BARGAINING WITH EMPOWERMENT: A MIXED-METHODS ANALYSIS OF THE CONTESTED CONCEPTUALIZATION, IMPLEMENTATION, AND EVALUATION OF WOMEN’S EMPOWERMENT By Vanessa Rene Rickenbrode A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Sociology – Doctor of Philosophy 2023 ABSTRACT While the use of the phrase “women’s empowerment” has become commonplace in academic, government, and grassroots circles, it has multiple meanings. These multiple meanings lead to confusion, incommensurable results, and competing interpretations across parties regarding what programs and policies empower women. The hidden dissensus over the apparent consensus of empowering women can have consequences that are unintentionally harmful to women and costly for agencies. In the face of an increasing interest and investment into programs of women’s empowerment, the time for interrogating the contested terrain is now. This dissertation uses a variety of methods to study women’s empowerment throughout the development process (conceptualization, implementation, and evaluation) across three studies. In the first study, I use the frame analytical approach to demonstrate the shared understandings and conflicts over women empowerment’s conceptualization in the development discourse. Using 54 books and articles written by women’s empowerment advocates from all over the world, and a grounded approach to coding, I demonstrate the 16 unique ways empowerment frames vary in the way they elaborate or amplify certain problems, solutions, and end goals. These unique collective action frames, or “stories,” of empowerment fall into one of three areas of emphasis: equal rights and participation, power, or agency. The findings from this frame analysis can help activists, scholars, and governmental bodies understand the tradeoffs of promoting and excluding certain frames of empowerment and select the one that aligns with their movement’s aims. The second study examines the way empowerment is conceptualized and then implemented in practice by using a case study of three non-governmental organizations (NGOs) working together to implement and evaluate an intervention in rural India that combines two strands of women’s empowerment funding: nutrition and self-help groups (SHGs). Applying a grounded approach guided by frame analysis to 10 semi-structured interviews with NGO staff, and content related to the project’s proposal, intervention materials, and evaluations, I show the way conceptualizations of women’s empowerment led to particular solutions (e.g., interventions) implemented on the ground. In this case, framing maternal (mal)nutrition as the ultimate problem and end goal of the project and conceptualizing empowerment as a means to this end, led to the implementation of women’s empowerment in the form of women’s SHGs. Findings from this case study reveal that factors informing the implementation of women’s empowerment are primarily due to the diagnostic frame (e.g., problem identification and attribution) and funding sources. Lastly, in the third study, I create a feminist quantitative model of women’s empowerment to evaluate women’s empowerment in rural Jharkhand, India. This study addresses the methodical gap left by feminists primarily utilizing qualitative methods and biases towards quantitative analyses in international development. It also addresses the very real and high rates of undernutrition and intimate partner violence (IPV) in India that disproportionately affect women. Using data from India’s National Family Health Survey 2019-21, I apply a feminist methodology to confirmatory factor analysis and structural equation modeling. In doing so, I model women’s empowerment as a process of women gaining transformative control over their lives (including their own nutrition) within a system of power inequalities based on caste, age, and forms of patriarchy (including men’s patriarchal values, IPV, and women’s internalized oppression). I find that agency (measured as decision making) is not a valid means to improve women’s strategic needs because it likely reflects burdened agency or patriarchal bargains in the context of rural Jharkhand. Moreover, the direct and indirect effects of men’s gender values and IPV suggest that development programs in India that are focused on improving women’s strategic empowerment should focus more attention towards mitigating IPV by addressing men’s patriarchal values. Copyright by VANESSA RENE RICKENBRODE 2023 To the empowered women and girls of the world, whose strength is inspiring. v ACKNOWLEDGEMENTS These last five years, I have been researching, writing, and living a dissertation devoted to women’s empowerment. My experience has proven to me that empowerment is not possible without the support of others. I am indebted to those who empowered me to accomplish a Ph.D. in Sociology; and to propose (and propose again), complete, and defend an internationally focused dissertation during a global pandemic. My greatest appreciation goes to my chair, mentor, and friend, Dr. Wynne Wright, whom I have learned so much from. I am a better feminist researcher because of her. She has guided me throughout my seven years of graduate education; supported and humbled me to experience international fieldwork in Tanzania; taught me how to conduct rigorous qualitative analyses; and has been the fiercest feminist, lavender farmer, Ph.D. to have ever advocated for me. I am grateful to my dissertation committee members for their insight and advice. Dr. Raymond Jussaume, who always challenged me to achieve greater clarity and provided me thoughtful feedback in all of my work, from my master’s thesis and comprehensive exams to this dissertation. Dr. Hui Liu, whose expertise in statistical methods and prompt generosity in providing guidance as a committee member allowed me to pursue my dreams of conducting structural equation modeling as part of my COVID-safe dissertation plans. Dr. Sejuti Das Gupta, whose expertise, readings, and meetings allowed me to form a deeper respect and understanding of conducting research in India. While not on my committee, Dr. Leif Jensen has been a mentor to me since I was an undergraduate student at the Pennsylvania State University. It is no exaggeration to say that I came to Michigan State University in pursuit of my Ph.D. in Sociology because of him. He has vi been a model of the kind of professional I aspire to be – kind, respected, and driven by a desire to improve people’s lives. This dissertation could not have materialized without the aid of several institutions and individuals. Michigan State University’s Graduate School, the College of Social Sciences, the Sociology Department, and the Center for Gender in Global Context, whose fellowships and funding financially aided my dissertation research and completion. I am also grateful to the individuals associated with the International Food Policy Research Institute (IFPRI), Professional Assistance for Development Action (PRADAN), and Public Health Resource Society (PHRS), who gave their time to be interviewed as part of this dissertation research. Thank you for sharing your experiences and expertise, for trusting me with your words, and for the lessons you taught me. I owe a million waves of gratitude to my colleagues, whom I met in the Sociology Graduate Program and became lifelong friends. I am especially grateful to Dr. Anna Wilcoxson, Dr. Kayleigh Ward, and Dr. Jennifer Lai for their integrity and generosity in sharing with me and showing up in ways that have been invaluable to me. My gratitude extends to Dr. Inna Mirzoyan, Dr. Brittany Tucker, and Marie Carmen Shingne - thank you for walking this path with me in solidarity, friendship, and with humor, especially when it came to us tackling our dissertations during COVID-19 lockdowns. The unwavering love and support from my family are pillars on which this dissertation stands and where my most intimate sources of empowerment reside. My parents, Denise and Jim Rickenbrode, who whole-heartedly supported my decision to pursue my Ph.D., cheered me on every second of the way, and always believed in me - even in times I did not. Their unconditional love continues to be my constant source of strength. I also owe my deepest depts and sincerest vii appreciation to my partner, Logan Taylor, who has proven to be my most trusted friend, a patient partner, and the most generous and empathetic person I have ever met. He has supported me in every sense of the word and continues to empower me in ways that are truly immeasurable. My last gratitude goes to my smallest family member, Winston, our cheeky chihuahua. While he might be the one that received the label “rescued” the day we adopted him, I believe we rescued each other that day – for he too empowers me with his unwavering love and support, albeit in his own little ways. viii TABLE OF CONTENTS CHAPTER 1: INTRODUCTION ................................................................................................... 1 REFERENCES .......................................................................................................................... 11 CHAPTER 2: CONCEPTUALIZING EMPOWERMENT FOR WOMEN IN THE GLOBAL SOUTH: A FRAME ANALYSIS OF CONTESTED CLAIMS IN THE DEVELOPMENT DISCOURSE ................................................................................................................................ 13 REFERENCES .......................................................................................................................... 61 APPENDIX A: FINAL SAMPLE............................................................................................. 63 APPENDIX B: SAMPLING PROCESSES .............................................................................. 66 APPENDIX C: PRACTICAL AND STRATEGIC GOAL CODES ........................................ 69 APPENDIX D: COLLECTIVE ACTION FRAME HEAT MAPS .......................................... 70 APPENDIX E: POWER CODES ............................................................................................. 76 CHAPTER 3: IMPLEMENTING EMPOWERMENT FOR WOMEN IN INDIA: THE CASE OF SELF-HELP GROUPS AS A MEANS TO IMPROVE NUTRITION .................................. 77 REFERENCES ........................................................................................................................ 115 CHAPTER 4: EVALUATING EMPOWERMENT FOR WOMEN IN JHARKHAND, INDIA: A FEMINIST STRUCTURAL EQUATION MODELING APPROACH ................................. 119 REFERENCES ........................................................................................................................ 152 APPENDIX A: REDUCED MEASUREMENT MODELS ................................................... 157 APPENDIX B: DIRECT AND TOTAL INDIRECT EFFECTS IN STRUCTURAL EQUATION MODEL ............................................................................................................. 158 CHAPTER 5: CONCLUSION ................................................................................................... 159 REFERENCES ........................................................................................................................ 168 ix CHAPTER 1: INTRODUCTION Women’s empowerment is prominently featured in the contemporary field of international development, used by feminist and grassroots activists, development practitioners, and governmental bodies in their policies, programs, and research on women in the Global South. The formal appearance of the term “women’s empowerment” in the development discourse can be credited to the Third World feminist movement DAWN (Development Alternatives with Women for a New Era) (Calvès 2009). These feminists articulated and promoted their alternative “empowerment approach” to women in development in the publication Development, Crises and Alternative Visons: Third World Women’s Perspectives (Sen and Grown 1987). From the perspective of Third World women themselves came an understanding that post-colonial development strategies exacerbated women’s oppression through systems of inequality based on gender that intersected with caste, class, race, region, and religion. They particularly criticized top-down development approaches that did not take the interests of women into account and treated women as instruments to achieve development goals (e.g., population control) (Sen and Grown 1987). Thus, the alternative “empowerment approach” was born. From this perspective, women’s empowerment was conceptualized as a collective feminist movement that respected the diversity of issues, immediate goals, and methods responsive to the different needs and concerns of women around the world, but remained connected by their core commitment towards breaking down systems of gender oppression to improve women’s lives (Sen and Grown 1987). Since its formal introduction into the development discourse more than 50 years ago, the term “women’s empowerment” has been adopted and elaborated on by feminists and development institutions. In particular, the 1990s witnessed a widespread co-optation of 1 women/gender/female empowerment, which for the ease of comprehension, I refer to all these terms simply as “women’s empowerment” for the remainder of this dissertation. Some scholars suggest women’s empowerment gained momentum after Kabeer’s (1999) seminal piece, where she elaborated a framework of empowerment consisting of resources, agency, and achievements (Priya et al. 2021). However, most identify the 1995 Fourth World Conference on Women in Beijing, and to a lesser extent the 1994 International Conference on Population and Development in Cairo, as playing a critical role in the institutionalization of term (Batliwala 2007; Calvès 2009). By the end of the 1990s, governments and international development organizations anxious to demonstrate a progressive stance on gender and development had adopted the term empowerment into their policies and programs (Batliwala 2007). Today, the magnitude of women’s empowerment in international development is evident in the United Nations Millennium Development Goal and Sustainable Development Goal agendas; in which the latter lists achieving gender equality and empowering all women and girls by 2030 as the fifth of 17 SDGs. The global spread of the concept’s importance is also seen at the level of governmental bodies and non-governmental organizations (NGOs) through their programs aimed at empowering women. For instance, in India, there are 41 government schemes and countless NGOs dedicated to the empowerment of women, mainly targeting women in rural areas (Sangwan and Bawa 2022). In particular, the Indian government, along with NGOs and funding agencies, have increasingly sought to use women’s self-help groups (SHGs) as a way to empower women and deliver development goals (particularly those related to income, agriculture, and child nutrition) (Kumar et al. 2021; Nichols 2021). As of July 2023, over 83 million households included members of the government’s National Rural Livelihood Mission (NRLM) SHGs (Ministry of Rural Development 2023). 2 Contested Terrain of Women’s Empowerment in Development On the outside, women’s empowerment has clearly gained consensus around its overall importance in the development discourse. But within the development discourse, it has become the most contested terrain today (Cornwall and Whitehead 2007). Feminists now suggest that the proliferation of women’s empowerment in the development discourse may seemingly signal an “objective”1 commitment towards women, but they warn that not all empowerment frames lead to the improvement of women’s lives. In some cases, they may even maintain or exacerbate women’s disempowerment (i.e., oppression, subordinate status). Feminists have noted the ways empowerment becomes depoliticized and denatured when taken up by development institutions (Cornwall and Whitehead 2007). They particularly take issue with empowerment frames that narrowly focus on women’s economic empowerment, those that use women as instruments to achieve household or national level development outcomes, and those applying essentialist claims about women’s “natural” inclination to be more caring and responsible compared to men (Batliwala and Dhanraj 2007; Cornwall and Whitehead 2007). I argue that one of the most mainstream empowerment frames in the development discourse today encapsulates these assumptions by arguing for women’s empowerment as an instrument (or means) to improving child nutrition on the basis that women are better allocators and caregivers compared to men. The debates described above illustrate what many of the leading empowerment advocates say is the most contested ways empowerment is conceptualized between feminists and development institutions: as either an intrinsic or instrumental goal (Ibrahim and Alkire 2007; Kabeer 1999; Narayan 2005). However, even among feminist scholars, concepts regarding what 1 The term “objective” is in quotations to reflect a positivist (rather than a feminist) epistemological interpretation of the word that is commonly used in empirical studies; as an all-encompassing, singular, and “value-free” truth (DeVault 1996). 