic experts who fully ascribe to the fundamental normative commitments the UN promotes. Just as the UN would not hire a programme manager who believed that torture is not a violation of human rights, so they should not hire a programme manager who believes that women should not have an equal voice to men in society. UN reform has been an ongoing discussion for many years; however the first comprehensive report on agency reform was only published in 2006 (Report of the High Level Panel on System Wide Coherence). Since then, reviews of agencies, programmes, funds, management and operations have occurred almost every year in varying forms.  Even UN reports on other topics have touched on recruitment reform. For example the so-called ―Brahimi Report‖ (Report of the Panel on United Nations Peace Operations, 2000) highlighted the difficulties associated with recruitment of appropriate staff and one of its key recommendations was: ―The Secretariat should also address, on an urgent basis,                                             178 UN Official, Interview by Ana Lukatela, 2010. Joint programs refers to a jointly implemented program by multiple UN agencies that both have expertise in the given area under consideration. For example an urban planning joint programme might include both UN Habitat and UNDP who have thematic expertise in housing and local governance respectively, while UN Women could provide the expertise to mainstream gender into housing and local governance issues.  239  the need to put in place a transparent and decentralized recruitment mechanism for civilian field personnel‖ (2000: xii). The Civilian Capacities report recognizes that the UN fails to ―recognize, recruit, reward and retain talent.‖ The primary determinant of hiring is number of years in UN service, while quality, relevance of knowledge, and experience are devalued. The current human resources system also ―struggles to cope with under-performance of staff... In the current fiscal climate, where austerity is the order of the day, the United Nations must be accountable to Member States and discontinue the employment of non-performing staff‖ (Civilian Capacity Report, 2011: 14). During its research the senior advisory group reported this to be one of the main concerns expressed by senior UN leadership. The findings of this dissertation support this concern with hiring – numerous sources claimed that results were measured only against budget delivery rates and not in terms of substantive performance.  This aligns with the conclusions of the Civilian Capacities Report, which found that staff is not held sufficiently accountable for their core functions let alone their mainstreaming responsibilities. To make matters worse non-performing staff face no sanctions and recruitment procedures are not effective at identifying the ability or skill of thematic experts to mainstream. The report on civilian capacities makes general recommendations for the improvement of hiring and retention of staff, including formalizing value of experience outside of the UN. However what the Civilian Capacities report does not discuss is a method to formalize not only the process of identifying expertise 240  but also a process of identifying normative commitments, which this dissertation argues is the real crux of the matter.  The current UN human resources recruitment and retention process is based on ―behavioral competency‖ questions, which are meant to measure a candidate‘s ability to associate behavior with outcomes (related to cooperation, leadership, conflict management, project management etc.), and less on substantive expertise questions. An example of a typical question a panel would pose to a candidate is ―Tell us of a time that you managed a conflict within your programme team.‖ Answers are then scored based on the number of positive values and skills mentioned by the candidate against a list the committee associates with each behavior.  This approach could also be used more effectively to identify whether a candidate is sensitive to basic manifestations of gender inequality in their surroundings. For example questions such as ―Tell us of the types of gender inequalities you observed in one of the previous programmes you managed? This would reveal normative commitment in so far as we would expect a candidate that is not normatively committed to gender inequality would have a difficult time elaborating on its manifestations. Other special measures that could target the right profile of programme staff includes a variety of administrative instruments, policies and practices, such as outreach or support programs; allocation of resources; preferential treatment; targeted recruitment, hiring and promotion; midterm and long term targets and quota systems. UN Member States have agreed already that using special 241  temporary measures to address structural inequalities and barriers is not discrimination (CEDAW Committee General Recommendation No. 25 1999). While the UN needs to renew its focus on normative commitments in recruitment processes (as discussed earlier in the section on staffing issues), it also needs to reconsider how it deals with the training of existing staff on gender mainstreaming within both operations and programmes. Managers pointed to training as the main vehicle for imparting technical knowledge on gender to programme staff, however trainings did not target norms and values with the goal of socialization and many respondents discussed the problem of ―training fatigue‖ or the concern that trainings imposed on unwilling staff could potentially do more harm than good by cementing opposition and hostility.179   The difficulty in providing organizational training to adults on a normative issue such as gender equality has been recognized by authors such as Rao (1999) and Tiessen (2005). The question of how to create change in attitudes and behaviors despite these difficulties is extremely important. Theories of adult learning, or "andragogy" stress that the learning process among adults differs greatly from children and youth.  Adults come with their own internalized motivations and norms, as well as experiences and knowledge, and resist information when they feel values are being imposed on them. In addition they require knowledge and skills being imparted to be relevant and practical (Lieb,1991; Fidishun 2000; Knowles, 1984). The approach to adult learning is                                             179 UN Official, Interview by Ana Lukatela, 2010. UNICEF Official, Interview by Ana Lukatela, Pristina, May 2010. 242  thus very much one of  ―action learning‖, or ―learning by doing‖, which involves collaborative problem-solving. Unfortunately the general training approach among UN agencies has been a passive training session where participants listen to a lecture on gender equality. Action learning, on the other hand, might mean participants go to project sites and observe firsthand how the experiences of men and women beneficiaries differ or situations where men and women face different challenges or have different needs. Action learning means experiencing gendered social structures in the specific context the staff member is working in and talking with beneficiaries – both men and women – to understand their views and solutions. Staff must be allowed to witness the benefit for individuals – men and women, children, families - of a project component that has been gender mainstreamed, and based on the experience internalize the change in their attitudes and values. This requires a significant move away from the top-down approach to change that UN agencies have adopted thus far, towards a strategy of transformational and active ‗ways of working‘.  In cases where deep-seated attitudes and behavior cannot be changed, eliminating non-performing staff is essential. Just as a UN agency would not condone a staff member who displayed attitudes contrary to universal human rights, so it cannot condone one who displays attitudes against women‘s rights and gender equality.  In addition, an important aspect of action learning must deal with fears of male staff that gender equality is a zero sum game in which power is taken away from men and given to women, or that gender equality is about ―changing culture.