3 is considered empowering differ. Conceptualizations of empowerment vary not just by place and time, but at its core, they vary based on their answers to how individuals are disempowered, what should be done to empower them, and why empowerment is important. In the frame analytic vocabulary, these questions translate to specific framing tasks claims-makers take to identify the problem (diagnosis), promote a set of solutions (prognosis), and actively mobilize (motivate) individuals to participate in the goals of women’s empowerment (Benford and Snow 2000; Snow and Benford 1988; Wright and Reid 2010). Frame analysis recognizes that the same phrase (e.g., women’s empowerment) can be used differently because “meaning is nuanced by the discursive context in which it is situated” (Mooney and Hunt 2009:471). In the process, several distinct claims and uses of the term arise in ways that are rarely scrutinized or comparatively analyzed in an empirically meaningful way. As scholars and activists continue to claim empowerment in various ways, the time for interrogating the term is now. My primary analytical approach to understanding these differences in conceptualizations in the development discourse draws from the frame analytical approach derived from the field of social movements. The framing perspective was outlined in a sequence of papers, most notably by David Snow and Robert Benford in the field of social movements (Snow et al. 1986; Snow and Benford 1988; Snow and Benford 1992). Drawing from Goffman’s (1974) idea of framing, Snow and Benford (1992:137) define collective action frames as “an interpretive schemata that simplifies and condenses the “world out there” by selectively punctuating and encoding objects, situations, events, experiences, and sequences of actions within one’s present or past environments.” In applying this perspective to the women’s empowerment and development discourse, the frame approach is analytically capable of demonstrating the contested ownership behind the apparent 4 consensus on women’s empowerment by drawing attention to the shared understandings and conflicts over its meanings (Benford and Snow 2000; Gamson 1992; Wright and Reid 2010). It also provides an analytical approach to understanding the ways language translates to mobilizing people to action, including how framing empowerment at the conceptual level shapes possibilities for program design and outcomes at the practical level (Benford and Snow 2000). In this respect, frames can be used to explain why certain programs and policies are implemented and others are not. At the same time, research that goes beyond policy analysis that incorporates interviews with organizational staff suggests there are multiple dimensions of gender and development discourse, including the various ways concepts like gender are translated in different cultures and implemented on the ground (Ferguson 2010). Taken together, I apply the frame analytical approach to demonstrate the multiple, and possibly contradictory, ways in which women’s empowerment is interpreted and deployed by various actors in the implementation stage of the development process. Amid these competing claims over women’s empowerment’s conceptualization and implementation in the development discourse have come newer debates over its measurement and evaluation (Narayan 2005). While there are some that argue empowerment is an immeasurable concept, in a context where feminists and organizations aim to empower women, effective empowerment frames must link concepts of empowerment to concrete observations that capture its intended meaning to know if empowerment is “happening” or “has happened” (Kabeer 1999). In other words, to measure and assess empowerment, one must operationalize the concept of empowerment. To date, the most popular indicators used to monitor and evaluate women’s empowerment in the development discourse stem from quantitative methods. One of these quantitative tools includes the Women’s Empowerment in Agriculture Index (WEAI). In 5 the context of agricultural change, the WEAI is the most popular empowerment methodology, used by academics and practitioners working on projects related to women’s empowerment in rural areas throughout the Global South (Malapit et al. 2019). In short, the WEAI is an aggregate index made up of two indices based on individual-level gender-disaggregated survey data. The majority (80 percent) of the index quantitatively measures men and women’s agency in five different domains related to their productive life, and the rest (20 percent) evaluates the rate of women that exhibit gender parity with men (Alkire et al. 2013). While the popular use of quantitative methods dominates the field of international development, it is contested by feminists advocating for qualitative methods on the basis that they tend to be more “feminist” by focusing on women’s daily practices and their analytical ability to interpret power and oppression. As a result, feminists have primarily relied on qualitative methods to empirically understand women’s empowerment processes (Rowlands 1995). At the same time, some feminists caution against the use of subjective interpretations as fulfilling the entire criterion for evaluating women’s empowerment (Rowlands 1997; Kabeer 1999). This idea comes from recognizing the more hidden forms of women’s disempowerment (i.e., internalized oppression), as women sometimes support and take part in activities they consider “empowering” despite the way the activities maintain their subordinate status and exacerbate their oppression. One proposed way around this is to recognize women’s internalized oppression in analysis and employ the distinction between women’s practical and strategic needs (Rowlands 1997). To clarify, women’s practical needs result from their position in society, whereas their strategic needs challenge the gender hierarchies and other mechanisms of subordination (Moser 1989). Using these understandings can strengthen the focus on women’s 6 strategic empowerment by not losing sight of the transformational changes required to systematically address gender inequalities. Despite these important refutes by qualitative feminists, quantitative indicators and analyses continue to be the most convincing type of data when it comes to influencing policy decisions outside of academia (DeVault 1996). At the same time, many feminist methodologists suggest that there is nothing inherently feminist or non-feminist about a research method, as methods refer to the techniques researchers employ to answer specific research questions (DeVault 1996; Hesse-Biber 2010:170). Both feminist and non-feminist methods can apply methodologies that involve bringing new topics into the discourse or producing data intended to impact policy decisions, but what makes it feminist is “its relevance to change in women’s lives or in the systems of social organization that control women” (DeVault 1996:34). Accordingly, the ultimate goal of this dissertation is to present and utilize analytic approaches that will contribute to the goals of feminist research. Namely, by using methods that seek to know more about women in ways that provide a more accurate account of society that will benefit them. Three Study Dissertation The climate we find ourselves in today is one that acknowledges women’s empowerment as an important goal. However, it is against this backdrop that distinct and contested claims over its meaning, implementation, and measurement lie. What are the theoretical, methodological, and practical consequences of claiming women are “empowered” if we mean different things? There is a dire need for understanding and demonstrating the differences in empowerment approaches, especially in the face of an increasing interest in implementing and evaluating women’s empowerment in the Global South. Not only is this of theoretical and methodical significance 7 within academic circles but is especially important in applied settings where the framing of women’s empowerment has tangible consequences on people’s lives. This dissertation uses a variety of qualitative and quantitative methods to address the current challenges presented by the contested terrain of women’s empowerment across three studies. Each study builds on the last and uses a different method of exploration to address the contested terrain of women’s empowerment throughout the entire development process (i.e., conceptualization, implementation, and evaluation). In the first study (chapter two), I use the frame analytical approach to demonstrate the shared understandings and conflicts over women empowerment’s conceptualization in the development discourse. Using 54 books and articles written by women empowerment advocates from all over the world, and a grounded approach to coding, I demonstrate the ways empowerment frames vary in the way they elaborate or amplify certain problems, solutions, and end goals. Specifically, I find 16 distinct collective action frames, or “stories,” of women’s empowerment in the development discourse that derive from one of the three areas of emphasis: equal participation and rights, power, or agency. These findings suggest there are 16 different interpretations of what “successful” empowerment looks like for women in the Global South. Thus, the study provides the defining features between frames that produce competing conclusions regarding whether a program or policy empowers women. Ultimately, this research can help activists, scholars, and governmental bodies understand the tradeoffs of promoting and excluding certain solutions and forms of feminisms over others and select the one that aligns with their movement’s aims. The second study (chapter three) picks up where the first study ends, by further examining the way empowerment is conceptualized and then implemented in development programs on the ground. I do this by offering a case study of three NGOs working together to 8 implement and evaluate a development project in rural India that combines two strands of gender empowerment funding: nutrition and self-help groups (SHGs). Applying a grounded approach guided by frame analysis to ten semi-structured interviews with NGO staff and content related to the project’s proposal, intervention materials, and evaluations, I show the way conceptualizations of women’s empowerment leads to particular solutions (e.g., interventions) that are implemented on the ground. In this case, implementers and evaluators similarly identify maternal (mal)nutrition as the ultimate problem and end goal of the project and conceptualize women’s empowerment as a means to this end. In practice, the concept of empowerment, as an instrument or platform for delivering nutrition behavior change communication, is implemented in the form of women’s SHGs. Findings from this case study reveal that the range of possible empowerment solutions that can be implemented on the ground depend on the development project’s diagnostic frame (e.g., problem identification and attribution), resources, and sources of funding. In the third study (chapter 4), I create a feminist quantitative model of women’s empowerment that can serve as an example for future evaluations of women’s empowerment in the development discourse. This study addresses the current methodical gap in the contested discourse between feminists primarily utilizing qualitative methods and the bias towards quantitative analyses in the field of international development. It also addresses the very real and disproportionate high rates of undernutrition and intimate partner violence (IPV) that characterize women’s disempowerment in India. Using nationally representative data from India’s National Family Health Survey 2019 - 2021 (NFHS-5), I apply a feminist methodology (guided primarily by Indian feminist theorists) to confirmatory factor analysis and structural equation modeling. In doing so, I model women’s empowerment as a process of women gaining transformative control over their lives (including their own nutritional status), in which women 9 overcome sources of disempowerment relevant to power, including caste, age, and three forms of patriarchy that overwhelmingly characterize married women’s daily lives in rural Jharkhand (i.e., men’s patriarchal values, women’s experience of IPV, and women’s internalized oppression). I find that agency (measured as decision making power) is not a valid means to improve women’s strategic needs, because it likely reflects burdened agency or a patriarchal bargain in the context of rural Jharkhand. Moreover, the direct and indirect effects of men’s gender values and IPV suggest programs and policies in India that aim to improve women’s strategic empowerment (including their nutritional status) should focus more attention towards mitigating IPV by addressing men’s patriarchal values. Following these three individual studies, I provide a concluding chapter (chapter 5), that ties together the findings of this dissertation research to discuss policy recommendations and the contributions these studies provide to the gender and development discourse, social movement frame analyses, the use of feminist quantitative methods. Finally, I conclude with a discussion regarding the theoretical, methodological, and most importantly, the practical implications on women’s lives – when claiming women’s empowerment in the contested discourse. 10 REFERENCES Alkire, Sabina, Ruth Meinzen-Dick, Amber Peterman, Agnes Quisumbing, Greg Seymour, and Ana Vaz. 2013. The Women’s Empowerment in Agriculture Index.” World Development 52:71–91. Batliwala, Srilatha. 2007. "Taking the Power Out of Empowerment–An Experiential Account." Development in Practice 17(4-5):557-565. Batliwala, Srilatha, and Deepa Dhanraj. 2007. "Gender Myths that Instrumentalize Women: A View from the Indian Front Line." Pp. 21-34 in Feminisms in Development: Contradictions, Contestations and Challenges, edited by Andrea Cornwall and Ann Whitehead. Zedbooks. Benford Robert D. and David A. Snow. 2000. “Framing Processes and Social Movements: An Overview and Assessment” Annual Review Sociology 26:611-39. Calvès Anne-Emmanuèle. 2009. “Empowerment: Généalogie d'un Concept clé du Discours Contemporain sur le Développement.” Revue Tiers Monde 200:735-749. Retrieved July 28, 2021: https://www.cairn.info/revue-tiers-monde-2009-4-page-735.htm Cornwall, Andrea, and Ann Whitehead. 2007. Feminisms in Development: Contradictions, Contestations and Challenges. Zed Books. DeVault, Marjorie L. 1996. “Talking Back to Sociology: Distinctive Contributions of Feminist Methodology.” Annual Review of Sociology 22:29-50. Ferguson, Lucy. 2010. “Interrogating ‘Gender’ in Development Policy and Practice: The World Bank, Tourism and Microenterprise in Honduras.” International Feminist Journal of Politics 12(1):3-24. Ibrahim, Solava, and Sabina Alkire. 2007. "Agency and Empowerment: A Proposal for Internationally Comparable Indicators." Oxford Development Studies 35(4):379-403. Kabeer, Naila. 1999. “Resources, Agency, Achievements: Reflections on the Measurement of Women’s Empowerment.” Development and Change 30(3):435–464. Kumar, Neha, Kalyani Raghunathan, Alejandra Arrieta, Amir Jilani, and Shinjini Pandey. 2021. "The Power of the Collective Empowers Women: Evidence From Self-help Groups in India." World Development 146:105579. Malapit, Hazel, Agnes Quisumbing, Ruth Meinzen-Dick, Greg Seymour, Elena M. Martinez, Jessica Heckert, Deborah Rubin, and the Gender Agriculture Assets Project Phase 2 (GAAP2) Study Team. 2019. "Development of the Project-level Women’s Empowerment in Agriculture Index (pro-WEAI)." World Development 122:675-692. 11 Ministry of Rural Development. 2023. “SHGs Profile Entry Status” Ministry of Rural Development, Government of India. Retrieved July 4, 2023: https://nrlm.gov.in/shgReport.do?methodName=showIntensiveStateWiseReport Mooney, Patrick H., and Scott A. Hunt. 2009. “Food Security: The Elaboration of Contested Claims to a Consensus Frame” Rural Sociology 74(4):469-497. Narayan, Deepa. 2002. Empowerment and Poverty Reduction: A Sourcebook. Washington, DC: The World Bank. Nichols, Carly E. 2021. "Self-help Groups as Platforms for Development: The Role of Social Capital." World Development 146:105575. Priya, Preeti, Anand Venkatesh, and Anshu Shukla. 2021. "Two Decades of Theorising and Measuring Women's Empowerment: Literature Review and Future Research Agenda." Women's Studies International Forum 87: 102495. Rowlands, Jo. 1997. Questioning Empowerment. Oxford: Oxfam. Sangwan, Richa, and Yashna Bawa. 2022. “Government Schemes for Women Empowerment” Proceedings of DGHE, Haryana Approved National Seminar on Gender Sensitive Issues and Women Empowerment. Retrieved May 2023: https://www.sdcollegeambala.ac.in/wp- content/uploads/2022/12/Gender-Issue-2022-P-37.pdf Sen, Gita, and Caren Grown. 1987. Development, Crises, and Alternative Visions: Third Word Women’s Perspectives. London: Earthscan Publications Snow, David A. and Robert D. Benford. 1988. “Ideology, Frame Resonance, and Participant Mobilization” International Social Movement Research 1:197-217. __________ 1992. “Master Frames and Cycles of Protest.” Pp. 133-155 in Frontiers in Social Movement Theory, edited by Aldon D. Morris and Carol McClurg Mueller. New Haven, CT: Tale University Press. Wright, Wynne and Taylor Reid. 2010. “Green Dreams or Pipe Dreams?: Media Framing of the U.S. Biofuels Movement” Biomass and Bioenergy 35(4):1-10. 12 CHAPTER 2: CONCEPTUALIZING EMPOWERMENT FOR WOMEN IN THE GLOBAL SOUTH: A FRAME ANALYSIS OF CONTESTED CLAIMS IN THE DEVELOPMENT DISCOURSE Introduction This study aims to address the most contested concept in the contemporary development discourse, women’s empowerment (Cornwall and Whitehead 2007). While empowerment is no longer a novel idea, its contestation by feminists, development practitioners, and governmental bodies, have given rise to several and sometimes conflicting claims about who empowerment is for, how women are disempowered, and what should be done to empower them. As a result, individuals and institutions today research, promote, and claim “empowerment,” but mean vastly different things. Understanding how empowerment frames vary is important because these frames ultimately inform policy and programs that have serious consequences on women’s lives. The contestation over empowerment’s meaning is not new; in fact, it has been acknowledged by feminists for decades. However, the recognition is often found in the literature review of empowerment studies showcasing the different options by which to frame their research or propose a new frame (Alsop et al. 2006; Rowlands 1997). Even among the most cited “comprehensive reviews” of empowerment are limited in either their analytical rigor to empirically identify the shared and/or contested grievances over its meaning (Ibrahim and Alkire 2007; Oxaal and Baden 1997) or the sample from which they draw from (Priya et al. 2021; Richardson 2018). As a result, scholars and organizations are left with a fear of adopting the “wrong” empowerment frame without any reference to a diverse and systematic study that showcases the defining features that lead to varying outcomes in terms its “successfulness” at empowering women. 13 The present study attempts to provide this need, which characterizes the theoretical gap in the empowerment and development discourse today. I find the frame analytical approach is particularly useful for providing a systematic review of empowerment frames because rather than treating frames as a given, from the frame analytical perspective, frames are understood to be the outcome of social interactions over a shared interpretation of social life and situations. In particular, I draw from the frame analytical approach to examine the various empowerment frames proposed by claims-makers (i.e., gender empowerment advocates) in the development discourse as a way to assess, describe, and compare these outcomes. Further, I examine the framing tasks taken by claims-makers to describe the processes by which grievances over an interpretation of empowerment are constructed and/or contested. In doing so, I can understand and compare the ways they identify solutions (prognosis framing) within a set of problems (diagnostic framing) to achieve particular goals (motivational framing) in their empowerment frame. At the same time, I take a feminist approach to this systematic study not only through a grounded approach to coding empowerment frames but also in my attempt to create a more diverse sample that includes voices from the Global South. I do this with the recognition that the most influential literature comes primarily from the Global North; contributing to a history of powerful groups in the Global North deciding what is “empowering” for all women (Cornwall and Whitehead 2007; Rowlands 1997). Even today, popular empirical reviews of women’s empowerment tend to start their sample timeline in the early or mid-1990s (Richardson 2018). However, this starting date reflects the institutionalization of the term into the development discourse when two United Nation (UN) conferences (i.e., 1994 International Conference on Population and Development in Cairo, 1995 Fourth World Conference on Women in Beijing) 14 gave the concept international visibility (Batliwala 2007; Calvès 2009). To honor my commitment of elevating the less acknowledged empowerment frames deriving from the Global South, I start my sample timeline in 1985 to reflect the reality of Third World feminists initially introducing the concept to the discourse in the mid-1980s (Batliwala 2007; Calvès 2009). In the sections that follow, I begin by describing the sampling procedures I took to create a more diverse and representative sample to examine the various ways empowerment advocates describe the meaning of women’s empowerment in the development discourse between 1985 and today. I then discuss the analytical approach I took to analyze this data, followed by a description of the resulting coding processes that led to the formation of conceptual groupings. I discuss these conceptual groupings and their defining features as distinct empowerment frames in the findings section and conclude with a discussion of this study’s contributions to the gender and development discourse as well as areas for future research. Methodology Sample Since the purpose of this study is to identify and explain the variations in how women’s empowerment advocates in the development discourse frame empowerment, the sample for the present study only includes articles and books that operationalize empowerment. I choose this criterion over those that simply define empowerment, as comparing definitions is not the best way to study differences in frames. I caution against this because the same definition of empowerment can be “used” or borrowed by various claims-makers that sometimes end up operationalizing and measuring it in very different ways. For this reason, I sample literature that may not explicitly define empowerment but attempt to operationalize and/or measure empowerment. Moreover, the following inclusion criteria was used to assess candidate pieces for 15 this study: 1) Original piece is obtainable (i.e., citations that can only be traced back to review pieces – not the original piece – were excluded); 2) Published between 1985 and 2021; 3) Written or translated to English; 4) Original research or conceptual framework (i.e., review articles were excluded); and 5) Operationalizes women/gender/female empowerment in the Global South (i.e., empowerment studies on women in the Global North were excluded, as well as articles solely focused on agency, autonomy, or equality/equity and not empowerment). A complete list of journal articles and books on women/gender/female empowerment in the development discourse was unavailable for sampling. Thus, two sampling procedures were used to systematically create a diverse and representative sample for this study. I began this process by systematically identifying the most recognized empowerment frames in development by generating a “most cited/Global North” sub-sample using a three-phase search and screen process. In the first phase, I used Google Scholar to conduct a reverse citation search for empowerment studies from four well-known reviews on the operationalization and measurement of women’s empowerment in the development discourse.2 Results from this search process yielded 85 unique studies (i.e., study inclusion criteria one through three). The second phase involved a review of study titles and abstracts to identify those that appeared to meet study inclusion criteria four and five, yielding 63 studies. The third phase involved narrowing down the number of studies to one piece per author and by selecting the most cited and seminal pieces of empowerment literature, yielding a final sub-sample of 24 “most-cited/Global North-centric” pieces for analysis. Furthermore, as one of the research goals of this study is to include and highlight empowerment frames from authors/feminists from the Global South, a second sampling 2 Pieces used in phase one: Alsop et al. 2006; Ibrahim and Alkire 2007; Oxaal and Baden 1997; Richardson 2018. 16 procedure was performed using a different three-phase search and screen process. To begin systematically generating this second sub-sample, I first created a list of feminists from the Global South using multiple sources; yielding 62 unique feminists that were born and/or currently live in the Global South. The second phase involved searching each feminist’s name using Google Scholar and reviewing titles and abstracts to determine if they published literature pertaining to women/gender/female empowerment (within the study inclusion criteria); yielding 38 feminists’ pieces of literature. The final phase involved full review of each study, cutting literature that was mainly reviews or did not elaborate on what empowerment meant in their study, yielding a total of 30 Global South feminists’ studies. I combine these 30 pieces with the 24 “most-cited/Global North-centric” pieces, for a total of 54 pieces of literature to submit to frame analysis.3 Analytical Approach The purpose of this study is to not only provide a comprehensive review of the gender empowerment literature, but to conduct rigorous analyses to reveal the differing ways empowerment is framed for women in the Global South. I do this by answering two research questions. The first research question asks: How do women’s empowerment advocates (e.g., feminists/scholars and organizations) frame empowerment? To answer this question, I use frame analysis to guide my qualitative coding. Frame analysis is the best approach to answering this question because it prevents researchers from treating meanings as a given, or part of a social structure detached from human agency (Wright and Reid 2010). From this approach, frames are understood to function as articulation mechanisms of tying together various elements of reality so that one set of meaning is conveyed 3 The final sample is listed in Appendix A. For more detail on the sampling process, see Appendix B. 17 rather than another (Snow 2004). In short, frames represent one story, and framing uncovers the process of why that one story is told instead of another. I therefore use frame analysis as a tool to understand how women’s empowerment advocates (scholars/feminists/organizations) articulate and promote particular frames and solutions of women’s empowerment and exclude alternatives. Specifically, I use the three framing tasks (i.e., diagnostic, prognostic, and motivational), as proposed by Snow and Bedford (1988), to guide how I am coding, to better understand how empowerment is framed. Yet, I use a grounded approach to allow the inductive formation of contextual codes. In other words, the contextual codes came from the text, to understand what the prognostic, diagnostic, and motivational frames are. I then use these codes to conduct analyses of all 54 pieces of literature in the sample. To clarify, the first framing task, diagnostic framing, serves to identify a problematic aspect of social life along with attribution of blame or causality to an actor or structure (Benford and Snow 2000; Snow and Benford 1988; Wright and Reid 2010). This framing task asks the questions “what is the larger social problem?” (problem identification consensus), and “what or who causes the social problem?” or “what is the source of the problem?” (attributional consensus). Applicably, I inductively code for the specific problem and focus of blame that is identified in each piece of literature. The second function, prognostic framing, proposes a solution to the diagnosed problem and specifies the strategies, tactics, and targets for remedying the condition (Snow and Benford 1988; Wright and Reid 2010). Because this framing task is concerned with “the range of possible “reasonable” solutions and strategies advocated” (Benford and Snow 2000:616), I inductively code for solutions or guidelines to empowerment, often times including what empowerment 18 means in practice.4 I also include deductive codes regarding the clarity of the solution. I do this because Snow and Benford (1988) suggest that some frames are unsuccessful at generating change because attention may be so focused on the diagnostic framing that prognostic considerations are neglected, leading to unclear guidelines for action. In other words, even when consensus around the nature and causes of a social problem is reached, it will do little to mobilize action if the answers to questions regarding what should be done (i.e., prognostic framing) and why (i.e., motivational framing) are left ambiguous. In addition, because the prognostic frame includes the targets for remedying the condition, I also code for who is being “empowered.” The last framing function, motivational framing, provides the motivational impetus for individuals to participate, in an effort toward redress. It provides a “call to arms” or rationale for engaging in ameliorative or corrective collective action that goes beyond the diagnosis and prognosis (Benford and Snow 2000; Snow and Benford 1988; Wright and Reid 2010). In the context of women’s empowerment, I inductively create motivational framing codes by asking the question “why does this issue matter?” To get at these motivations, I specifically add codes that provide answers to the question “what is the goal of women’s empowerment?” or “why is empowering women important?” Consequently, these motivational codes help to better understand who the solution is (really) for. In addition to coding for the motivational frames, I also include codes regarding who can participate in solving the issue. I do this because Snow and Benford (1988) suggest that one reason action mobilization is unsuccessful is because consensus mobilization (i.e., problem and solution) is framed in a way that shudders participation. 4 I coded all segments pertaining to the author’s empowerment framework, not just the definition. I do this because some authors do not provide a definition, the same definition can be operationalized differently, and I found that sometimes the definition the author offers does not align with what they count as “empowering” in practice. 19 Furthermore, the second research question of this study asks: What factors influence the interpretation of how successful a solution (e.g., development program, policy) is at empowering women in the Global South? In other words, I am interested in uncovering what solutions (prognoses/indicators) are considered “empowering” and why. Answering this question will provide a better understanding of the defining features that make empowerment frames produce varying conclusions regarding whether a solution is successful or not at empowering women. To understand these defining features, I borrow Benford and Snow’s (2000) concept of frame elaboration (also known as frame amplification or punctuation) to analytically group pieces of literature - into collective action frames of empowerment - based on their similarities and differences. I do this as frame elaboration is a discursive mechanism through which collective action frames become modified and generated (Benford and Snow 2000). Moreover, I found that the pieces in my sample were mentioning the same diagnoses and prognoses to the point that there was so much cross over it was hard to find definitive groupings constituting them as “collective action frames.” However, I knew that two pieces sharing the same codes could be drastically different. Benford and Snow’s (2000:624) concept of frame elaboration is helpful in this case, as they state: “The key to understanding the evolution of frames resides in the articulation and amplification processes rather than in the topics or issues comprising the frames.” In other words, grouping the empowerment literature based on differences in topics was not surprising because most frames do not differ based on diagnoses or prognoses mentioned, but the way they are discussed. Specifically, frame elaboration involves emphasizing and focusing on some topics, issues, and ideas more than others, making them more salient in a hierarchy of movement-related topics (Snow et al. 2019). 20 While Benford and Snow (2000) mention that this kind of analysis is highly labor intensive and therefore less researched, most collective action frames vary in this way. It is clear in this case of women’s empowerment that advocates identify similar topics, but the space and rationale they give to certain issues and solutions vary. I therefore find frame elaboration helpful in understanding the different collective action frames of empowerment, which lend in identifying factors causing differing interpretations of how successful a solution is at empowering women in the Global South. Coding Empowerment for Whom? Instrumental versus Intrinsic Goals Snow and colleagues (1986:469) suggest that one variety of frame amplification is value amplification, referring to the goals or end states that movements seek to attain or promote. Following this recommendation, I chose to first divide the literature into two groups based on their end goal of empowerment: one representing empowerment as an instrumental goal (i.e., empowerment is a means to an end), and the other representing empowerment as an intrinsic goal (empowerment is the end goal itself).5 Specifically, literature coded with the motivational code “for development” were categorized into the instrumental group and all others were categorized in the intrinsic group. I did this because those coded with “for development” were used in cases where women’s empowerment was explicitly framed for development, as well as in cases where authors’ framed women’s empowerment as a goal “in itself” but also suggested it was instrumental to other goals. 5 Theoretically, this is a good starting point because many of the leading empowerment advocates make it a point to highlight the tradeoffs of framing empowerment as an intrinsic or instrumental goal: In other words, deciding if empowerment is a means to an end, an end in itself, or both (Ibrahim and Alkire 2007; Kabeer 1999; Narayan 2005). 21 Types of Feminism: Practical and Strategic Goals While most of the codes in this study emerge from a grounded approach, the codes related to intrinsic motivational codes are the only caveat. I elaborate on Moser’s (1989) binary distinction between women’s practical and strategic goals, to account for the multiplicity of feminist goals in empowerment frames, by creating five intrinsic codes. The first two codes account for practical goals, including emic value (i.e., valuing the needs of the women in question) and if the author (indirectly) identifies a patriarchal bargain as an example of women's empowerment. The other three codes account for the three ways strategic goals of empowerment can be defined by authors including: 1) “Alternatives for women” which focuses on creating alternatives for women that allow them not to be oppressed by structures that once oppressed them, but not necessarily with the goal of living more equitability with men; 2) “Alternatives for women to live more equitably,” however, identifies alternatives that challenge gender inequality in ways that allow women to live more equitably with men, sometimes by transforming gender inequality through redistribution, but always within current social institutions; and lastly, 3) “Alternative organization/relationship between men and women” represents the most radical of the three strategic goals in its overt rejection to the status quo and identification of breaking down structures that (re)create gender inequality with the goal that it will lead to a complete transformation of gender organization and relationships between men and women.6 Collective Action Frames of Empowerment As I mentioned, many of the diagnostic, prognostic, and motivational codes are found across the study’s sample, creating significant overlap between pieces in terms of the codes applied. As a result, I implement frame elaboration to understand potential groupings of 6 For more detail on practical and strategic goal codes, see Appendix C 22 collective action frames of empowerment. Snow and colleagues (2019) suggest studying frame elaboration/amplification by seeing how much of a topic is given a certain amount of space compared to others. To do this, I use the Code Matrix Browser in MAXQDA and run separate analyses for each grouping of codes (diagnostics, prognostics, and the various motivational code groups) to create a heat map; allowing me to see which codes are used most often in each piece. While the heat maps are helpful in seeing the amplification of certain codes, they are insufficient to identify which pieces of literature align to create subsequent groupings. To create these groupings, I initially grouped literature together based on their most defining features that were identified in the coding process. I mapped these groupings into a concept map and then used the heat maps to triangulate the analyses of my groupings. The final concept map is shown in Figure 1.1 on the next page and the heat maps of each of the literature groupings can be found in Appendix D. Each of the groupings created in this process represent collective action frames of empowerment (grey squares in Figure 1.1); as the pieces of literature making up each collective action frame align in the way they amplify certain issues and solutions over others. Further, each of the collective action frames are derivative of a more generic master frame. Master frames perform the same functions as collective action frames but on a larger scale. In a sense, master frames are to collective action frames as paradigms are to theories (Snow and Benford 1992:138). The three master frames of empowerment include: 1) Power; 2) Equal participation and rights; and 3) Agency (colored squares in Figure 1.1). 23 Figure 1.1. Concept map of collective action frames and master frames of women’s empowerment Notes: The three colored boxes above represent master frames, and the thirteen grey boxes represent unique collective action frames. The three grey boxes with a bolded black border are frames that all other collective action frames within the master frame stem from. Citations in blue font indicate pieces with instrumental framing; those in purple font indicate intrinsic framing. Bolded citations represent pieces from the Global South feminist sub-sample; non-bolded ones are those from the most-cited/Global North-centric sub-sample. Lastly, citations with (PB) indicate that a patriarchal bargain was identified in the piece. 