‖ The focus of gender mainstreaming initiatives should 243  be on changing discriminatory structures, practices, attitudes and behaviours, not on taking power ―away from men‖.  The unequal distribution of power in organizations is relevant not only to gender but also to differences of race or culture. This issue was recognized in Rao‘s study on international development NGOs; she found that within the same offices there was a deep divide between international and national staff, with the perception that international staff is superior in gender equality norms to national staff, and that gender mainstreaming was done ―better‖ by internationals (Rao et al, 1999:98-101). In general in the world of development assistance there is a recognized divide between national staff (locals) who are paid less for the same work as is done by international staff, often under the pretense they do not have the same skill-set as their ex-pat colleagues (Jackson, 2005).  This study found that female national staff were most likely to understand the structures of gender inequality in the context of their office and were in the position of least power.180 There was also a sense among some respondents that male national staff could be excused for having negative gender norms and values since they were a product of their patriarchal society.181 In addition it must be recognized that to some extent national staff members of UN agencies are elites within their own societies, often with connections to a regime or a powerful political faction, which means they may not necessarily identify with the problems                                             180 In fact power and benefits were clearly divided between national and international female staff. For example in one UNICEF office an international female staff member was afforded the opportunity to bring her baby with her to a seminar so as to continue breastfeeding while a national female staff member in the exact same position was not afforded the same benefit. UNICEF Official, Interview by Ana Lukatela, 2012.  181 Ingrid Baken, Interview by Ana Lukatela, Tirana, March 8, 2011. 244  of poor or vulnerable groups within their own society. There should be no excuse for staff that do not value gender equality, whether they are national male staff or international female, all should be held to the same expectation that they unequivocally value men‘s and women‘s equality as a development goal.  Accountability Structures for Crosscutting Policies Accountability mechanisms are the single biggest gap in the mainstreaming approach and this in turn stems from the lack of gender sensitive indicators, sex disaggregated data and monitoring and evaluation systems that use these tools to show results on gender mainstreaming in development programmes. The Civilian Capacity Report (2011) found the UN must urgently take:   …measures to improve accountability, particularly with respect to implementation and gender equality. It is recommended that the United Nations adopt a results-based audit culture that seeks to enable and improve implementation, rather than just punish administrative non-compliance (2011:8) This dissertation confirms these conclusions.  In UN development agencies the accountability of senior leadership for mainstreaming results is non-existent and with no sanctions senior leaders have little incentive to press staff below them for implementation. Respondents mused about the differences in accountability structures between military hierarchies and the UN development agency hierarchy, arguing that until sanctions were introduced there would be no 245  progress. 182 As long as mainstreaming results are not included as a core component of performance reviews for both middle and senior managers then the present situation will continue – where only a handful of individuals with personal commitment drive the agenda forward at the middle management level encouraged by some key norm champions above them. For example, a rigorous monitoring and evaluation system that is underpinned by a baseline with sex disaggregated data would be a valuable resource for learning lessons across contexts. If the monitoring and evaluation system is paired with sanctions for failures to meet targets it would be all the more effective. The lack of sanctions for failure to take action was a theme brought up in an overwhelming number of the interviews. In fact one respondent felt that the ―UN as an organization has no functional accountability mechanism… Programmes are rarely if ever held accountable and therefore generating a workplace accountability mechanism for gender equality in the UN seems to be a dream.‖183 Another UN official felt that if personal staff evaluations had a gender equality component and evaluations were a factor in promotions and step increases then there would be some incentive for behavior change.184 Bring the Focus back to Operations and Structural Transformation According to the research in this dissertation, gender mainstreaming has become confused with women in development in the minds of many international                                             182 Osnat Lubrani, Interview by Ana Lukatela, Pristina, May 2010. UN Official, Interview by Ana Lukatela, Sarajevo, June 2010. UN Official, Interview by Ana Lukatela, 2012. 183 UN Official, Interview by Ana Lukatela, 2012. 184 UN Official, Interview by Ana Lukatela, 2012. 246  development professionals. The term is in the world polity without a definition and only a vague understanding and lack of background knowledge by senior managers. This has meant that the pockets of interventions and attempts at implementation that exist have been piecemeal and have focused on programming. Gender mainstreaming's original goal of transforming the structures of international organizations themselves, to address their gendered nature and the inequality they perpetuate, never happened. Without putting its own ―house in order‖ the UN cannot be successful at gender mainstreaming.  Bringing a focus back on operations within the UN, and not just on its external development programming, is critical. The privileges and inequalities of different categories of staff within the UN system, often divided by class, gender and ethnicity/race, national/international, are serious impediments to the UN‘s attempts to promote gender equality. Changing this requires fulfilling the above policy recommendations – changing the profile of staff who are hired to reduce inequalities and introducing accountability systems so that privileges cannot provide immunity from sanctions for lack of implementation. Reconnect Gender Mainstreaming with the Fulfillment of Human Rights  Although there is scholarly agreement that gender mainstreaming has failed it is evident from the continuing critical engagement of scholars and of practitioners that the concept persists and has not been replaced within international development organizations. There are a three distinct positions regarding the future of the concept that have emerged out of the literature on gender mainstreaming and the accompanying lessons learned. While all three 247  strands recognize the failures of mainstreaming and are generally critical of the way in which it has been implemented thus far, they differ in their views on whether the concept should continue to be supported in the future.  The first strand is critical of mainstreaming but maintains that it is an important approach that should remain but can be improved on in terms of its  implementation. Parpart (2013) argues that challenging gendered power structures is the key to achieving the transformational goals of mainstreaming and that this can happen through small incremental changes. Gender-sensitive staff members can influence the perspectives of their peers on women's and men's relations in the workplace; research findings that make evident ongoing inequalities can challenge the norms and assumptions of staff; and champions among leaders have the potential to push for answers to uncomfortable questions on how organizations function (Parpart, 2013: 391).  