24 Findings My analysis reveals that the phrase “women’s empowerment,” in the context of development, is used quite differently across empowerment advocates. Specifically, I find 16 collective action frames of empowerment; each made up of 1 - 11 pieces of literature. Moreover, I contend three master frames provide women’s empowerment advocates the ideational boundaries to form deriving collective action frames by elaborating on its basic problem-solving schema (or amplified prognostic). In particular, I show how four collective action frames center around the problem-solving schema of “power,” five collective action frames derive from the master frame “equal participation and rights,” and seven collective action frames share a common focus on the prognostic “agency.” Thus, amplified prognostics (i.e., power, equal participation and rights, or agency) not only draw collective action frames together within a master frame, they also define the differences between master frames. I also reveal collective action frames differ in the way they elaborate on their master frame’s prognostics (prognostic frame) to particular end goals of empowerment (motivational frame) within a set of amplified problems (diagnostic frame). I present these findings in three major sections, beginning with the “Power” empowerment frames (consisting of four collective action frames and ten pieces of literature) because the earliest concepts on empowerment for women in the Global South fall under this frame. I then turn to the “Equal Participation and Rights” empowerment frames, consisting of five collective action frames and twenty-five pieces of literature, including pieces credited for institutionalizing empowerment in the development discourse. I then conclude my findings by describing the “Agency” empowerment frames, consisting of seven collective action frames and nineteen pieces of literature. Each of the three sections opens with a description of the master 25 frame and a summary comparison of their collective action frames. Following each overview, I describe each deriving collective action frame in terms of its diagnostic, prognostic, and motivational framing tasks. In doing so, I reveal how each collective action frame tells a unique “story” about how women’s empowerment is achieved within a set of problems, and what “successful empowerment” looks like, given the frame’s amplified framing tasks. Master Frame 1: Power Third World feminist movements started spreading the idea of women’s empowerment in development in the 1980s, with the seminal piece Development, Crises, and Alternative Visions – written by Gita Sen and Caren Grown with the guidance of feminists and activists involved in DAWN (Development Alternatives with Women for a New Era). The piece offered the first master frame of empowerment, allowing numerous groups to elaborate their grievances in terms of its basic problem-solving schema: power. For all four collective action frames deriving from the master frame “Power,” questioning the unequal power dynamics that create inequality in the first place is at the heart of empowerment. All four collective action frames (and ten pieces of literature) frame empowerment as an intrinsic goal and identify the feminist goal of an alternative gendered organization as an important end goal of empowerment. Notwithstanding, these collective action frames vary from one another in terms of what constitutes power, how it is operated, and thus the appropriate prognostics of how empowerment should happen. To understand this variance between the four collective action frames that fall under the “Power” master frame, I establish six codes accounting for the various ways power is conceptualized (see Appendix E). In the following discussions, I present how these four collective action frames vary from taking just one to all conceptualizations of power. In addition, what I find is: the more notions of power that are identified and explained, the clearer the solutions become. In the same 26 vein, when power is only conceptualized as taking one of these forms, the prognostics become more unclear and only accessible to “experts.” Power-to Heavy This collective action frame consists of two pieces of literature (Antrobus 1989; Rowlands 1997), that aim to empower “Third World” women from sources of power inequality (i.e., systems of domination and subordination). While systems of oppression including imperialism, colonialism, and neo-colonialism are identified as sources of power inequality; this frame particularly highlights aspects of patriarchy and culture in generating zero-sum models of power that create material (overt) and psychological (internalized) barriers to women’s exercise of power in comparison to men. Specifically, the literature espousing this collective action frame discusses these diagnostic codes together in a way that resembles hegemonic masculinity. Hegemonic masculinity refers to a type of man that embodies the most culturally celebrated masculine traits idealized by men and women in that society. All other men and women are measured against this; creating not just a gender hierarchy that privileges male control and superiority, but maintains an ideology that justifies women’s subordination and domination. While neither author explicitly names “hegemonic masculinity” as the problem, they both point to gendered ideologies and values as the major component to women’s exploitation and subordination to men – ideas that men and women alike endorse and perpetuate. For example, in her Honduras case study, Rowlands (1997:34) identifies a specific form of masculinity that defines men’s superiority in society – namely, machismo, and to a lesser extent caudillismo (meaning “strong male leader”): “The social and cultural ethos of Honduras, heavily dominated by machismo, reinforces the subordinate position of women.” 27 This “Power-to Heavy” collective action frame gets its name by highlighting generative forms of power; where empowerment is achieved through power processes where one’s increase in power does not diminish another’s. Further, these forms of power are interconnected: it is understood that empowerment begins with “power within,” by undoing internalized forms of “power over” through critical consciousness, leading to feelings of agency and dignity as well as the ability to identify sources of “power over” that need to be radically challenged. Further, “power to” is intrinsically linked to “power with” and “power within.” The ultimate feminist goal for these frames is an alternative organization between men and women. In theory, this means empowerment processes require less “power over” relationships, and more generative forms. This frame acknowledges that this kind of power transformation cannot happen without those with privilege. For Antrobus (1989), she says for empowerment to happen, critical consciousness is not just something the women in question need to do, but something those with privilege need to engage in as well. For Rowlands (1997), she suggests that personal and collective empowerment can happen without others, yet when it comes to practicing empowerment within close relationships, it is the most difficult area of change for women and requires men changing their attitude and behavior. She says when men change, the possibilities for more generative forms of power begin to flourish: “If men become less willing to use 'power over' in their relationships with women, whether in the home, in the workplace or in the community, it not only directly reduces the obstacles women face in their own empowerment process, but also helps to create an environment where the use of power in other forms becomes more possible” (Rowlands 1997:132). 28 Only Power-to This next collective action frame also consists of two pieces of literature (Corrêa 2010; Mendoza 2019) but is more extreme than the previous frame in its call for “power to,” because it is the only conceptualization of power used. This conceptualization of power borrows from a Foucauldian understanding that informs their feminist goal of creating an alternative gendered organization - so much so, that it takes the most radical form of creating a society that aims to diminish gender altogether. This frame’s feminist goal is aligned with the frame’s diagnostic: while it is concerned with all forms of power inequality that oppress women, as well as power imbalances between women along lines of race, class, sexuality, and place, it identifies Eurocentric conceptualizations of the gender binary as the source of all power inequalities. For example, Mendoza (2019) identifies colonialism and modernity as the source power inequality (in terms of gender, sexuality, and race) in dehumanizing women and indigenous men, because before colonial intrusion these kinds of inequalities did not exist in indigenous societies. Across all three master frames, this collective action frame is the most explicit about understanding gender as a continuum. The diagnostic of the problem leads to the frame’s ultimate goal of mimicking models of a genderless society before colonialism. It therefore views “reversing hierarchies” as an ineffectual prognostic. For example, Corrêa (2010:186) is explicit about the need of injecting a Foucauldian conceptualization of power when it comes to framing empowerment and in doing so, hints at the fact she does not see patriarchy as an institution that creates binary outcomes of “men with power” and “women with no power,” like the previous frame. Similarly, Mendoza’s (2019:6) conceptualization of power is tied directly to the goal of an alternative gendered society, based on creating horizonal relations between men and women: 29 “In this sense, it would seem that there is a consensus around the epistemic and political value of the theories of African and Indian feminists who discuss the existence of gender in the pre-intrusive world. These theories show us the possibility of creating models of relations between men and women based on horizontality and show us other ways of inhabiting bodies, of exercising sexuality, but also, beyond the reorganization of social relations between men and women, the possible reorganization of the economy, politics, spirituality, etc.” While this genderless aim may be the most radical in comparison to all other collective action frames, it is also the most unclear frame in describing how this can happen and who can be a part of this change. Not only is this process unclear, but this collective action frame is left in the theoretical for other experts (e.g., post and decolonial feminists) to tackle. For example, Corrêa (2010) begins suggesting it is up to the individuals in question (i.e., those who need empowering) to do the empowering themselves. However, she concludes her article by addressing a “we” – specifically a “we” that includes those engaged in theorizing empowerment to participate in better understanding empowerment and its blind spots. Similarly, and ironically (as a decolonial feminist), Mendoza keeps this frame available only to those in her field. Even though the article was originally published in Spanish and translated to English, the high-level academic language used throughout (e.g., “coloniality of gender,” “epistemology”) keeps this collective action frame in the theoretical by making it exclusively an issue accessible to those familiar with decolonial theory. Power-over and Power-to So far, the power-centered collective action frames identified “power to” as the center of the prognostic for an alternative gendered organization. This frame, however, includes the most 30 conceptualizations of power. Specifically, this frame recognizes the need to have “power over” and strategic goals of alternatives for women to live more equitably, for forms of “power-to” to form and for the ultimate strategic goal of creating alternative relations between men and women to be successful. This collective action frame I refer to as “Power-over and Power-to,” includes three pieces of literature (Sen and Grown 1987; Bhatt 1989; Jain and Chacko 2008), including Development, Crises, and Alternative Visions. The logic of the diagnostic frame is that colonial rule created - and post-colonial development (e.g., neoliberalism and globalization) exacerbated - power inequalities between nations, classes, genders, and ethnic groups. For women in particular, power inequalities are reinforced by gender relations that manifest in violence against women and forms of gender discrimination (including gendered division of labor) that value women less. Most importantly, this frame underscores the need to link all forms of discrimination. While Sen and Grown (1987:9) were the first to state this, Jain and Chacko (2008:13) reiterate these sentiments well: “Gender discrimination has always operated through other forms of exclusion and oppression: colonialism, caste, class, race, region, religion, and so forth. So gender relations operate through these other social and institutional structures, determining their status not only vis-a-vis men and women of their own social group, but also men and women of other groups.” Similar to the frames previously discussed, an alternative gender organization is the ultimate strategic goal of this frame, with “power to” as the defining prognostic frame. For example, Sen and Grown (1987) explain that alternative gender organization is not going to happen by mimicking “power over” models, but by creating a new model of power that stems from “Third World” women’s experiences (i.e., emic-value) within systems of oppression. At the 31 same time, this frame is unlike the others emphasizing “power to”, as it is one of the few that does not view “power to” with rose-colored glasses. While “power to” is the goal, the authors do not fool themselves in believing it is easy nor a solution free of its own set of problems.7 Further, as Sen and Grown (1987:79) suggest, a variety of methods become appropriate given the diversity of women and their emic issues. Thus, this frame is unique in recognizing prognostics of “more power over” and women’s rights as steppingstones to the ultimate feminist goal. Specifically, it recognizes that women’s more “power over” and equality with men in established institutions opens the possibilities for new gender relations to form. In addition, cultural change is identified as a critical component to empowerment and an alternative organization between men and women. Specifically, this takes the shape of a “just” social order: one that is equitable and humane (life-affirming for all people and holds people liable). In creating this alternative organization (and culture), internalized oppression is not framed as a personal problem women are responsible for undoing, but a process that includes people with power. In other words, empowerment includes processes of cultural change that hold those with relative power and privilege accountable for systems of oppression and undoing them. Bhatt (1988:1064) explains the importance of making policy makers aware of women’s problems and perspectives so they can act in their favor. She refers to this process as “political visibility” and says that this will help lead to a society that is more just and humane. Similarly, Sen and Grown (1987:88) state, “The level of awareness about women's subordination has to be raised through popular culture, the media, and formal and informal education …Governments must be pressured to give us a greater voice in radio, T.V., film, and other mass media, and to generate more 7 Sen and Grown (1987) lay out multiple practical issues that arise in the process of forming “power to” that arise in searching for non-hierarchal structures including difficulty in forging relationships with organizations and weakened voices in defining public policies, because no one is authorized to speak due to unclarity in authority. 32 funding for such programs …We already know that research into our history, networking among scholars, and curriculum development are vital aids to raising our own consciousness, as well as that of men.” Zero-sum Power This last power-focused collective frame is distinct from the other power-focused collective action frames in two ways: 1) All three pieces making up this frame (Batliwala 1994; Samuel 2001; Raju 2005) conceptualize power (as a diagnostic and prognostic) as zero-sum, and 2) It gives equal attention to the two strategic goals of alternatives for women to live more equitably and alternative relations between men and women. To begin, this frame’s prognostic and strategic goals are directly aligned with the diagnostic frame: power is understood as those who have control over (or the ability to influence the distribution of) material assets, intellectual resources, and ideology, and thus have “power over” those who do not have this kind of control or influence. Women are deemed relatively powerless in comparison to men, given this definition, where the underlying ideology that makes this so is patriarchy. Correspondingly, the goals of empowerment include enabling women to gain access and control over resources, challenging patriarchal ideology, and transforming structures that preserve power inequality. Because power is conceptualized as zero-sum, “more power over” is the dominant prognostic. Further, empowerment is framed as a process of shifting political, social, and economic power between and across both individuals and social groups. Specifically, this means redistributing these forms of power from more privileged people to those who are disempowered. By framing empowerment processes as redistribution, this frame does not sugarcoat the fact that women’s empowerment inevitably involves men’s loss of power. Because patriarchal ideologies and structures create and preserve men’s power and women’s subordination, and empowerment 33 involves challenging these ideologies and structures of power inequality, successful empowerment ultimately necessitates men’s loss of power. However, men’s loss of power in this process is not framed as “disempowerment” but rather results in men’s own liberation and empowerment. At the same time, it is instructed that once women gain control and power, that they do not use it in the same exploitative and corrupt way as it is used in hierarchical, male- dominated societies (Batliwala 1994). At the heart of this collective action frame is an understanding that empowerment strategies intervene at the level of women’s “condition” (i.e., their emically defined needs) while also transforming their “position” (i.e., strategic goals). Because women’s emically defined needs are a product of their inferior status, solutions need to address the source of the problem. Thus, this frame finds women’s “more power over” to be “successful” or “genuine” empowerment when women do not use traditional gendered roles to mobilize empowerment. This is because patriarchy is the root of the problem (women’s subordination), and all prognostics and strategic goals are evaluated against understanding local patriarchal values. Put another way, unless women’s “more power over” involves questioning, disturbing, and challenging the existing patriarchal order (relations and ideology) that exists at all levels and spaces in women’s lives, it is not empowerment. For example, Samuel (2001) explains that tactics using traditional gendered norms to get more “power over” (i.e., patriarchal bargains) may be the only option in a context of extreme violence; but because these tactics do not push through oppressive gendered roles, they do not empower political decision making and fail to generate transformative results for women and society. To realize these strategic goals women must first become aware of the ideology that legitimizes their subordination. For women to demand strategic goals, it is understood to not 34 come spontaneously from the condition of being oppressed, but rather needs an external agent (that has an awareness that the social order is unjust) to prompt undoing internalized oppression with processes of raising women’s critical consciousness. In turn, women can begin to feel worthy of empowerment (and their innate rights) that will propel them to challenge patriarchal ideologies. At the same time, while empowerment is initially prompted by external agents, it is solely for the purpose of providing women with new ideas and information that will change their consciousness and encourage action (specifically for strategic change). Thus, empowerment cannot be top-down, but only possible when large numbers of women are mobilized to press for change (i.e., “power with”). Master Frame 2: Equal Participation and Rights While all of the previous frames uphold the original emphasis on notions of power in framing empowerment, the frames deriving from the master frame “Equal Participation and Rights” shifts the focus of empowerment to women’s equality in terms of participation and/or rights in comparison to men. In particular, the 1995 Beijing Platform for Action (the Fourth World Conference on Women) serves as the master frame from which all five collective action frames derive. In this analysis, the document is grouped with four other pieces that share the same accentuated parts of empowerment. In the following discussions, I will show the way two of the collective action frames deriving from this master frame emphasize equality and participation, and the ways the other two derive primarily from the aspects of women’s rights as human rights. Human Rights & Gender Mainstreaming Gender mainstreaming was the overall agreed and adopted strategy to support the global goal of gender equality in the 1995 Beijing Platform for Action (Mukhopadhyay 2004). The goal 35 of gender mainstreaming has been said to bring women’s issues to the center of development agendas (Mohanty and Miraglia 2012). The document is also explicitly an agenda for women’s empowerment, which is never defined but quickly becomes intertwined with the goal of equality between men and women. Specifically, the defining properties for all five pieces making up this collective action frame (UN 1996; Moghadam and Senftova; Chattier and Slatter; Yiping 2013; Rodríguez 2017) is the strategic goal of gender equality and the prognostic goal (to reach this) is women’s equal participation and rights in relation to men. This prognostic is twofold: the focus on women’s participation and women’s rights as human rights. It stems from the first few pages of the 1995 Beijing Platform for Action (UN 1996:1), where it states: “Women’s empowerment and their full participation on the basis of equality in all spheres of society, including participation in the decision-making process and access to power, are fundamental for the achievement of equality, development and peace;” and further, “Women’s rights are human rights.” The conflation between gender equality and empowerment is apparent in other pieces of this collective action frame. For example, Moghadam and Senftova (2005:290) adopt this idea so literally that it becomes the defining goal of empowerment: “The goal, then, is gender equality, or – to use the call from the Beijing Conference – women’s empowerment in all its dimensions.” However, as I mentioned, the defining prognostic is human rights, but it is not the only prognostic covered in the Beijing Conference. While the Platform for Action “is an agenda for women’s empowerment” (UN 1996:7), no actual definition is provided in the entire document: leading to certain countries to agree with some parts of the Platform and have reservations about others; highlighting the obvious lack of clarity regarding which parts of “empowerment” are non-negotiable. Throughout the document (UN 1996), empowerment is combined and used 36 synonymously with prognostics of the advancement of women, full participation, full realization of all human rights and fundamental freedoms of all women, equality between women and men, autonomy of women, and women’s social, economic, and political status. Thus, because of its many meanings and connections, the concept of empowerment becomes so grand, the meaning and prognostics are diluted in the end. In line with the goal of gender equality, this frame identifies gender discrimination as the problem to be fixed (i.e., diagnostic) and mainly “non-discrimination” as a means (i.e., prognostic) to both empowerment and the ultimate goal of gender equality. For instance, “Non- discriminatory education benefits both girls and boys and thus ultimately contributes to more equal relationships between women and men” (UN 1996:26). Governing structures are identified as the main means by which empowerment can happen. The difference between this collective action frame and the other gender mainstream collective frame (i.e., the following frame, “Gender Parity”), is the recognition of the need for both bottom-up and top-down approaches. However, it is not explicitly directed which needs to come first: women’s emic value or governing structures in defining the goals of empowerment and obtaining gender equality. Gender Parity This collective action frame mimics gender mainstreaming without the heavy focus on rights. It also represents the most “popular” collective action frame in terms having the most (11) pieces of literature making up the frame across all other frames (UNDP 1995; Banerjee and Roy 2004; Lokshin and Ravallion 2005; Grootaert 2005; Wedderburn 2006; Barroso and Sippel 2011; Alkire, Meinzen-Dick, Peterman, Quisumbing, Seymour, and Vaz 2013; Sen and Mukherjee 2014; Sharma 2015; Shettar 2015; Mahata, Kumar, and Rai 2017). This frame starts by conceptualizing the diagnostic frame as gender inequality; yet discusses and operationalizes it as 37 mainly a problem of gender discrimination and measures it as a problem of gender gaps using gender-disaggregated data. In turn, the aligned prognostic of gender inequality is not gender equality in its most feminist and radical form, but rather gender parity (i.e., proportionate representation of men and women in a given group). Further, the majority of the literature making up this frame understands empowerment to be an instrumental goal for gender parity and development. The most famous examples included in this frame include the UNDP’s (United Nations Development Programme) Human Development Report (1995) describing the Gender Empowerment Measure and Alkire and colleagues’ (2013) Women’s Empowerment in Agriculture Index. For this frame, the core of the problem is essentially framed as women’s lack of opportunities and resources compared to men in many areas of life: some authors include dimensions of literacy, education, and jobs, but the most popular dimensions are economic and political gender gaps captured by disparities in pay and political seats (UNDP 1995). In addition, some of the authors use sex-ratios either as an indicator of gender inequality or gender gap and efforts are made to identify the factors that lead to the imbalance or balance of the sexes (Sharma 2015; Mahata et al. 2017). Further, gender gaps are described as symptoms of gender discrimination, which leads many to use quantitative data in understanding gender discrimination (and in effect measuring gender gaps). For example, Sen and Mukherjee (2014) state, “One symptom of discrimination is persistent gender wage gaps where men systematically earn more than women in all countries and all sectors.” This frame’s prognostic is aligned with their diagnostic. Specifically, with gender discrimination and gender gaps as the focus of the problem women face, this collective action frame ends up conflating equal opportunities (or gender parity) with women’s empowerment 38 (UNDP 1995; Sharma 2015). To achieve empowerment, or rather gender parity, the frame favors solutions that remove discrimination, often reflecting positive discrimination in favor of women, in closing the gender gap and reaching gender parity (mainly in terms of increasing resources and participation). Some authors explicitly use this terminology while others imply positive discrimination as a prognostic for women’s empowerment. For example, Sen and Mukherjee (2014:9) suggest, “Women’s freedoms and agency can be promoted by tilting the unequal distribution of resources – assets, rules and norms – in favor of women so that they can exercise a much wider set of choices and autonomy.” Since removing gender discrimination is the center of the prognostic frame, state government and non-governmental organizations (NGOs) are identified as playing the most important role in making sure women become “empowered.” This point is compounded by some authors’ suggestion that empowerment is not necessarily a grassroots or bottom-up activity, but one in which the state is a key actor (Grootaert 2005). Further, because reducing disparities between men and women is the prognostic of this empowerment frame, “success” of a solution is based on if it has given women more resources and opportunities to be on par with men. For example, Shettar (2015) explains this need to be on par with men for the purpose of women elevating their status in society. Participation + Emic = Empowering This next collective action frame consists of four pieces (Malena and Heinrich 2005; Jad 2010; Bieri, Ott, Frey, Cross, Partenio, and Fernández Álvarez 2011; Reid, Reddock, and Nickenig 2014) centered around participation and women’s emic needs, to the point where if these two factors are considered, programs are deemed successful at empowering women. Many times, this conflation occurs because of a perceived identification that “power with” or what is 39 often termed “civil society” is the space where empowerment is developed. This is perfectly described by Malena and Heinrich (2005:342), “In this sense, civil society can be understood as the principal societal space where the empowerment of citizens is developed and practiced (especially poor people and other marginalized groups), that is, where citizens may participate in, negotiate with, influence, control, and hold accountable institutions that affect their lives.” Further, there is an understanding that empowerment is essentially about having women participate not only to fix power imbalances but to be heard in decision-making spaces that impact their lives. This collective action frame therefore often resembles community- participatory-action-research. Many pieces making up this collective action frame reference this practice as embodying empowerment. For example, Reid and colleagues (2014:261) state, “Community-based research is increasingly recognized as a viable approach to developing relationships with communities to address complex public health problems. Rationales for using this approach include the opportunity for researchers and community members to learn from one another, address power imbalances, empower participants, democratize knowledge, enhance the relevance of research, and connect research to larger social change efforts.” However, the irony of these frame’s highlighting women’s emic value and participation as empowerment, is that many times the problem to be fixed (identification) is decided by outsiders first and then it is up to the women participating (emic) to figure out how to implement the solutions (from the ground-up) to fix these problems. 40 Women’s Rights Starts with Emic This collective frame of empowerment, “Women’s Rights Starts with Emic,” is a descendent of the “Human Rights & Gender Mainstreaming” frame. Unlike the previous two collective action frames, this frame emphasizes both gender equality and women’s rights as human rights. Specifically, for this collective action frame, the strategic goal of empowering women is gender equality; which is often conflated with the goal of women’s rights or is at the forefront (prognostic) in obtaining empowerment. The four pieces making up this frame either operationalize empowerment by explicitly using a human rights framework (Malhotra and Schuler 2005; Mukherjee 2002), or by conflating empowerment with women’s rights (Friedmann 1992; Taylor 2010). For example, Friedmann (1992:119) explains that strategic goals require long-term struggle and should focus on women’s rights as human rights, “The ultimate goal is very clear: because women's rights form part of a constellation of basic human rights, they must be inscribed in the very heart of an alternative development.” In another example, written almost 20 years later, Taylor (2010:167) suggests that to fight for gender equality and equity, women’s human rights need to be at the forefront. In addition, women’s basic needs are discussed as part of these rights and identified as important (as a problem and something to work towards); stressing the equal importance of rights and basic needs. One way this frame sets itself apart from the previous collective action frames is with the identification of patriarchy as the ultimate problem. If other problems are mentioned, it is suggested that patriarchy is the source of these problems and transmits itself into these systems. For example, Taylor (2010:165) says, 41 “The adaptive character of patriarchy and the new and changing forms of oppression are manifest in institutionalized violence, violence in the home as well as continuing deprivations in health care, in education, in access to decent waged work and to food.” In addition, this frame is different to the previous frames in its recognition that empowerment cannot happen without the women in question realizing it is in their interests first. In other words, empowerment needs to start with women’s emic value; but as important as this first step is, it does not mean it is on women alone to make it happen. For empowerment to happen, those with power, including men and governments, have responsibility to empower women. In her concluding paragraph, Taylor (2010:167) emphasizes the role women’s emic needs must be heard as legitimate and suggests that “all of us who have the power to make a difference to the lives of the poorest women and peoples in our countries are complicit in their continued marginalization and oppression.” For Mukherjee (2002), those with power in making a difference includes men’s responsibility (in sharing responsibilities in reproductive and productive life) as well as developmental and governmental organizations. Some are more adamant about the need for the state’s role in ensuring women’s empowerment in the long run (Friedmann 1992). Malhorta and Schuler (2005) explain this collective action frame best by acknowledging the power governments and multilateral organizations have in strengthening gender equality for women, but if the claims being implemented by institutions to improve women’s position in society do not stem from women themselves (i.e., emic value; grassroots beginning), then it is good for gender equality, but it is not considered empowerment. Gender Minorities’ Rights as Freedom This frame is represented by one piece of literature (Chetty and Faleatua 2015), and while on the surface it may seem very similar to the previous frame in the shared concern for rights, it 42 differs most notably in the framing of who empowerment is for, how empowerment is realized, and takes a stronger stance on strategic rather than emic goals. These shifts present a latent shift in the overall focus from human rights/basic needs (with the goal of gender equality) to sexual and reproductive health and rights (SRHR) (with the goal of autonomy and freedom). To begin, this collective action frame, “Gender Minorities’ Rights as Freedom” is concerned not only with the rights of women but all gender minorities including girls, LGBTQI (lesbian, gay, bisexual, transgender, queer, intersex) persons, and the intersections of these gender and sexual identities. The shift in focus from women to gender minorities is tied to the frame’s diagnostic focus on identifying other cultural influences and systems of inequality besides patriarchy, including neoliberalism, heteronormativity, and religion in being responsible for violations to the right of SRHR. Second, this frame differs precisely in how empowerment is realized. Specifically, it does not consider equality in patriarchal institutions empowering. Rather than focusing on human rights and basic needs through the help of established institutions, this frame is more radical in the sense that SRHR is more than “just rights” but includes a strong emphasis on personal power (autonomy/agency) and freedom. As this frame illustrates, women and gender minorities are considered disempowered in terms of control over their own bodies including negotiating pleasure and sex, leading to a prognostic frame emphasizing right to bodily autonomy and the pessimism of empowerment occurring in patriarchal institutions. As a result of avoiding patriarchal institutions as a means to “empower” women and gender minorities, this frame favors strategic solutions for women and for an alternative gendered organization. For example, when Chetty and Faleatua (2015) discuss the topic of sex work, they frame it as work rather than as a 43 form of violence against women, leading to recommendations of decriminalization rather than legalization, as a means for sex workers to gain “formal equality.” Lastly, this frame takes a stronger stance for strategic versus emic goals, despite its additional focus on women’s personal power as constituting empowerment. This difference comes precisely from the stance that culture cannot be an excuse for not moving feminist goals (in this case gendered rights as personal freedom) forward. While the previous “Woman’s Rights Starts with Emic” frame would suggest that unless the struggle for women’s rights makes sense to the women in question it cannot lead to change, this “Gender Minorities Rights as Freedom” frame points to the diagnostic of cultural barriers to programs on sexual and reproductive health as creating a context in which it is difficult for marginalized young women and LGBTQI persons to get the information they need to fully enjoy and make decisions about their bodies. This leads to Chetty and Faleatua (2015:20) to propose the following prognostic, which entirely challenges the former women’s rights frame to empowerment: “Pacific SRHR advocates must continue to unpack arguments about culture and tradition, and not allow duty bearers to hide behind conservative, simplistic definitions of culture that restrict young people’s access to SRHR.” Master Frame 3: Agency The last master frame of empowerment does not focus on women’s rights and participation, nor power, as the focus of empowerment, but rather on agency. Some feminists suggest that one of the defining features of empowerment that separates it from other terms like “gender equality” or “women’s status” is concept of women’s agency (Malhotra and Schuler 2005; Oxaal and Baden 1997). While I have already established this is not the case for all empowerment frames, it is the case for all seven collective action frames and 19 pieces of literature that put agency at the center of their prognostic frame. In putting agency at the center, 44 many of the collective action frames that derive from this “Agency” master frame end up equating (expanded) agency as empowerment.8 However, as I show, not every agency-centered collective action frame conceptualizes empowerment the same way. Many of the following frames closely mirror one of the most famous and influential empowerment frames in the development discourse by Kabeer (1999). With a number of factors leading to varying interpretations of how successful a solution is at empowering women, I review seven unique agency-centered collective action frames in the following sections. Agency Heavy: Balanced Kabeer’s (1999) collective action frame acts as a master frame of empowerment centered around agency. While Kabeer’s frame has been borrowed by empowerment advocates in the sole form of borrowing her definition of empowerment,9 her interrelated framework,10 or bits and pieces of her frame, I find that no other piece of literature replicates Kabeer’s frame exactly (making it the sole piece of this frame). Kabeer’s (1999) frame centers around agency because she conceptualizes power in terms of the ability to make choices. Every part of her frame is aligned with the operationalization of choice, including the denial of choice (or disempowerment). Kabeer’s diagnostic frame is understood to be women’s disempowerment; a condition caused mainly by gendered values that manifest in overt material ways (i.e., gender inequality in well-being) and covert psychological ways (i.e., internalized oppression). Empowerment is “inescapably wound up with the condition of disempowerment” (Kabeer 1999:437). Kabeer is precisely interested in the inequalities of people’s capacity to make 8 Analytically, all seven of the collective action frames discussed in this section emphasize the prognostic code “agency/personal autonomy” describing the extent to which women gain or exert control over their lives. 9 Kabeer (1999:437) defines empowerment as “the expansion in people’s ability to make strategic life choices in a context where this ability was previously denied to them.” 10 Many empowerment advocates know Kabeer’s empowerment framework as an interrelated model comprising of resources (potential choice), agency (process of choice), and achievements (outcomes of choice). 