Parpart argues, and this research agrees, that the  gendered nature of societies, politics, economies and international organizations are at the heart of the problem of why the neoliberal mainstreaming approach promoted by international organizations185 has not been successful. A second  strand also recognizes the problems of the mainstreaming approach, however, is non-committal as to whether it has a future, arguing that the concept has limitations that seem unlikely to change (Tolhurst et al, 2012).                                             185 Parpart argues that international organizations generally hold widespread support for neoliberal global solutions to development problems and that this has also influenced the way that gender mainstreaming is interpreted and implemented in these organizaitions. The values that are associated with leadership and power in the global neoliberal economy are "typical hegemonic masculine traits" and men and women who "follow local gendered power rules are seen as  the ‗natural‘ leaders of society" (2013: 390). 248  Based on their research across fifteen different country contexts of gender mainstreaming in the health policy sector Tolhurst et al argue that the disconnect between mainstreaming policy and practice has served as a significant obstacle to achieving its goals. This disconnect has worsened in particular through the lack of collaboration and consultation between academia and practitioners. The authors argue that greater collaboration between the two is critical and that a new strategic direction to gender mainstreaming focusing in particular on intersectionalities with race and sexuality is the only hope for the policy to possibly achieve results in the future.  The third strand argues that a completely different approach to gender mainstreaming is required that would fundamentally change its conceptualization (Hankivsky, 2013). For example, based on a five country comparison (Canada, Australia, Sweden, the United Kingdom, and Ukraine) Hankivsky argues that instead of adjustments to how gender mainstreaming is implemented, a complete rethinking of its conceptualization is required. If mainstreaming were understood as a strategy to bring together relevant diversities and address intersectionality, instead of only a gender perspective, it would address inequalities more comprehensively and prevent the zero sum game of understanding mainstreaming through the binary categories of "men" and "women". This could perhaps bring on board individuals who otherwise feel threatened by the perceived focus that mainstreaming is a feminist project that is interested only in women's rights. Given these views, this strand is far more critical as to whether mainstreaming has a viable future.  249  This debate remains relevant from both a theoretical and policy standpoint and this dissertation attempts to contribute to the discussion. The position of this dissertation is consistent  with elements of the first strand in that the biggest obstacle to gender mainstreaming is the gendered nature of society and organizations – such as the UN -  and the individuals they are composed of and how they  are relied on to implement mainstreaming. Within international organizations specifically, transformative changes depend on a critical mass of normatively committed middle management as well as a number of normatively committed champions at the senior level. Without individuals who  are willing to continually challenge the gendered nature of organizations and the societies that these organizations are linked to, there can be no success. At the same time, the position of the third strand also warrants attention especially the need for  better understandings of diversities through   intersectionality analyses. This could inform the work  of  programme managers as they grapple with the kinds of inequalities relevant in their context instead of trapping themselves in the approach that gender mainstreaming is "only about women" or is a zero sum game of taking power away from men and giving it to women.  However the concern that this dissertation raises with this third approach of "diversity or intersectionality mainstreaming" is that it would still rely on middle management; and there seem little grounds to assume that they would have a commitment to analyzing the different aspects of diversity that may be important, such as gender, race, class, sexual orientation, etc. If a normative commitment to 250  gender equality is missing among implementers it seems unlikely that a normative commitment to these other values would be present.  The question of how diversity mainstreaming would deal differently with the pitfalls of gender mainstreaming identified in this research is thus not clear. Given the theoretical implications of this dissertation, diversity mainstreaming would seem like a useful strategy with improving implementation by programme managers who are already normatively supportive of inequality issues; however, it is not clear that diversity mainstreaming would be any more successful in its implementation phase than gender mainstreaming with those middle managers who are normatively opposed.  Despite the difficulties with implementing gender mainstreaming, alternative strategies proposed by academics and scholars for promoting gender equality in development programming have not appeared within the discourse of professionals and practitioners. When specifically asked about alternatives in the course of interviews respondents either said that they knew of no other alternatives, or felt that gender mainstreaming could not be abandoned after so much had been invested. Only one respondent felt that given the failure of gender mainstreaming it might be worth it to abandon the strategy in favor of a return to a ―rights-based approach.‖186 These responses and the continuing development of new and revised Gender Mainstreaming strategies by international development agencies (recall                                             186 UN Official, Interview by Ana Lukatela, 2012. The rights based approach is essentially a focus on women‘s human rights as outlined foundational human rights documents but also additional legal obligations such as CEDAW and Beijing Platform for Action. 251  Table 2 in Chapter One) make it clear that mainstreaming will continue and will not fall by the wayside, punished for its failures, anytime soon. Osnat Lubrani, Kosovo‘s UN Resident Coordinator, responded to the question with concern: ―I would be worried about giving up on gender mainstreaming; I think that is a dangerous thing.‖187 She felt gender mainstreaming‘s philosophical focus on transforming social power relations was the key to progress and that it still had the potential to be successful. Giving up on mainstreaming would mean giving up on transformative change in her view. While abandoning mainstreaming would be dangerous, this does not mean that it cannot reconnect with the rights discourse in order to reestablish the link between human rights and gender equality in practitioner‘s minds. In her study of gender mainstreaming Eyben also despairs that ―the language of rights has disappeared‖ from the gender equality strategies of international development organizations (Eyben, 2010). A number of UNIFEM/UN Women staff expressed interest in re-thinking the use of the rights-based approach to gender equality, and reconnecting it to mainstreaming. However they had difficulty in articulating what this relationship might look like. Their responses overall left the impression that the gender mainstreaming discourse was increasingly perceived as an instrumental tool towards improving economic development indicators, as opposed to a tool towards achieving human rights and social transformation.                                             187 Osnat Lubrani, Interview by Ana Lukatela, Pristina, May 2010. 