45 choices (rather than differences in achievements/choices they make) that are relevant to power. The two conditions that must be fulfilled for a woman to have “meaningful choice” related to power are: alternatives must exist and must be seen to exist. The first condition relates to factors that limit women’s material possibility of having meaningful choice. In other words, we can first know if women are disempowered if there are striking gender inequalities in basic well-being achievements between men and women in a society. However, she says this operationalization alone may hide covert forms of power women have internalized. Therefore, the second condition of choice relates to factors that limit women’s perceived possibility of having meaningful choice. She finds this condition important in operationalizing empowerment because simply accepting women’s behavior as “real choice” may be completely wrong, and instead reflect the denial of choice that stem from and reinforce women’s subordinate status. Further, Kabeer (1999) operationalizes two consequences of choice as further measures to evaluate empowerment; indicating a prognostic frame centered around agency and critical consciousness, a motivational frame requiring the women in question and “power with” others for empowerment to happen, and the often missed (yet critical) feminist goal of an alternative gendered organization. The name for this frame, “Agency Heavy-Balanced,” comes from Kabeer’s instruction to give equal importance to the two consequences of choice: namely, the need for women’s emic value (i.e., first-order strategic choices) and feminist goals of an alternative relationship between men and women (i.e., transformatory significance). Unlike many other agency-centered collective action frames, Kabeer is not interested in counting any form of women’s increased agency (or achievements) as “successful” empowerment. For her, achievements (or outcomes of agency) are considered empowerment when they allow women to step out of their prevailing gender norms and have transformatory effects on structures of gender 46 inequality. Notwithstanding, she is clear that transformative agency and achievements have to start with women themselves, which is why critical consciousness becomes a starting point for empowerment in this frame. While individual changes in critical consciousness and agency are important for beginning processes of empowerment, they are not by themselves enough to challenge the systemic reproduction of gender inequality; this requires longer-term processes of changing the culture and structures that reinforce women’s subordinate status through “power with” others in ultimately creating alternative relationships between men and women. Agency Heavy: Strategic > Emic This collective action frame, “Agency Heavy: Strategic > Emic” aligns with Kabeer’s empowerment frame the most. Like Kabeer’s, this frame’s most prominent prognostic frame is agency and focuses on critical consciousness as creating the basis for achieving strategic goals (i.e., empowerment). Further, the two pieces making up this collective action frame (Moser 1989; Mosedale 2005) hold women’s emic value as important and thus the need for women and “power with” others in creating empowerment. This is the first of two collective action frames I review to show the discrepancies between frames deriving from Kabeer’s “balanced” frame, when valuing a certain “consequence of choice” more than the other. As it is insinuated by its title, this collective action frame is more extreme in defining the strategic goals of empowerment as more important than women’s emic value, compared to Kabeer’s frame. Specifically, it ultimately favors feminist goals of an alternative relationship between men and women above women’s practical needs. For example, Mosedale (2005:252) shows her similarities and differences to Kabeer’s frame by specifically referencing it and admitting her definition is closely related. However, she says it has two differences. The first is its focus on women (which is not entirely different because Kabeer’s diagnostic is centered 47 around gender inequality). However, the second stated difference is the factor that distinguishes it from Kabeer’s, in that this frame values strategic end goals more highly (whereas Kabeer’s frame is more balanced). Mosedale (2005:252) describes this difference as having “more of an emphasis on women achieving a change that expands options not only for themselves but also for women in general both now and in the future.” Further, in deciding whether or not agencies should intervene if they believe actions will not improve women’s situation in this way, Mosedale (2005:253) shows her favoritism over strategic goals by suggesting, “Ultimately, of course, if the agency is convinced that the proposed action will actually be harmful it has the power to decline to get involved. However this should be a last resort.” This frame also believes in a bottom-up approach in achieving strategic goals. For Mosedale (2005), she explicitly states power cannot be bestowed, it must be won. Similarly, Moser (1989) describes the ways development approaches, including the equity approach, while concerned with women’s strategic goals, are unsuccessful because they do not begin with women’s needs. To make this point, she cites Third World activists’ feelings that this kind of approach is like taking “feminism to a woman who has no water, no food and no home is to talk nonsense” (Bunch 1980:27; cited in Moser 1989:182). Thus, while strategic goals are indeed the favored goal between emic and strategic, there is an understanding that empowerment must come from women first, from a place of critical consciousness, and not from the “top-down.” Agency Heavy: Emic > Strategic This collective action frame, like the former two, is heavily centered around women’s agency and women’s critical consciousness as the basis of the prognostic frame; upheld by the motivational frame of women themselves and “power with” others in creating empowerment processes. However, this frame differs in the ways it ends up counting any form of agency as 48 empowering, as long as women want it. Thus, unlike the previous frame, this collective action frame values women’s emic value higher than any strategic goal. Both pieces making up this frame (Wieringa 1994; León 2017) focus more on prognostics than on identifying or creating a diagnostic frame. Specifically, both pieces’ definition of empowerment is linked to the emic interests of those who are deprived of power (i.e., choice). For example, León (2017:25) defines empoderarse, empower in Spanish, as, “People gain control of their lives, gain the ability to do things and define their own agenda.” Additionally, Wieringa shows that her conceptualization of power is centered around choice. This leads to a recognition that this frame is not concerned about reversing power hierarchies per se, but strictly with increasing women’s agency of any kind: “The central focus of this approach is a critique of the way power and development are interlinked; it seeks ways to ‘empower’ women, not in the sense of reversing existing power hierarchies but rather in empowering women and/or women’s groups to make their own choices, to speak out on their own behalf and to control their own lives” (Wieringa 1994:833). While the previous frames consider transformatory forms of agency to be evidence of “successful” empowerment, this frame does not. Wieringa (1994) published her piece before Kabeer (1999) and wrote hers in direct opposition to Moser’s piece (1989) on the basis that the binary distinction between strategic and practical (and the favoring of the former) denies women the possibility to define their problems. For this frame, it does not matter which feminist goal the author identifies as important, but that they ultimately prioritize women’s emic needs over strategic feminist goals. For example, in siding with emic (i.e., women’s value) over etic (i.e., outsiders’ value, in this case planners) in defining strategic goals, Wieringa (1994) provides a 49 cultural relativist understanding women’s empowerment as both empathetic to the severe constraints of women’s empowerment in the context of violence and patriarchy, and prioritizing women’s survival needs over everything else. In practice, by implementing this empathetic (emic heavy) frame, women gaining the ability to do things emphasizes “the lives they want” rather than on “expanding women’s possibilities.” When empowerment becomes heavily defined by the women in question, patriarchal bargains begin to be considered a legitimate way of expanding women’s possibilities, especially in a severe patriarchal context involving gender violence. In addition, because she very much favors women’s emic value in defining empowerment, Wieringa (1994) proposes measuring empowerment by tracking the progression of women’s ability to form strategic goals of their own (from critical consciousness) over time. In practice then, this means measuring “successful” empowerment based on the progression of this skill, rather than women’s ability to act on these goals. Thus, while these pieces have definitions that center around women gaining the ability to do things, these “things” primarily boil down to women building critical consciousness and defining their own agendas, not necessarily acting on them. Agency Heavy Alternatives for Women The dominant prognostic in this collective action frame, “Agency Heavy Alternatives for Women,” as the name infers, is agency. It is made up of two pieces of literature (Sadiqi 2008; Mohanty and Miraglia 2012), and like the collective action frames already discussed, this frame too, takes women’s emic value seriously in defining the goals of empowerment. However, unlike the other agency-centered frames that value both emic and strategic goals, this frame does not mention critical consciousness as a prognostic for empowerment. Rather, this frame is so heavily entrenched with valuing women’s emic value and forms of agency, that it is assumed women 50 know what they need in creating strategic gender goals, given their positionality of oppressed. This positionality, rather than assuming gives women internalized oppression, offers a space for them to critically question (providing a fuller picture of the problem) and create alternatives for women. Specifically, these alternatives can be found in forms of resistance these women practice (i.e., empowerment). As Mohanty and Miraglia (2012:111) state, “Engaging the perspectives and experiences of those most marginalised by capital provides a fuller understanding of how capital operates and points towards strategies of resistance that can fundamentally transform social inequalities from the ground up.” Mohanty and Miraglia (2012) do not explicitly use the word “empowerment” because of the ways instrumental frames have used and popularized its approach. In rejecting this word, and its implied generalized downfalls, they suggest focusing on women’s agency in creating an alternative approach. They describe their approach in the context of resource management: “Rather than impose an empowerment agenda … we argue for attention to women’s agency in the context of gender justice and imagining alternative futures – not in defending a patriarchal status quo or in the service of conservative social movements. Starting with women’s transformative agential practices is a “bottom-up process” that builds on and broadens the agency that women already express. (Mohanty and Miraglia 2012:110) The feminists making up this collective action frame are transnational feminists, concerned with the empowerment of women in the Global South. Therefore, the identified systems that women are oppressed by, and resisting to, include colonial legacies, neoliberalism, religion, and especially patriarchy. Because these systems are the identified problem and resisting and creating alternatives for women is the ultimate goal, prognostics like participation are not considered evidence of empowerment; only the ways women become active agents of 51 change in resisting these structures are considered empowerment. For example, Sadiqi (2008) discusses the ways women code-switch in a context of Arab-Muslim patriarchy that gives them empowerment in conversational enterprises. Specifically, she describes how women code-switch to languages (like French) to have people, especially men, listen to what they have to say; as it is a way of saying, “I am here and I want to speak!” (Sadiqi 2008:38). Because code-switching is a conscious activity and involves choice, “it denotes women’s agency in everyday all-female or mixed interactions.” Thus, this example helps illustrate the way this collective action frame considers women’s resistance to systems of oppression, using agency, as a way to create alternatives for women (i.e., creating and full-filling strategic goals). Agency Heavy Beyond Alternative for Women Like all the frames discussed thus far in this section, this collective action frame is also agency heavy. It begins by framing the problem of women’s empowerment in the context of hegemonic paradigms that produce systems of unequal and unearned privileges – which are supported by ideologies that reproduce inequalities along the lines of race and gender. While the two pieces of literature that make up this frame (Sanyal 2014; Viveros-Vigoya and Ruette- Orihuela 2021) are case studies from opposite sides of the world, they both consider agency to be empowerment when it takes the form of freedom. Specifically, forms of agency that emancipate women from harmful ideologies. Therefore, not all forms of agency are considered empowerment, but only when agency becomes synonymous with women’s freedom. This leads Sanyal (2014) to adopt the term “improved agency” and Viveros-Vigoya and Ruette-Orihuela (2021) to use the phrase “emancipatory agency,” to signify empowerment. Accordingly, this frame considers new ideas and practices outside of the dominant ideology and systems of domination and subordination to be “successful” or evidence of 52 empowerment, as they signify forms of agency that denounce the systems that oppress them. For example, Viveros-Vigoya and Ruette-Orihuela (2021) describe empowerment in ways black “Afro-Columbian” women actively challenge the “white gaze” through individual choices in black aesthetic (care) practices. In the context of West Bengal, Sanyal (2014) considers women using unrestrained freedom of agency (e.g., physical mobility) as evidence of their empowerment. She further suggests this can happen through women’s interaction with others – specifically those that act as a second “habitus” - in creating a critical consciousness that push against patriarchal ideologies. At the same time, unlike the previous agency heavy collective action frame, this frame acknowledges the shortcomings of having “alternatives for women” as the only strategic goal of empowerment. While alternatives for women are considered as evidence of empowerment, this frame is cautious about the repercussions of this sole goal. Specifically, this collective action frame worries that strategies and programs that are successful in “empowering women” using this goal alone, will lead to new and deepening inequalities. Without the additional strategic goal of creating an alternative gendered organization, this frame posits that new inequalities will form on the basis of those who achieve “empowerment” and those that do not (either because the strategy/program was only accessible to certain women, or the alternative(s) only worked/were successful for some women). Thus, the ultimate feminist goal for women’s empowerment in this frame is to strive for an alternative gendered organization; but the frame still recognizes the importance of creating alternatives for women in the meantime. This is explained in both pieces, but Viveros-Vigoya and Ruette-Orihuela (2021:121) describe this caveat particularly well. In their ethnographic study, the authors find that the strategy taken by primarily urban black- Colombian women - of affirming aesthetic care for “natural” and “afro-textured” hair - helped 53 erode racist stereotypes and produced emancipatory agency for them. Yet, after considering the following questions, the authors conclude the strategy was limited by hiding (and potentially creating future) asymmetrical social relations between black women: “To what point is the aesthetic value produced by black women accumulated and kept in cities, in digital activism spaces, and in student feminist niches, and not redistributed to women living in low-income sectors or rural areas? To what point do these young, urban women’s subjective empowerment processes concern women who do not have the chance of using these social networks or products?” Agency & Participation This next collective action frame differs from the ones discussed so far, because agency is not alone in getting the most attention in the prognostic frame. In addition to agency, this frame includes access to resources, participation, and expanding opportunities as important prognostics for women’s empowerment. These prognostics all stem from the main diagnostic of the problem, gender discrimination. In addition, like many other agency-centered empowerment frames, this frame sees empowerment processes as starting from within - in the development of critical consciousness – which is viewed as a personal problem for the women in question. At the same time, NGOs are seen as contributing greatly to women’s empowerment. Access to resources is identified as an important prognostic because all four pieces making up this frame (Eade and Williams 1995; Oxaal and Baden 1995; Goetz and Gupta 1996; Mehra 1997) focus on the empowerment of poor women. Further, participation and agency/autonomy seem to go hand-in-hand in this frame, as women gaining control over their lives is part of understanding empowerment, and one way to do that is through participating in decision-making spaces that influence their lives. At the same time, while participation is very 54 important to this frame in defining empowerment, the frame is also critical of participation as always counting as “successful” empowerment. Thus, agency or personal autonomy is still at the heart of understanding empowerment because women deciding to participate or not is more important than the act of participation, especially if it is forced. Two of the four pieces making up this “Agency & Participation” collective action frame present a possible contradiction or shortcoming in their prognostic of “expanding opportunities.” The two pieces (Goetz and Gupta 1996; Mehra 1997) set out understanding expanding opportunities as important in changing women’s gender status, but in practice end up counting things women already do as forms of empowerment; leading them to consider patriarchal bargains as empowerment. For example, Mehra (1997:138) defines empowerment first as “the expansion of choices for women and an increase in women’s ability to exercise choice” and then as “the need for development policies and programs that can enable girls and women to challenge current norms and change conditions – a process that can be termed “empowerment”.” However, when Mehra (1997:148) describes what made a certain empowerment program successful at empowering women in India, she says the success is based on women using their gender assigned roles easier and not expanding their gendered possibilities. While the