252  One respondent argued that reconnecting mainstreaming to human rights could bring significant benefits as it would be easier to mobilize action on ―human rights‖ than on ―gender equality‖ since more individuals are normatively committed to human rights and will not reject the discourse offhand.188 This view is very much in accordance with the literature on women‘s rights as human rights (Bunch 1987), as well as the constructivist approach to ―framing‖ new norms as part of a more mainstream discourse (Nadelmann, 1990). Another important difference this official identified was that by working through the discourse of human rights more diversity is brought to the implementation as different categories of vulnerable women are usually those that have the least opportunities to exercise their rights (minority women, poor women, rural women, elderly women, illiterate women, women with disabilities).  Ultimately she felt that framing gender mainstreaming in terms of women‘s human rights would get more traction with programme managers because it would effectively capitalize on and mobilize internalized human rights norms. However the question of what this would look like in practice was not explained by any respondents. In general this dissertation recommends that policy papers and strategies on gender mainstreaming ensure that the link to human rights is clearly established and that indicators developed for monitoring frameworks use rights-based language. Policies and strategies should focus more on what it takes to socialize their organizational culture into understanding the link between gender equality and human rights. In the evaluations of UN gender                                             188 UNIFEM Official, Interview by Ana Lukatela, 2010. 253  mainstreaming policies, presented above in Figure 1, it is curious in light of the results of this study, that only two evaluations discussed the problem of an organizational culture that is not supportive of gender equality. Organizations' could improve their implementation by focusing more on transforming organizational culture and presenting gender equality as a fundamental human rights issue. Framing gender mainstreaming as a tool and strategy to achieve human rights and equality of all persons is thus one option.     Conclusion  Gender mainstreaming has gained the increasing attention of the international development community since 1995, despite an overwhelming agreement by scholars and practitioners that it has not contributed to improving women‘s political, social or economic outcomes, nor has it led to institutional change within the international organizations adopting it (Tolhurst et al, 2012; Powell, 2005; Squires, 2005; Daly, 2005; Lang, 2009; Hankivsky, 2005; Hankivsky, 2009; Hankivsky, 2013; Lewis, 2006; Bendl and Schmidt, 2013; True and Parisi, 2013). This study has considered two central questions. What factors explain the differences in adoption and implementation of gender mainstreaming policies in international development organizations and what these dynamics tell us about the reasons for the persistence of such policies in the face of recognized failure. The theory proposed by this research argues that each stage of the policy process is influenced by different mechanisms.  While the study has found that 254  the answer to the first question is that gender mainstreaming has become part of the values of the world polity and embedded itself in the minds of the international civil servants working in the senior management of the UN system, the answer to the second question points to a break down in the dynamics of policy adoption vs. policy implementation and an underestimation of the influence of middle management. In the second phase of the policy process a different mechanism tips the scales – the normative commitments of mid-level management coupled with the standard operating procedures they can use to their advantage to influence the behavior of otherwise passive colleagues. This study represents a break from the typical IR approach of using one theoretical argument to explain an entire phenomenon and reasserts the importance of opening up the black box of organizations to examine actors at different levels and how they interact. It recognizes the complexities and multiple factors that influence international development organizations.  Importantly, it also allows that given different factors and responsibilities regarding policy outcomes, senior and middle management behavior can be explained by different theoretical explanations, world polity vs. normative constructivism, that are not necessarily incompatible due to the differing policy context that affect actors. To develop this analysis, the study first examined how policy makers describe their decisions, interest and commitment to adopting gender mainstreaming policies. Respondents in this category were overwhelmingly unable to explain gender mainstreaming as a concept or in its operation, neither could they pinpoint the origins of their support for the concept beyond an 255  understanding that gender mainstreaming was ―the way things were done‖. Because they are not required to take any direct action beyond formulating political support and directives they could remain passive and did not find themselves in situations where their values and norms were being tested or questioned.  The study then considered the process of policy implementation and the responses of mid-level management in terms of how they do or do not operationalize gender mainstreaming policy directives. In this case the key obstacle was the lack of normative commitment or active opposition to the issue of gender equality and women‘s empowerment.  Because this level of the organization is not passive, but rather are the ones effecting action and implementation, if their norms run counter to their implementation responsibilities they will quickly find themselves facing a dilemma of action.  The study situated the four approaches described in the theoretical framework within the different stages of the non linear policy process (policy context, policy adoption and policy implementation) and identified the mechanisms influencing each in separate ways. This has important implications for how International Relations literature deals with policy processes in international development organizations, as well as with how mainstreaming/cross-cutting policies are conceptualized and implemented. In this conclusion the implications for these types of policies will be discussed. In all three countries respondents from certain UN development agencies, such as UNFPA, WHO and UNIFEM, felt that the staff working there had self-selected themselves according to normative commitments such as the 256  Hippocratic Oath (as in the case of WHO or UNPFA) or women‘s rights/feminism (as in the case of UNIFEM). However with UNDP there was no such clear response related to the norms of staff in a country office. Instead, respondents from UNDP discussed multiple examples of staff members and colleagues that had no normative attachment to the values underpinning the UN‘s development work, such as gender equality, poverty reduction or social inclusion, and expressed concern that it continued to be possible for these types of people to be hired. The motivation for applying for a job with the UN was to gain personal status and financial gain, and certainly not normative.  The importance of middle management‘s interest in mainstreaming gender into their programmes was also stressed by Osnat Lubrani, the UN‘s Resident Coordinator in Kosovo, who herself recognized that despite her strong policy leadership on gender mainstreaming, she could not do the work of her programme managers and follow up with the implementation of each and every programme. She felt that: Having a Programme Manager of a big programme who can do a gender analysis… is critical. It has to be in there, we cannot do a separate programme for women returnees; it has to be embedded where the big money is otherwise you are marginalizing it. You do also need separate projects because it drives things; it is a boost and addresses specific issues… and that combination is what we are trying to do here now.  257  Almost 40 UN specialist agencies, the Secretariat and 28 peace operations189 employ 44000 staff around the world in more than 180 countries with an annual budget of approximately 31 billion USD (Global Policy Forum, 2012). If all these programs and budgets and staffing structures were gender mainstreamed the impact on women‘s political, economic and social outcomes would be dramatic.190  While scholars agree that there is a global increase in indicators related to equality of access (employment, education and health care) this progress has been extremely uneven with the least improvements in the countries with the worst trends. The pervasiveness of the trend and its weak correlation to economic growth and religious tradition leads Dorius and Firebaugh to conclude that the change is exogenous and cultural diffusion is responsible - the effect of the world polity they argue. However this improvement in access has not translated into improvement in equality of outcomes (Dorius and Firebaugh, 2010); women are still less than 20% of parliamentarians globally (19% currently compared to 12% in 1995191) and enter highly sex segregated labour markets (Charles and Grusky, 2004). The gender gap in political and economic outcomes as measured on an annual basis by the Global Gender Gap Report since 2006 has not decreased significantly (World Economic Forum, 2011).                                               189 See http://www.un.org/en/aboutun/structure/index.shtml for a list of UN system agencies. 190Evidence shows that increasing women‘s employment in post-conflict and conflict contexts for example has an impact on increasing not only family welfare, but also community welfare. If the temporary employment programmes organized by the UN would mainstream gender considerations and ensure that 40% of beneficiaries were women, we can infer this would have a marked impact on families and communities.  Justino, 2012. 191 See data on women in parliament at http://www.ipu.org  258  While empowering women would have an impact on political, economic and social development indexes in many different contexts, what cannot be let out of sight is the essential human need for dignity, rights and freedom for which men and women around the world equally yearn. Hillary Clinton‘s famous statement to the Fourth World Conference of Women in Beijing in 1995 made a passionate plea for women‘s human rights and touched on issues as far ranging as sex selection, trafficking in women, honour crimes, rape as a war crime and domestic violence and femicide;  grim acts that remain prevalent even 20 years later (see for example Saurabh, 2012; Jakobsson, 2013; Chesler, 2010). The Beijing Platform for Action is seen by gender equality advocates as an extremely strong set of commitments to gender equality. While the world polity of international development professionals remains strong on the value of gender equality, the usual member state suspects remain vocally against broadening the concept.  In 2012, for only the second time since the UN Commission on the Status of Women began holding its annual meetings, Member States failed to reach Agreed Conclusions on measures to improve the Status of Rural Women. There was a push-back by a vocal minority of conservative member states (Caglar, 2013) led by the Vatican against the use of the term ―gender stereotypes‖ in favor of the enigmatic ―feminine genius‖; presumably in the view of these actors ―gender stereotypes‖ do not exist. While the vast majority of UN development professionals present at negotiations laughed at what they perceived a horribly backward approach to gender issues, women‘s local NGOs present at CSW were not laughing. While gender equality is safe at the world 259  polity level at which these international development professionals operate, it is far from an agreed norm at the local level where these NGOs are fighting their battles, as this research confirms. One UNDP gender advisor described her work at the country level to promote women‘s rights as a constant struggle and likened it to ―riding a bicycle up a mountain‖; in her particular society if one stops pedaling at any moment in time all the gains are lost and the bicycle rolls back down the mountain.192 Gender mainstreaming as a tool to achieve the empowerment of ordinary women in contexts around the world will not be effective until it is understood as one step in achieving universal human rights and until the UN transforms its recruitment process to make it a necessary condition that newly hired middle management staff have internalized women‘s rights as human rights. Senior management commitment may be a necessary condition but it is far from sufficient for managers trying to implement mainstreaming in their portfolio; instead the power to move mainstreaming from paper to practice in international organizations lies with the middle management section of the structure, where normative commitment is a necessary condition and can be sufficient even without accompanying senior management commitment or SOPs. The traditional insistence of IR theory to apply theoretical approaches to international organization policy processes that neglect the role of middle management in policy implementation prevents a multi-layered understanding of the nature of the policy process in these organizations and how goals can be achieved.                                              192 UNDP Official, Presentation on Gender Mainstreaming in UNDP, New York, 2012. 260  Bibliography Adler, Emanuel. ―Seizing the Middle Ground: Constructivism in World Politics.‖ European Journal of International Relations 3.3: (1997): 319-363.  Allison, Graham. Essence of Decision: Explaining the Cuban Missile Crisis. 1st Edition. Boston: Little Brown, 1971.   Allison, Graham and Phillip Zelikow. Essence of Decision: Explaining the Cuban Missile Crisis. 2nd Edition. New York: Longman, 1999.  Alter, Karen. The European Court‘s Political Power: Selected Essays. New York: Oxford, 2009.   Arneil, Barbara. Politics and Feminism. Oxford: Blackwell, 1999.  Ashcraft, Karen. "Organized dissonance: Feminist bureaucracy as hybrid form." Academy of Management Journal 44 (2001): 1301-1322.  Avant, Deborah, Finnemore, Martha and Susan Sell. Who Governs the Globe? Cambridge: Cambridge University Press. 2010.  Bacchi, Carol, and Joan Eveline. "Gender Mainstreaming or Diversity Mainstreaming? The Politics of ―Doing‖. Nordic Journal of Feminist and Gender Research 17.1 (2009): 2-17.   Barkin, Samuel. International Organizations: Theories and Institutions. New York: Palgrave Macmillan, 2006.  Barnett, M. Eyewitness to Genocide: The United Nations and Rwanda. Ithaca: Cornell University Press, 2002.  Barnett, Michael and Finnemore, Martha. "The Politics, Power, and Pathologies of International Organizations," International Organization 53:4 (1999): 699-732.  Beckfield, Jason. ―Inequality in the World Polity: The Structure of International Organization.‖ American Sociological Review. 68. 3 (2003): 401-424.  Beckfield, Jason. ―The Dual World Polity: Fragmentation and Integration in the Network of Intergovernmental Organizations.‖ Social Problems. 55.3 (2008): 419–442.  Beckfield, Jason. "The Social Structure of the World Polity." American Journal of Sociology 115.4 (2010): 1018-1068.  261  Beer, Michael. ―The Critical Path for Change: Key to Success and Failure in Six Companies.‖ In Corporate Transformation: Re-vitalizing Organizations for a Competitive World. Ed Ralph Kilmann, et al. San Francisco: Josey-Bass, 1989.  Beijing Declaration and Platform for Action. September 1995. http://www1.umn.edu/humanrts/instree/e5dplw.htm  Bendl, Regine and Angelika Schmidt. "Gender Mainstreaming: An Assessment of Its Conceptual Value for Gender Equality." Gender, Work and Organization. 20.4 (2013): 364-381.  Beneria, Lourdes and Savitri Bisnath. Eds. Gender and Development: Theoretical, Empirical and Practical Approaches, Vol 1. Cheltenham: MPG Books, 2001.  Benner, Thorsten, Andrea Binder and Phillip Rotman. ―Learning to Build Peace? United Nations Peacebuilding and Organizational Learning: Developing a Research Framework.‖ Global Public Policy Institute Research Paper Series. 