Self Employed Women’s Association (SEWA) can be praised for a number of reasons, Mehra (1997) praises the success of women’s empowerment based on ongoing activities and steering clear from “new and unfamiliar” activities; which does not match the radical goal that she describes in her definition. Agency & Structure All six of the pieces in this collective action frame (Kishor 2000; Mayoux 2000; Bennett 2002; Mason and Smith 2003; Narayan 2005; Alsop, Bertelsen, and Holland 2006) understand 55 empowerment as an instrumental goal and highlight agency and resources as the prognostic for making institutions more accessible and equitable. To this frame, empowerment is conceptualized as the resulting interaction of agency and opportunity structure. Analytically, for this collective action frame, agency represents most of the direct measures for evaluating “successful” empowerment, and opportunity structure represents the diagnostic (and aligned prognostic) frame. First, agency is at the center of understanding if empowerment has occurred. Alsop and colleagues (2006:10) provide a good example of the way this collective action frame closely ties agency to empowerment with their definition of empowerment as, “A group’s or individual’s capacity to make effective choices, that is, to make choices and then to transform those choices into desired actions and outcomes.” Similarly, Kishor (2000:131) operationalizes women’s empowerment as an end-result by “directly measuring women’s control over their lives and environment” (i.e., agency/autonomy) by using measures such as earnings controlled by women currently and before marriage, and indices measuring their direct access to money. Second, opportunity structure is understood to represent the broader social context that provides the “rules of the game” for exercising agency and determining its effectiveness. Thus, to a large extent, this frame identifies culture, specifically social norms, as the source of the problem for those who are disempowered. However, rather than identifying cultural change through radical transformation of structures (that create inequality in the first place) as the solution, the prognostic for empowerment is to make institutions more accessible and equitable by removing the formal and informal institutional barriers that limit ones’ choices. For this frame, radical transformation of institutions is not the solution because their analyses of power and gender inequality is operationalized as a problem of distribution and access (i.e., 56 discrimination). Relatedly, there tends to be an implicit tendency to conflate rich with powerful and poor with disempowered; leading to a framing of women not having control/autonomy over resources as the problem and resources seen as enabling agency by providing an opportunity for choice to exist. Specifically, it is mostly up to development organizations and government to “make” empowerment happen in one of two ways: interventions or involving those with relative power to make institutions more equal. Thus, the irony of this frame focusing on agency is that it is not necessarily those who are disempowered that are responsible for their own empowerment but done from “positions of relative power within the existing institutional framework” (Bennett 2002:7) attempting to change power relations between poor and powerful actors (Narayan 2005). Essentially, this frame suggests that increasing disempowered actors’ agency is the goal and is most effectively/efficiently done through changing institutional policies. The problem with this becomes clear with one example: Mason and Smith (2003) include their indication of opportunity structure by emphasizing the goal of women’s freedom in the domestic sphere by measuring women’s physical freedom of movement and family decision-making power. Based on their findings, they suggest policy recommendations to increase women’s empowerment such as raising the female age at marriage. In other words, they suggest restricting freedom (e.g., raising female age at marriage) so women can gain a different freedom (i.e., a freedom the researchers think is more important/indicative of empowerment). In essence, this frame focuses on agency, but only forms of agency that are predetermined as important to outsiders, rather than the women in question. 57 Conclusion My aim in this study was to conceptually clarify how advocates (scholars/feminists/organizations) frame empowerment for women in the Global South. The frame analysis in this study suggests 16 unique collective action frames, or “stories,” of women’s empowerment in the development discourse. Notably, I revealed the ways empowerment frames are similar and different from one another in the way they elaborate on certain solutions (prognostic) to particular end goals (motivational) within a set of amplified problems (diagnostic). This finding contributes not only to the gender and empowerment discourse, but also to the field of frame analysis where this finding supports Benford and Snow’s (2000) claim that most collective action frames do not vary based simply on the topics or issues identified, but by the way they are discussed by claims-makers (e.g., women’s empowerment advocates). Theoretically, 16 collective action frames of empowerment suggest there are 16 interpretations of what “successful empowerment” looks like in development. Using frame elaboration, I uncovered the factors that influence the interpretation of how successful a solution is at empowering women. Specifically, these factors are the elaborated framing tasks (diagnostic, prognostic, motivational) that make each collective action frame unique. In other words, the elaborated framing tasks that make each collective action frame unique, produce varying conclusions regarding whether a solution (e.g., policy, program, etc.) is successful or not at empowering women in the Global South. This means ideas about what solutions are successful at empowering women do not simply vary by time or place, but at its core, they vary based on the empowerment frame one uses to understand if empowerment is “happening” or “has happened.” Thus, while scholars already know the dangers of generalizing solutions across women in different contexts, my 58 findings suggests that now, more attention needs to be drawn to what empowerment frame is being used to promote and assess these solutions in the first place. Practically, this means that we must do more than simply define empowerment in development work, but rather show transparency in how we frame it. As my findings revealed, each meaning or frame of women’s empowerment implies a particular set of solutions (i.e., implementations) and way of knowing (i.e., evaluating) if these solutions are “successful” or not in terms of women’s empowerment. Thus, by fully and transparently framing empowerment, we ultimately practice more intention in how we implement and evaluate solutions for women in the Global South. Moreover, we should ensure our operationalization and measurement of “successful empowerment” aligns with the goals our proclaimed solutions (e.g., programs, policies, collective action, etc.) intend to achieve. In using frame analysis to systematically uncover the defining features of empowerment’s shared and contested aspects, the present study can help academics, development practitioners, and activists involved in implementing and/or evaluating women’s empowerment to identify - and align with - an appropriate collective action frame of empowerment. For example, a feminist scholar whose idea of feminism or personal research ethics are against telling women what to do, would take the findings presented here to identify a frame that does not involve 1) an external agent prompting critical consciousness or 2) heavily weigh strategic goals as an indicator of “successful empowerment.” Or maybe a development practitioner or NGO is interested in a community-driven frame of empowerment. Choosing a collective action frame that heavily weighs emic value may be an appropriate choice to guide the formation of programs and solutions implemented, and then using that frame’s idea of “successful empowerment” to evaluate the outcomes. 59 Finally, since frames are created over a shared meaning and reading of social life and situations, it is inevitable that more empowerment frames already exist and will evolve in the future. This frame analysis provides the well-needed groundwork in revealing the similarities and differences that allow for the bridging (or connecting) of empowerment frames, by highlighting the defining features that can be elaborated or suppressed to advance them further. This study specifically focused on the ways empowerment is framed for (and to an extent by) women in the Global South. Further research is needed on the ways empowerment is framed for women in the Global North to understand the potential for bridging collective action frames between the Global North and Global South and for the development of cross-national indicators of empowerment. Lastly, I attempted to produce a more inclusive frame analysis of empowerment than what currently exists in the development discourse. However, one bias that remained was that all pieces were originally written in a colonizer language (e.g., English or Spanish). Thus, an important area of future research that can contribute to the present study will include a larger sample that captures a diversity of languages. Doing so will advance the inclusiveness of women’s empowerment frames across all spaces and languages, which is the goal of feminist research. 60 REFERENCES Alsop, Ruth, Mette Bertelsen, and Jeremy Holland. 2006. Empowerment in Practice: From Analysis to Implementation. Washington, DC: The World Bank Publications. Batliwala, Srilatha. 2007. "Taking the Power Out of Empowerment–An Experiential Account." Development in Practice 17(4-5):557-565. Benford Robert D. and David A. Snow. 2000. “Framing Processes and Social Movements: An Overview and Assessment” Annual Review Sociology 26:611-39. Calvès Anne-Emmanuèle. 2009. “Empowerment: Généalogie d'un Concept clé du Discours Contemporain sur le Développement.” Revue Tiers Monde 200:735-749. Retrieved July 28, 2021: https://www.cairn.info/revue-tiers-monde-2009-4-page-735.htm Carty, Linda, and Chandra Talpade Mohanty. 2015. "Mapping Transnational Feminist Engagements." Pp.82-115 in The Oxford Handbook of Transnational Feminist Movements, edited by R. Baksh and W. Harcourt. Oxford: Oxford University Press. Cornwall, Andrea, and Ann Whitehead. 2007. Feminisms in Development: Contradictions, Contestations and Challenges. Zed Books. Drydyk, Jay. 2013. “Empowerment, Agency, and Power.” Journal of Global Ethics 9(3):249- 262. Ibrahim, Solava, and Sabina Alkire. 2007. "Agency and Empowerment: A Proposal for Internationally Comparable Indicators." Oxford Development Studies 35(4):379-403. Kabeer, Naila. 1999. "Resources, Agency, Achievements: Reflections on the Measurement of Women's Empowerment." Development and Change 30(3):435-464. Kandiyoti, Deniz. 1988. “Bargaining with Patriarchy.” Gender & Society 2(3):274-289. Malhotra, Anju and Sidney Ruth Schuler. 2005. “Women’s Empowerment as a Variable in International Development.” Pp. 71–88 in Measuring Empowerment: Cross- Disciplinary Perspectives, edited by Deepa Narayan. Washington, DC: World Bank, Gender and Development Group. Mohanty, Chandra Talpade, and Sarah Miraglia. 2012. "Gendering Justice, Building Alternative Futures." Pp. 115-148 in Alternatives to Privatization, edited by D. McDonald and G. Ruiters. New York, NY: Routledge. Moser, Caroline ON. 1989. "Gender Planning in the Third World: Meeting Practical and Strategic Gender Needs." World Development 17(11):1799-1825. Mukhopadhyay, Maitrayee. 2004. “Mainstreaming Gender or “Streaming” Gender Away: Feminists Marooned in the Development Business.” IDS Bulletin 35(4):95-103. 61 Narayan, Deepa. 2005. “Conceptual Framework and Methodological Challenges.” Pp. 3-38 in Measuring Empowerment: Cross-Disciplinary Perspectives, edited by D. Narayan. Washington, DC: World Bank, Gender and Development Group. Oxaal, Zoe, and Sally Baden. 1997. “Gender and Empowerment: Definitions, Approaches and Implications for Policy.” Briefing prepared for the Swedish International Development Cooperation Agency (Report No. 040). Brighton, England: Bridge Development— Gender, Institute of Development Studies. Retrieved March 1, 2019 (http://www.bridge.ids.ac.uk/sites/bridge.ids.ac.uk/files/reports/re40c.pdf). Pyke, Karen D. 2010. “What is Internalized Racial Oppression and Why Don’t We Study It? Acknowledging Racism’s Hidden Injuries.” Sociological Perspectives 53(4):551-572. Richardson, Robin A. 2018. "Measuring Women’s Empowerment: A Critical Review of Current Practices and Recommendations for Researchers." Social Indicators Research 137(2):539-557. Shiva, Vandana. 2015. "Save Our Soil!" The OECD Observer: 54. Snow, David. 2004. "Framing Processes, Ideology, and Discursive Fields." The Blackwell Companion to Social Movements 1:380-412. Snow, David A. and Robert D. Benford. 1988. “Ideology, Frame Resonance, and Participant Mobilization.” International Social Movement Research 1:197-217. Snow, David A., E. Burke Rochford Jr., Steven K. Worden, and Robert D. Benford. 1986. “Frame Alignment Processes, Micromobilization, and Movement Participation.” American Sociological Review 51(4):464-81. Snow, David A., Rens Vliegenthart, and Pauline Ketelaars. 2019. "The Framing Perspective on Social Movements: Its Conceptual Roots and Architecture." Pp. 392-410 in The Wiley Blackwell Companion to Social Movements, edited by D.A. Snow, S.A. Soule, H. Kriesi, and H.J. McCammon. 2nd ed. Hoboken: Wiley Blackwell. Wright, Wynne and Taylor Reid. 2010. “Green Dreams or Pipe Dreams?: Media Framing of the U.S. Biofuels Movement.” Biomass and Bioenergy 35(4):1-10. 62 APPENDIX A: FINAL SAMPLE Table 1.1. Pieces of literature in final sample, by year published (N = 54, 1987 - 2021) Year Number Title Sen, Gita, and Caren Grown. 1987. Development, Crises, and Alternative Visions: Third Word Women’s Perspectives. London: Earthscan 1987 1 Publications. Antrobus, Peggy. 1989. “The Empowerment of Women.” Pp. 189-207 in The Women and International Development Annual. First 1989 25 Edition. Routledge. 1989 26 Bhatt, Ela. 1989. "Toward Empowerment." World Development 17(7): 1059-1065. Moser, Caroline O.N. 1989. "Gender Planning in the Third World: Meeting Practical and Strategic Gender Needs." World Development 1989 2 17(11):1799-1825. 1992 3 Friedmann, John. 1992. Empowerment: The Politics of Alternative Development. Oxford: Blackwell. Batliwala, Srilatha. 1994. "The Meaning of Women's Empowerment: New Concepts form Action.” Pp. 127-138 in Population Policies 1994 4 Reconsidered: Health, Empowerment, and Rights, edited by G. Sen, A. Germain, and L. C. Chen. Wieringa, Saskia. 1994. "Women's Interests and Empowerment: Gender Planning Reconsidered." Development and Change 25(4):829- 1994 5 848. Eade, Deborah and Suzanne Williams. 1995. The Oxfam Handbook of Relief and Development. Vol 1, Approaches, Focusing on People, 1995 6 Capacity Building. Oxford: Oxfam. UNDP (United Nations Development Programme). 1995. Human Development Report 1995: Gender and Human Development. New 1995 7 York: Oxford University Press. Goetz, Anne Marie, and Rina Sen Gupta. 1996. “Who Takes the Credit? Gender, Power, and Control over Loan Use in Rural Credit 1996 8 Programs in Bangladesh.” World Development 24(1):45–63. UN (United Nations). 1996. The Beijing Declaration and the Platform for Action: Fourth World Conference on Women: Beijing, 4-15 1996 9 September 1995. New York: United Nations. Mehra, Rekha. 1997. "Women, Empowerment, and Economic Development." The Annals of the American Academy of Political and 1997 27 Social Science 554(1):136-149. Oxaal, Zoe, and Sally Baden. 1997. “Gender and Empowerment: Definitions, Approaches and Implications for Policy.” Briefing prepared 1997 10 for the Swedish International Development Cooperation Agency (Report No. 040). Brighton, England: Bridge Development-Gender, Institute of Development Studies. Retrieved March 1, 2019 (http://www.bridge.ids.ac.uk/sites/bridge.ids.ac.uk/files/reports/re40c.pdf). 1997 11 Rowlands, Jo. 1997. Questioning Empowerment. Oxford: Oxfam. Kabeer, Naila. 1999. “Resources, Agency, Achievements: Reflections on the Measurement of Women’s Empowerment.” Development 1999 12 and Change 30(3):435–464. Kishor, Sunita. 2000. “Empowerment of Women in Egypt and Links to the Survival and Health of their Infants.” Pp. 119-156 in Women’s 2000 13 Empowerment and Demographic Processes: Moving Beyond Cairo, edited by H. B. Presser and G. Sen. New York: Oxford University Press. Mayoux, Linda. 2000. “From Access to Empowerment: Gender Issues in Micro-Finance.” CSD NGO Women’s Caucus Position Paper for 2000 14 CSD-8. CSD. 63 Table 1.1. (cont’d) Samuel, Kumudini. 2001. "Gender Difference in Conflict Resolution." Pp. 184-204 in Gender, Peace and Conflict. Edited by I. 2001 28 Skuelsboek and D. Smith. Oaks, CA: Sage. Mukherjee, Vanita Nayak. 2002. "Gender Matters." Beyond Numbers: A Symposium on Population Planning and Advocacy. Pp. 511:67- 2002 29 75. Bennett, Lynn. 2002. “Using Empowerment and Social Inclusion for Pro-Poor Growth: A Theory for Social Change.” Background Paper 2002 15 for the Social Development Sector Strategy Paper. World Bank. Mason, Karen Oppenheim, and Herbert Smith. 2003. Women’s Empowerment and Social Context: Results from Five Asian Countries. 2003 16 Washington, DC: Gender and Development Group, World Bank. Banerjee, Nirmala, and Poulomi Roy. 2004. Gender in Fiscal Policies: The Case of West Bengal. New Delhi: United Nations 2004 30 Development Fund for Women. Grootaert, Christiaan. 2005. “Assessing Empowerment at the National Level in Eastern Europe and Central Asia.” Pp. 309-340 in 2005 17 Measuring Empowerment: Cross-Disciplinary Perspectives, edited by D. Narayan. Washington, DC: World Bank, Gender and Development Group. Lokshin, Michael, and Ravallion, Martin. 2005. “Self-Rated Power and Welfare in Russia.”, Pp. 177-195 in Measuring Empowerment: 2005 18 Cross-Disciplinary Perspectives, edited by D. Narayan. Washington, DC: World Bank, Gender and Development Group. Malena, Carmen, and Volkhart Finn Heinrich. 2005. “Measuring Empowerment at the National Level: The Case of the CIVICUS Civil 2005 19 Society Index (CSI).” Pp. 341-364 in Measuring Empowerment: Cross-Disciplinary Perspectives, edited by D. Narayan. Washington, DC: World Bank, Gender and Development Group. Malhotra, Anju, and Sidney Ruth Schuler. 2005. “Women’s Empowerment as a Variable in International Development.” Pp. 71–88 in 2005 20 Measuring Empowerment: Cross-Disciplinary Perspectives, edited by D. Narayan. Washington, DC: World Bank, Gender and Development Group. Moghadam, Valentine M., and Lucie Senftova. 2005. "Measuring Women's Empowerment: Participation and Rights in Civil, Political, 2005 31 Social, Economic, and Cultural Domains." International Social Science Journal 57(184):389-412. Mosedale, Sarah. 2005. “Assessing Women’s Empowerment: Towards a Conceptual Framework.” Journal of International Development 2005 21 17(2):243–257. Narayan, Deepa. 2005. “Conceptual Framework and Methodological Challenges.” Pp. 3-38 in Measuring Empowerment: Cross- 2005 22 Disciplinary Perspectives, edited by D. Narayan. Washington, DC: World Bank, Gender and Development Group. Raju, Saraswati. 2005. “Limited Options-Rethinking Women’s Empowerment ‘Projects’ in Development Discourses: A Case from Rural 2005 32 India.” Gender, Technology and Development 9(2):253-271. Wedderburn, Judith. 