7 (2007): 1-81. http://www.gppi.net/fileadmin/gppi/GPPiRP_LearningToBuildPeace_Feb2007final.pdf  Bennett, Andrew. "Case Study Methods: Design, Use and Comparative Advantages." In Models, Numbers and Cases: Methods for Studying International Relations, eds. Detlef F. Sprinz and Yael Wolinsky-Nahmias. Ann Arbor: University of Michigan Press, 2004.  Bennett, Andrew and Alexander George. Case Studies and Theory Development in the Social Sciences. Cambridge: MIT Press: 2005.  Berkovitch, Nitza and Karen Bradley. ―The Globalization of Women's Status: Consensus/Dissensus in the World Polity.‖ Sociological Perspectives. 42.3 (1999): 481-498.  Beveridge, Fiona, and Sue Nott. ―Mainstreaming: A Case for Optimism and Cynicism.‖ Feminist Legal Studies 10 (2002): 299-311.  Beveridge, Fiona, et al. ―Mainstreaming and the engendering of policy-making: A Means to an End? Journal of European Public Policy 7.3 (2000): 385-405.   Boli, John. “World Authority Structures and Legitimations.” In Constructing World Culture: International Nongovernmental Organizations Since 1875. Eds. Boli, John and George M. Thomas. Stanford: Stanford University Press, 1999: 267-302.  262  Boli, John and George Thomas. "World Culture in the World Polity: A Century of International Non-Governmental Organization." American Sociological Review. 62.2 (1997): 171-190.  Boli, John and George M. Thomas, Eds. Constructing World Culture: International Nongovernmental Organizations Since 1875. Stanford: Stanford University Press, 1999.  Bonacker, Thorsten, Wolfgang Form and Dominik Pfeiffer. "Transitional Justice and Victim Participation in Cambodia: A World Polity Perspective." Global Society 25:1 (2011): 113-134.  Booth, Christine, and Cinnamon Bennett. ―Gender Mainstreaming in the European Union: Towards a New Conception and Practice of Equal Opportunities?‖ European Journal of Women‘s Studies 9.4 (2002): 430-446.  Boserup, Esther. Women‘s Role in Economic Development. London: George Allen and Unwin, 1970.  Bowornwathana, Bidhya and Ora-orn Poocharoen. "Bureaucratic Politics and Administrative Reform: Why Politics Matters." Public Organization Review 10 (2010):303–321.  Broome, André and Leonard Seabrooke. "Seeing like an International Organisation." New Political Economy 17.1 (2012): 1-16.  Bryson, Valerie. Feminist Political Theory: An Introduction 2nd ed. New York: Palgrave Macmillan, 2003.  Bunch, Charlotte. Passionate Politics: Feminist Theory in Action London: St. Martin‘s Griffin, 1987.   Butler, Judith. Gender Trouble: Feminism and the Subversion of Identity. New York: Routledge, 2006.  Burawoy, Michael et al. Ethnography Unbound Berkeley: University of California Press, 1991.  Burt, Sandra. Changing Methods: Feminists Transforming Practice. Toronto: University of Toronto Press, 1995.  Butcher, Kate and Clare Dickinson. Evaluation of the Joint UNDP/ World Bank/ UNAIDS Programme on mainstreaming AIDS into national development plans and processes. HLSP, March 2010. http://www.undp.org/content/dam/aplaws/publication/en/publications/hiv-263  aids/evaluation-of-the-joint-undp--world-bank--unaids-programme-on-mainstreaming-aids-into-national-development-plans-and-processes/Final%20UNDP%20report%2029%2004%2010.pdf  Cabrera-Belleza, Mavic, et al, UN Reform: What‘s in it for Women? Washington: Heinrich Boll Foundation North America, 2006.  Caglar, Gulay. "Gender Mainstreaming." Politics and Gender 9.3 (2013):336-344.  Calas, M.B. and L. Smircich. "Past Postmodernism? Reflections and Tentative Directions." Academy of Management Review 24.4 (1999): 649-671.  Carney, Gemma. ―Communicating or Just Talking? Gender Mainstreaming and the Communication of Global Feminism.‖ Women and Language 26.1 (2003): 52-60.  Convention on the Elimination of All Forms of Discrimination against Women, General Recommendation Number 25 on Temporary Special Measures. 1999. http://www.un.org/womenwatch/daw/cedaw/recommendations/General%20recommendation%2025%20%28English%29.pdf  Charles, Maria, and David Grusky. Occupational Ghettos: The Worldwide Segregation of Women and Men. Stanford University Press, 2004.  Checkel, Jeffrey T. "It‘s the Process Stupid! Process Tracing in the Study of International and European Politics." Arena, Centre for European Studies, University of Oslo. Working Paper No.26. 2004.  Chesler, Phyllis. "World Wide Trends in Honor Killings." Middle East Quarterly 17.2 (2010): 3-11.  Chesterman, Simon. ―Introduction: secretary or general?‖ In Secretary or General? The UN Secretary-General in World Politics. Ed. Chesterman, Simon. Cambridge: Cambridge University Press, 2007.  Civilian Capacity in the Aftermath of Conflict: Independent Report of the Senior Advisory Group (A/65/747 – S/2011/85). New York: United Nations, 2011. http://www.civcapreview.org  Clark, Rob. Technical And Institutional States: Loose Coupling in the Human Rights Sector of the World Polity." The Sociological Quarterly 51 (2010): 65–95.  Clark, Rob. "Child Labor in the World Polity: Decline and Persistence, 1980-2000." Social Forces 89.3 (2011): 1033-1055. 264   Clegg, Liam. "In the Loop: Multilevel Feedback and the Politics of Change at the IMF and World Bank." Journal of International Relations and Development 13.1 (2010): 59–84.  Cooley, Alexander. Logics of Hierarchy: The Organization of Empires, States, and Military Occupation Ithaca: Cornell University Press, 2005.   Cousens, Elizabeth and Charles K. Cater. Toward Peace In Bosnia : Implementing The Dayton Accords Boulder: Lynne Rienner Publishers, 2001.  Cueva-Beteta, Hanny, et al. ―What Women Want: Planning and Financing for Gender-Responsive Peacebuilding.‖ UN Women Sourcebook on Women, Peace and Security. New York: UN Women, 2012.  Daly, Mary. ―Gender Mainstreaming in Theory and Practice.‖ Social Politics 12.3 (2005):433-450.  Daley, Elizabeth, Dore-Weeks, Rachel and Umuhoza, Claudine. ―Ahead of the Game: Land Tenure Reform in Rwanda and The Process Of Securing Women's Land Rights.‖ Journal of Eastern African Studies. Volume 4.1 (2010): 131-152.  David, Maxine and Roberta Guerrina. "Gender and European external relations: Dominant discourses and unintended consequences of gender mainstreaming." Women's Studies International Forum 39 (2013): 53–62.  Debardeleben, Joan. "Applying constructivism to understanding EU–Russian relations." International Politics 49.4 (2012): 418–433.  Diethelm, Pascal and Martin McKee. "Denialism: What is it and How Should Scientists Respond?"  European Journal of Public Health 19 (2009): 2–4.  Dijkzeul, Dennis and Yves Beigbeder. Rethinking International Organizations: Pathology and Promise. New York: Berhghan, 2003.  Dili Declaration. A New Vision for Peacebuilding and Statebuilding. 2010. http://www.pbsbdialogue.org/documentupload/44927821.pdf  Dix, Sarah, Miranda Diego and Charles Norchi.  ―Innovations for Post Conflict Transitions: The United Nations Development Programme in the Democratic Republic of Congo.‖ Ash Center for Democratic Governance Research Paper. February 2010. 265  http://www.cd.undp.org/mediafile/Innovations%20for%20Post-Conflict%20Transitions.pdf  Dorius, Shawn F. and Glenn Firebaugh. "Trends in Global Gender Inequality." Social Forces. 88.5 (2010): 1941–1968.  DPKO. "Former Yugoslavia - UNPROFOR." 1996. Accessed 23 January 2014. http://www.un.org/Depts/DPKO/Missions/unprof_b.htm  Drori, Gili, John W. Meyer and Hokyu Hwang. Globalization and Organization: World Society and Organizational Change. Oxford: Oxford University Press, 2006.  Elliot, MA. "Human Rights And The Triumph Of The Individual In World Culture." Cultural Sociology 1.3 (2007): 343–363.  Ellis, David. ―The Organizational Turn in International Organization Theory.‖ Journal of International Organizations Studies.1.1 (2010): 11-28.  Elsey, H, R. Tolhurst, and S. Theobald. "Mainstreaming HIV/AIDS in development sectors: Have we learnt the lessons from gender mainstreaming?" AIDS Care 17.8 (2005)" 988-998.  Emerson, R, R. Fretz and L. Shaw. Writing Ethnographic Field Notes. Chicago: The University of Chicago Press, 1995.  Enloe, Cynthia. Bananas, Beaches and Bases: Making Feminist Sense of International Politics. Berkeley: University of California Press, 2000.  Eveline, Joan, Carol Bacchi and Jennifer Binns. "Gender Mainstreaming versus Diversity Mainstreaming: Methodology as Emancipatory Politics" Gender, Work and Organization. 16.2 (2009): 198-216.  Eyben, Rosalind. "Subversively Accommodating: Feminist Bureaucrats and Gender Mainstreaming." IDS Bulletin 41.2 (2010): 54-61.   Falleti, Tulia and Julia Lynch. "Context And Causal Mechanisms In Political Analysis." Comparative Political Studies 42.9 (2009): 1143-1166.  Farazmand, Ali. "Bureaucracy and Democracy: A Theoretical Analysis." Public Organization Review (2010) 10: 245–258.   Fearon, James. "Rationalist Explanations for War." International Organization 49 (1995): 379-414.  266  Fidishun, D. (2000) "Andragogy and Technology: Integrating Adult Learning Theory as we Teach with Technology." [Conference Paper]: 5th Annual Instructional Technology Conference. Accessed 23 January 2014. http://www.lindenwood.edu/education/andragogy/andragogy/2011/Fidishun_2005.pdf  Finnemore, Martha. "International Organizations As Teachers Of Norms: The United Nations Educational, Scientific, And Cultural Organization And Science Policy." International Organization 47.4 (1993): 565-597.  Finnemore, Martha. “Rules of War and Wars of Rules: The International Red Cross and the Restraint of State Violence.” In Constructing World Culture: International Nongovernmental Organizations Since 1875. Eds. Boli, John and George M. Thomas. Stanford: Stanford University Press, 1999: 149-168  Finnemore, Martha, and Kathryn Sikkink. ―International Norm Dynamics and Political Change‖. International Organization 52 (1998): 887-917.  Friedan, Betty. The Feminine Mystique. New York: Norton, 1963.  Gallarotti, G. ―The Limits of International Organization: Systematic Failure in the Management of International Relations.‖ International Organization, 45.2 (1991): 183–220.   Gillespie, Andra. "Participant Observation and the Political Scientist: Possibilities, Priorities, and Practicalities." APSA Symposium: 'Pracademics': Mixing an Academic Career with Practical Politics April 2011: 261-265.  Gilligan, Carol. In a Different Voice. Boston: Harvard University Press, 1982.  Gilpin, Robert. Global Political Economy: Understanding the International Economic Order. Princeton: Princeton University Press, 2001.  Givens, Jennifer and Andrew Jorgenson. "Individual Environmental Concern In The World Polity: A Multilevel Analysis." Social Science Research 42 (2013): 418–431.  Global Policy Forum. "Total UN System Contributions." Accessed 25 January 2014. http://www.globalpolicy.org/un-finance/tables-and-charts-on-un-finance/un-system-budget/27505.html  Goetz, Anne Marie. ―Gender and Administration.‖ IDS Bulletin 23.1 1992.  Goetz, Anne-Marie, ed. Getting Institutions Right for Women in Development. London: Zed Books, 1997. 267   Goetz, Anne Marie. ―Local Heroes: Patterns of Field Worker Discretion in Implementing GAD Policy in Bangladesh.‖ IDS Discussion Paper no. 358, University of Sussex, 1996.  Grabska, Katarzyna. ―Constructing ‗Modern Gendered Civilised‘ Women and Men: Gender-mainstreaming in Refugee Camps.‖  Gender and Development  19 (2011): 81-93.  Grieco, Joseph. "Anarchy and the Limits of Cooperation: A Realist Critique of the Newest Liberal Institutionalism." International Organization 42.3 (1988): 485-507.  Guy, Mary.‖Three Steps Forward, Two Steps Backward: The Status of Women‘s Integration into Public Management.‖ Public Administration Review. 53.4 (1993): 285-292.  Haas, Ernst. Beyond the Nation-state: Functionalism and International Organization. Stanford University Press,1964.  Haas, Peter. ―Introduction: Epistemic Communities and International Policy Coordination." International Organization 46:1 (1992): 1-35.  Habermas, Jurgen. The Structural Transformation of the Public Sphere. Hermann Luchterhand Verlag, 1962.  Hafner-Burton, Emilie, and Mark Pollack. ―Gender Mainstreaming Global Governance.‖ Feminist Legal Studies 10 (2002): 285-298.  Hafner-Burton, Emilie, and Mark Pollack. ―Mainstreaming Gender in the European Union: Getting the Incentives Right.‖ Forthcoming in Comparative European Politics. 2009.  Halliday, Terence and Bruce Carruthers. Bankrupt: Global Lawmaking and Systemic Financial Crisis. Stanford University Press, 2009.  Hankivsky, Olena. ―Gender vs. Diversity Mainstreaming: A Preliminary Examination of the Role and Transformative Potential of Feminist Theory.‖ Canadian Journal of Political Science 38.4 (2005): 977-1001.  Hankivsky, Olena. “Gender Mainstreaming in Neoliberal Times: The Potential of “Deep Evaluation.” In Public Policy for Women: The State, Income Security and Labor Market Issues. Cohen and Pulkingham Eds,Toronto: Univesrity of Toronto Press, 2009.  268  Hankvisky, Olena. "Gender Mainstreaming: A Five-Country Examination." Politics and Policy 41.5 (2013): 629-655.   Harrison, Kathryn and Lisa McIntosh Sundstrom. ―Introduction: The Comparative Politics of Climate Change.‖ Global Environmental Politics 7.4 (2007): 1-19.  Hill Collins, Patricia. Black Feminist Thought: Knowledge, Consciousness and the Politics of Empowerment. New York: Routledge, 2000.  Hooghe, Liesbet. The European Commission and the Integration of Europe. Cambridge: Cambridge University Press, 2002.  Hooghe, Liesbet. ―Several Roads Lead To International Norms, But Few Via International Socialization. A Case Study of the European Commission.‖ International Organization 59.4 (2005): 861-898.    Howlett, Michael and M. Ramesh. Studying Public Policy: Policy Cycles and Policy Subsystems. Oxford University Press, 2003.  Howlett, Michael, M. Ramesh and Anthony Pearl. Studying Public Policy: Policy Cycles and Policy Subsystems. Oxford University Press, 2009.  Hunt, Frances. "Making Education Inclusive For All." Insights 81 2010.  Inglehart, Ronald and Pippa Norris Rising Tide: Gender Equality and Cultural Change Around the World. Cambridge University Press, 2003.  International Crisis Group, ―The Dangers of Albania's Disputed Election.‖ May 2011. http://www.crisisgroup.org/en/publication-type/blogs/the-balkan-regatta/the-dangers-of-albanias-disputed-election.aspx  IPU. "Women in National Parliament - Inter-Parliamentary Union." Accessed 25 January, 2014. http://www.ipu.org/wmn-e/world.htm  Jackson, Cecile, and Ruth Pearson. Eds. Feminist Visions of Development: Gender Analysis and Policy. London: Routledge, 1998.    Jackson, Jeffrey. The Globalizers: Development Workers in Action. Baltimore: Johns Hopkins University Press, 2005.  Jahan, Rounaq. The Elusive Agenda: Mainstreaming Women in Development. New York: Zed Books, 1995.  Jakobsson, Niklas and Andreas Kotsadam. " The Law and Economics of International Sex Slavery: Prostitution Laws and Trafficking for Sexual 269  Exploitation." European Journal of Law and Economics 35.1 (2013)" 87-107.  Jarvis, Christopher. "The Rise and Fall of Albania's Pyramid Schemes." Finance and Development 37.1 (2000).  Jenkins, Rob. Peacebuilding: From Concept to Commission. Routledge, 2013.  Johnson, Jeffrey, Christine Avenarius and Jack Weatherford. "The Active Participant-Observer: Applying Social Role Analysis to Participant Observation." Field Methods 18.2 (2006 ): 111-134.    Johnstone, Ian. ―The Secretary-General as Norm Entrepreneur.‖ In Secretary or General? The UN Secretary-General in World Politics. Ed. Chesterman, Simon. Cambridge: Cambridge University Press, 2007.  Judah, Tim. Kosovo: War and Revenge. Yale University Press, 2002.   Justino, Patricia. ―Women Working for Recovery: The Impact of Female Employment on Family and Community Welfare after Conflict.‖ UN Women Sourcebook on Women, Peace and Security. New York: UN Women, 2012.  Kabeer, Naila, Reversed Realities: Gender Hierarchies in Development Thought. London:  Verso, 1994.  Kantola, Johanna. "Gender and the European Union." Political Studies Review 11.1 (2010): 130–131.  Kantor, Rosabeth. Men and Women of the Corporation. New Edition. New York: BasicBooks, 1993.  Kardam, Nuket. Bringing Women In: Women‘s Issues in International Development Programs. Boulder: Lynee Rienner, 1991.  Kardam, Nuket. "The Emerging Global Gender Equality Regime from Neoliberal and Constructivist Perspectives in International Relations." International Feminist Journal of Politics 6:1 (2004): 85–109.  Katzenstein, Peter. The Culture of National Security: Norms and Identity in World Politics. New York: Columbia University Press, 1996.  Keck, Margaret and Kathryn Sikkink. Activists Beyond Borders. Ithaca: Cornell University Press, 1998.  270  Kennedy, Paul. The Parliament of Man: The Past, Present, and Future of the United Nations. New York: Random House, 2006.  Keohane, Robert. After Hegemony: Cooperation and Discord in the World Political Economy. Princeton: Princeton University Press,1984.  Keohane, Robert. "Institutional Theory and the Realist Challenge after the Cold War." In Neorealism and Neoliberalism: The Contemporary Debate. Ed. Baldwin, David. New York: Columbia University Press, 1993: 269-300.  King, Gary, Robert O. Keohane and Sidney Verba. Designing Social Inquiry: Scientific Inference in Qualitative Research. Princeton: Princeton University Press, 1994.  Kingdon, John. Agendas, Alternatives and Public Policies. Boston: Little Brown, 1984.  Klotz, Audie. Norms in International Relations: The Struggle against Apartheid. Ithaca: Cornell University Press, 1996.  Knowles, Malcolm. The Adult Learner: A Neglected Species. Houston: Gulf Publishing, 1984.  Kolb, Deborah and Debra Meyerson. ―Keeping Gender in the Plot: A Case Study of the Body Shop.‖  in Gender at Work: Organizational Change for Equality. eds Rao et al. West Hartford: Kumarian Press: 1999: 129-186.  Koenig, Matthias. "Institutional Change In The World Polity: International Human Rights And The Construction Of Collective Identities." International Sociology 23.1 (2008): 95–114.  Koenig, Matthias and Julian Dierkes. "Conflict In The World Polity – Neo-Institutional Perspectives." Acta Sociologica 54.1 (2011): 5–25.  Koremenos, Barbara, Charles Lipson and Duncan Snidal. The Rational Design of International Institutions. Cambridge: Cambridge University Press, 2004.  Krasner, Stephen. International Regimes  Ithaca: Cornell University Press,1983.  Krasner, Stephen. ―Global Communications and National Power: Life after the Pareto Frontier,‖ World Politics 43:3 (1991): 336-366.  Lang, Sabine. "Assessing Advocacy: European Transnational Women's Networks and Gender Mainstreaming." Social Politics: International Studies in Gender, State and Society 16.3 (2009): 327-357. 271  Lavena, Cecilia and Norma Riccucci. "Exploring Gender Mainstreaming in the European Union." International Journal of Public Administration 35:2 (2012): 122-136.  Lieb, Stephen. ‖Principles of Adult Learning." Vision Journal Fall (1991). http://www.lindenwood.edu/education/andragogy/andragogy/2011/Lieb_1991.pdf  Lind, Amy. Gendered Paradoxes: Women‘s Movements, State Restructuring, and Global Development in Ecuador. Pennsylvania: Pennsylvania State University Press, 2005.  Lee, Jongkon. "The Administrative Broker: Bureaucratic Politics in the Era of Prevalent Information." The American Review of Public Administration 43:6 (2013): 690–708.  Lee-Gosselin, Helene, Sophie Briere and Hawo Ann. "Resistances To Gender Mainstreaming In Organizations: Toward A New Approach." Gender in Management: An International Journal. 28. 8 (2013): 468-485.  Legro, Jeffrey. Rethinking the World: Great Power Strategies and International Order. Ithaca: Cornell University Press, 2005.  Lerner, G. The Creation of Feminist Consciousness: From the Middle Ages to 1870. Oxford: Oxford University Press, 1993.  Lewis, Stephen. Race against Time. Toronto: Anansi Press, 2005.  Lewis, Stephen. ―Remarks to a High Level Panel on UN Reform‖ Geneva, February 7, 2006, http://www.sarpn.org/documents/d0002065/index.php (accessed July 15, 2012).  Lumsdaine, David Halloran. Moral Vision in International Politics: The Foreign Aid Regime. Princeton: Princeton University Press, 1993.  MacKinnon, Catherine. Toward a Feminist Theory of the State. Cambridge: Harvard University Press, 1989.  Mainstreaming a Gender Perspective into UN Policies and Programmes. Speech delivered by UN Women Executive Director Michelle Bachelet at the ECOSOC Substantive Session, Geneva, 13 July 2011. http://www.unwomen.org/2011/07/mainstreaming-a-gender-perspective-into-un-policies-and-programmes/  272  Mansbridge, Jane, and Aldon Morris, eds. Oppositional consciousness: The Subjective Roots Of Social Protest. Chicago: University of Chicago Press, 2001.  Marsh, Kevin. "The Intersection of War and Politics: The Iraq War Troop Surge and Bureaucratic Politics." Armed Forces and Society 38.3 (2012): 413 - 437.  Mazur, Amy. Theorizing Feminist Policy. Oxford: Oxford University Press, 2002.  Mazey, Sonia. ―Introduction: Integrating Gender – Intellectual and Real-World Mainstreaming.‖ Journal of European Public Policy 7.3 (2000): 333-345.  Mazey, Sonia. ―Gender Mainstreaming Strategies in the EU: Delivering on an Agenda?‖ Feminist Legal Studies 10 (2002): 227-240.  Mearsheimer, John J., ―The False Promise of International Institutions,‖ International Security 19 (1994): 5-49.  Meier, Petra and Emanuel Lombardo. "Gender Quotas, Gender Mainstreaming And Gender Relations In Politics." Political Science 65 (2013): 46-62.  Meyer, Christopher and Eva Strikmann."Solidifying Constructivism: How Material and Ideational Factors Interact in Market Studies. 49.1 (2011): 61–81.  Meyer, John, John Boli, George Thomas, and Francisco Ramirez. "World Society and the Nation-State." American Journal of Sociology. 103 (1997): 144-81.  Mohanty, Chandra "Under Western Eyes: Feminist Scholarship and Colonial Discourses" Feminist Review 30 (1988): 61-88.  Morgenthau, Hans. Politics Among Nations: The Struggle for Power and Peace. Third Edition. New York: Knopf, 1960.  Moser, Caroline and Annalise Moser. "Gender Mainstreaming since Beijing: a Review of Success and Limitations in International Institutions." Gender and Development 13.2 (2005): 11-21.  Murphy, Craig. The United Nations Development Programme: A Better Way? Cambridge: Cambridge University Press, 2006.  Nadelmann, Ethan. ―Global Prohibition Regimes: The Evolution of Norms in International Society.‖ International Organization 44 (1990): 479–526.  Nelson, Paul. The World Bank and Non-Governmental Organizations. New York: St. Martin‘s, 1995. 273   Nelson, Barbara. ―Public Policy and Administration: An Overview.‖ in A New Handbook of Political Science. Goodin and Klingemann, Eds. Oxford: Oxford University Press, 1996.  Niskanen, William. Bureaucracy and Representative Government. Chicago: Aldine, 1971.  North, Douglass. Institutions, Institutional Change and Economic Performance. Cambridge University Press, 1990.  Njenga, Mary, Nancy Karanja, Gordon Prain, Diana Lee-Smith and Michael Pigeon. "Gender Mainstreaming In Organisational Culture And Agricultural Research Processes." Development in Practice 21.3 (2011): 379-391.  O Riain, Sean. "States and Markets in an Era of Globalization." Annual Review of Sociology. 26 (2000): 187-213.  Oh, Minjoo and Jeffrey Jackson. "Animal Rights vs. Cultural Rights: Exploring the Dog Meat Debate in South Korea from a World Polity Perspective." Journal of Intercultural Studies 32.1 (2011): 31-56.   Page, Margaret. "Gender Mainstreaming — Hidden Leadership?" Gender, Work and Organization. 18.3 (2011): 318-336.  Paris Declaration on Aid Effectiveness. March 2005. http://www.oecd.org/document/18/0,2340,en_2649_3236398_35401554_1_1_1_1,00.html  Parpart, Jane. "Exploring the Transformative Potential Of Gender Mainstreaming in International Development Institutions." Journal of International Development 26 (2014): 382–395.  Payne, Rodger. ―Persuasion, Frames and Norm Construction.‖ European Journal of International Relations. 7.1 (2001): 37-61.  Peters, B. Guy. The Politics of Bureaucracy. Third Edition. New York: Longman, 1989.  Peters, Julie and Andrea Wolper (Eds.) Women‘s Rights, Human Rights: International Feminist Perspectives. London: Routledge, 1995.  Peters, B. Guy. The Politics Of Bureaucracy. Fifth Edition. Routledge. 2001.  Pew Research Center. ―Global Attitudes Project: Gender Equality Universally Embraced but Inequalities acknowledged.‖ Washington: Pew Research 274  Center, 2010. http://www.pewglobal.org/files/2010/07/Pew-Global-Attitudes-2010-Gender-Report-July-1-12-01AM-EDT-NOT-EMBARGOED.pdf  Phillips, Lynne. ―Gender Mainstreaming: The Global Governance of Women?‖ Canadian Journal of Development Studies 26 (2005): 651-663.  Pollack, Mark. ―Delegation, Ag