2006. "Gender, Trade Liberalization and the CARICOM Single Market and Economy: Challenges and Options for 2006 33 Civil Society." Caribbean Quarterly 52(2-3):138-155. Alsop, Ruth, Mette Bertelsen, and Jeremy Holland. 2006. Empowerment in Practice: From Analysis to Implementation. Washington, DC: 2006 23 The World Bank Publications. Jain, Devaki, and Shubha Chacko. 2008. "Unfolding Women's Engagement with Development and the UN: Pointers for the Future." 2008 34 Forum for Development Studies 35(1):5-36. Sadiqi, Fatima. 2008. "Language and Gender in Moroccan Urban Areas." International Journal of the Sociology of Language 190:145- 2008 35 165. 2010 36 Corrêa, Sonia. 2010. "Sexuality, Gender and Empowerment." Development 53(2):183-186. 64 Table 1.1. (cont’d) 2010 37 Jad, Islah. 2010. "Palestinian Women Contesting Power in Chaos." IDS Bulletin 41(5):81-88. 2010 38 Taylor, Viviene. 2010. "Emancipation and its Failurest." Development 53(2):164-168. Barroso, Carmen, and Serra Sippel. 2011. "Sexual and Reproductive Health and Rights: Integration as a Holistic and Rights-Based 2011 39 Response to HIV/AIDS." Women's Health Issues 21(6):S250-S254. Bieri, Sabin, Cordula Ott, Ada Freytes Frey, Cecilia Cross, Florencia Partenio, and María Inés Fernández Álvarez. 2011. "A Tool for Thought and Transformation: Gender-Considerate Global Change Research in Practice." Pp. 185-205 in Research for Sustainable 2011 40 Development: Foundations, Experiences, and Perspectives, edited by U.M. Wiesmann and H. Hurni. Bern, Switzerland: Geographica Bernensia. Chattier, Priya, and Claire Slatter. 2012. "Regional Dialogue on Sustainable Development and the Post-2015 Development Agenda." 2012 41 Pacific Contextual Paper. Mohanty, Chandra Talpade, and Sarah Miraglia. 2012. "Gendering Justice, Building Alternative Futures." Pp. 115-148 in Alternatives 2012 42 to Privatization, edited by D. McDonald and G. Ruiters. New York, NY: Routledge. Alkire, Sabina, Ruth Meinzen-Dick, Amber Peterman, Agnes Quisumbing, Greg Seymour, and Ana Vaz. 2013. The Women’s 2013 24 Empowerment in Agriculture Index.” World Development 52:71–91. Yiping, Cai. 2013. "Re-vitalize, Re-strategize and Re-politicize the Chinese Women's Movement in the New Era." Asian Journal of 2013 43 Women's Studies 19(1):113-126. Reid, Sandra D., Rhoda Reddock, and Tisha Nickenig. 2014. "Breaking the Silence of Child Sexual Abuse in the Caribbean: A 2014 44 Community-Based Action Research Intervention Model." Journal of Child Sexual Abuse 23(3):256-277. 2014 45 Sanyal, Paromita. 2014. Credit to Capabilities: A Sociological Study of Microcredit Groups in India. Cambridge University Press. Sen, Gita, and Avanti Mukherjee. 2014. "No Empowerment Without Rights, No Rights Without Politics: Gender-Equality, MDGs and the 2014 46 Post-2015 Development Agenda." Journal of Human Development and Capabilities 15(2-3):188-202. 2015 47 Chetty, Tara, and Rachel Faleatua. 2015. "Regional Advocacy Tool." 2015 48 Sharma, Vibha. 2015. “Women Empowerment Through the ICDS Centres in Haryana.” New Horizons XII:8-17. Shettar, Dr. (Smt.) Rajeshwari M. 2015. “A Study on Issues and Challenges of Women Empowerment in India. IOSR Journal of Business 2015 49 and Management 17(4):13-19. 2017 50 León, Magdalena. 2017. "Women Facing Power. A Reflection on Empoderamiento." Cahiers du Genre 2:23-43. Mahata, Dinabandhu, Amit Kumar, and Ambarish Kumar Rai. 2017. "Female Work Force Participation and Women Empowerment in 2017 51 Haryana." International Journal of Humanities and Social Sciences 11(4):1039. Rodríguez Enríquez, Corina. 2017. "SDG 5." Spotlight on Sustainable Development 2017: Reclaiming Policies for the Public. Retrieved 2017 52 August 13, 2021 (https://twaweza.org/wp-content/uploads/2021/01/spotlight_report_2017.pdf#page=64) 2019 53 Mendoza, Breny. 2019. "La Question de la Colonialité du Genre." Les Cahiers du CEDREF 23:90-116. Viveros-Vigoya, Mara, and Krisna Ruette-Orihuela. 2021. "Care, Aesthetic Creation, and Anti-Racist Reparations." Pp. 107-123 in Care 2021 54 and Care Workers, edited by N. Araujo Guimarães and H. Hirata. Springer Cham. Notes: Each piece has a “number” (column two) that corresponds to its sub-sample and its chronological order within that sample: Pieces 1 – 24 derive from sub- sample 1 (Global North-centric); 25 – 54 include sub-sample 2 (Global South) pieces with authors in bold as the feminist of interest (see Appendix B, part 2). 65 APPENDIX B: SAMPLING PROCESSES The following section describes in detail the sampling process for both sub-samples in this study. I begin by describing the process for my first “most cited/Global North-centric” sub- sample, and then describe my second sampling process for the “Global South feminists” sub- sample.11 Part 1: Most Cited Pieces (Global North-centric) Sampling Process I began my sampling processes by first systematically creating a sub-sample of the most- cited/popular pieces of female/women/female empowerment in the development discourse by obtaining well-known published reviews on the term’s operationalization and measurement (i.e., Alsop et al. 2006; Ibrahim and Alkire 2007; Oxaal and Baden 1997; Richardson 2018).12 Moreover, I used these four reviews to conduct a reverse citation search on Google Scholar of the selected studies that were cited as examples of empowerment frameworks (either labeled as concepts or approaches to measuring empowerment), for a beginning total of 112 pieces. After removing 15 duplicate citations across all four reviews, this search process yielded 97 pieces. Eleven articles were unobtainable/could not find/only existed as citations in reviews; and one was only available in French; leaving 85 studies for the title and abstract phase. The second phase of the search-and-screen process involved a review of study titles and abstracts to identify those that appeared to meet the study inclusion criteria, specifically criteria four (e.g., original research or framework) and five (e.g., on empowerment in the Global South). In total, phase two cut 22 pieces of literature, yielding 63 studies.13 However, to conduct a thorough content analysis, additional parameters were introduced in the third phase to pinpoint the most seminal and cited pieces used for analysis (described on the following page).14 11 I use the terms “Global North/South” in the way Mohanty (2003:226) defines it, “used to distinguish between affluent, privileged nations and communities, and economically and politically marginalized nations and communities, as is “Western/non-Western”.” I realize the implication this binary language creates in terms of feminist epistemology, particularly in exacerbating misleading geographical and ideological binarisms. However, I use this imperfect language only to analytically distinguish relatively dominant/privileged frames of gender empowerment that often derive from the Global North (Cornwall and Whitehead 2007), from more marginalized ones - in an attempt to acknowledge and uplift the latter. 12 When choosing reverse citations from Richardson (2018), I only chose articles that were directly related to empowerment. I do this because throughout the article, Richardson conflates empowerment measures with agency measures. The clearest example comes from a section titled “Use Direct Indicators of Empowerment (i.e., Agency) When Possible” (Richardson 2018:546). Because empowerment is not the same thing as agency (Drydyk 2013), and the purpose of this paper is to cover the ways empowerment specifically is framed, I exclude any citations that are part of agency-related discussions in the Richardson (2018) article. 13 Many studies cut in this process severely conflated/used agency and empowerment interchangeably and few were cut because empowerment was only used as a causal variable/its operationalization was not explained nor the focus of the study. 14 To balance the time it would take to thoroughly analyze this literature, I chose to introduce additional parameters that would cut the number of individual literature for a thorough content analysis. However, future research should include the 39 “popular pieces” that were cut in the third stage of this literature selection. 66 1. The first additional parameter introduced in the third phase narrowed down the number of pieces to one per author (or one per grouping of same authors).15 The one piece that remained from this selection process was kept for analysis: this process cut 11 pieces and put seven pieces of literature in the final sample. 2. Next, I identified literature that was classified as having a duplicate from the initial search and screen process between the four empowerment reviews: seven studies were kept for analysis using this parameter. 3. Lastly, I carefully selected seminal works that remained on the list of most-cited pieces. Selection in this process relied heavily on my own examination of the literature based on my knowledge of the gender, empowerment, and development discourse. Specifically, based on my experience reading in this field, I selected pieces written by authors that I perceived as widely-recognized contributors to the discourse, including those whose seminal pieces are frequently cited by others that either use, build on, or critique their conceptualization of empowerment. Ten additional pieces were selected during this process; contributing to a sub- sample total of 24 “most-cited/Global North-centric” pieces for analysis. Part 2: Pieces Written by Feminists from the Global South Sampling Process The method I used for systematically generating the second sub-sample for this study, of highlighting and analyzing pieces authored by those from the Global South, began with the creation of a list of feminists from these countries. To create this list of feminists, I gathered names from: 1) DAWN’s list of founding members, current board members, and executive committee;16 2) A list of “key feminist scholar-activists” from various countries in the Global South that were interviewed and featured in a book chapter on transnational feminism by Linda Carty and Chandra Mohanty (2015); and lastly, 3) I added names of feminists from the Global South gathered through readings suggested to me by professors at Michigan State University and/or from decolonial feminist or transnational feminist literature.17 After removing duplicates of feminists, this first process generated a list of 62 feminists from the Global South. I purposely say “from” because while the intent was to include only empowerment activists that reside in the Global South, not all of them currently reside in the Global South but all of them were born or from a country located in the Global South. I then used Google Scholar to search for a piece of literature by each of the names on my Global South feminist list. Specifically, I searched each feminist’s name individually as the author, whose work was published between the years 1985 and 2021 and must include the word 15 For this process, I chose the piece that was either known as seminal pieces to the women’s empowerment literature, if one of them was a duplicate from the initial search and screen process, the one with more citations according to Google Scholar, and/or the piece with the more elaborate explanation/operationalization of empowerment. In addition, some pieces were cut because they proved to be the same operationalization and list of authors (but in a different authorship order or “focus”): In one case, the article with five times the number of citations (according to Google Scholar on 8/15/2021) was kept for full review. Other cuts in this process followed similar procedure. 16 I chose DAWN (Development Alternatives with Women for a New Era) because they were the first group of “Third World Feminists” (i.e., researchers, activists, and political leaders from the Global South) to introduce the “empowerment approach” to the role of women in the development discourse (Calvès 2009; Moser 1989). 17 Suggested readings came from Dr. Wynne Wright and Dr. Soma Chaudhuri’s graduate-level courses on gender (Gender and the Environment, and Gender and Power, respectively), as well as Dr. Sejuti Das Gupta’s curated reading list for my dissertation research in India. Decolonial and transnational feminist names were gathered either directly from my comprehensive exam reading list or citations/authors from these readings. 67 “empowerment” or “empower,” and at least one of the following words: “gender,” “woman,” “women,” or “female” anywhere in the text. This search either automatically eliminated non- English literature given the searched terminology, leaving work that was originally published in English and/or translated to English from another language.18 If there were multiple works to choose from this search process, I selected one piece of literature based on the following (in order of importance): 1) the piece attempted to operationalize empowerment clearly, 2) had the most citations, and 3) the searched feminist was either the sole author or first author to represent the empowerment piece by that specific Global South feminist.19 Part of this process eliminated feminists from the initial list because they do not necessarily write about or discuss women/gender/female empowerment: after reviewing titles and abstracts during this discovery phase, 24 feminists were cut from the list, leaving 38 feminists’ pieces of literature for further review. Phase three cut literature that did not attempt to elaborate on what empowerment means in their study (i.e., author uses the term, but does not define nor operationalize it) or were reviews (i.e., did not propose their own ideas on empowerment). For example, Vandana Shiva’s (2015:1) sole piece that turned up in the search phase mentions “the empowerment of women” only once and is combined with the idea “gender equality” with no attempt at operationalizing empowerment or elaborating on the concept fully. Eight pieces like this example were cut, leading to a total of 30 Global South feminists’ studies. I combine these 30 pieces with the 24 “popular/Global North-centric” pieces), for a total of 54 pieces of empowerment literature for the frame analysis of this study. 18 There was only one piece where the abstract was translated but the entire piece was not (which was ultimately eliminated from the search process). 19 This was the case unless the most cited piece was an article/study from a more comprehensive book of the same study (e.g., Paromita Sanyal’s (2014) most cited article thoroughly discusses her empowerment analysis in her book, although her 2009 article from the same interviews is the most cited literature on Google Scholar). 68 APPENDIX C: PRACTICAL AND STRATEGIC GOAL CODES I account for practical gender needs with either or both of the following codes: 1) emic value and/or 2) patriarchal bargain. Emic is captured by the first half of the idea of practical gender needs; while patriarchal bargain captures the second half. Specifically, “emic value” indicates that the goals of empowerment come from what the women in question identify as important. Regardless of outsiders playing a role in defining the goals of empowerment, segments are coded as “emic value” if the author starts with and takes seriously women's personal needs and goals as important to defining women's empowerment. Often these goals are based on women’s human survival; an immediate necessity perceived by women in a certain context. While it is more straight-forward in identifying accounts where an author highly values the perspectives of the women in question in (emicially) defining the goals of empowerment, identifying examples of patriarchal bargains are not so overt. In this context, I coded segments with “patriarchal bargain” if the author (indirectly) identifies a patriarchal bargain as an example of women's empowerment. Meaning, the author proposes or identifies a strategy or action as “empowering” that involves a tradeoff made within patriarchal constraints – specifically, it helps women carry out their gender-assigned roles easier/safer or women use their traditional gender roles to get some desired outcome.20 Borrowing from Moser’s description of “strategic gender interests” and the ways many others summarize her idea, I understand and define strategic gender interests as goals that challenge prevailing forms of gender subordination/oppression/inequality that creates alternatives (for women (to live more equitably)) (gender organization/relationship between men and women). The italicized parts of the definition are the most notable difference amongst authors identifying “strategic gender interests” as important to empowerment. I therefore dissect this italicized part into three motivational codes pertaining to the frame’s rationale (i.e., why empowerment is important) into the following: 1) alternatives for women, 2) alternatives for women to live more equitably, and 3) alternative gender organization/relationship between men and women. In doing so, I can better identify the specific parts of “strategic gender interests” that are claimed by authors, and which parts are not. 20 I created a set of questions to help in identifying occurrences of patriarchal bargains by asking the following questions: 1) Does this goal help women carry out their gender-assigned roles? (From Moser’s (1989) definition); 2) Are women using their gender-assigned roles as a means to empower? 3) What are the patriarchal constraints (ideological and material) in this example of empowerment? (Both from Kandiyoti’s (1988) definition); and 4) Where is the internalized sexism in this example of empowerment? (From Pyke’s (2010) internalized racism study). 69 APPENDIX D: COLLECTIVE ACTION FRAME HEAT MAPS The tables included in Appendix D (Tables 1.2 – 1.4.2) represent heat maps of each piece of literature, presented in separate tables by master frame and grouped by collective action frame within each table. Each grouping of codes (e.g., diagnostic, prognostic, etc.) were run separately in MAXQDA to determine the most amplified codes in each piece of literature. In other words, every grouping of codes (e.g., diagnostic, etc.) and each piece of literature has its own proportion (i.e., number of codes) within each grouping to determine which diagnostic codes, prognostic codes, etc., were amplified the most in each piece of literature. The columns indicate pieces of literature, indicated by a number, and grouped by collective action frame (see Appendix A for identifying the piece of literature indicated by each number). The rows indicate the inductive qualitative codes used for analysis. Each grouping of codes is indicated by different colors (e.g., diagnostic codes are presented in a green heat map). The most discussed/amplified codes within each grouping/color are indicated with darker shades. Lastly, motivational codes were broken down into three categories that prompted the inductive formation of codes responding to specific question(s) about the motivational frame: I. “Motivation codes I” answer question: why does this issue matter? As well as related questions regarding the rational/end goal of empowerment (including strategic goals). II. “Motivation codes II” answer the question: who can participate in fixing the issue? III. “Motivational codes III” answer the question: who is the solution for? 70 Table 1.2. Power collective action frame heat maps Power-to Power-to Power-over Zero-sum Heavy Only & Power-to Power 25 11 36 53 1 26 34 4 28 32 Diagnostic Codes . . . . . . . . . . Gender gap 0 0 0 0 0 0 0 0 0 0 Gender discrimination 0 0 0 0 4 2 3 2 2 1 Gender inequality 0 2 0 1 5 0 5 2 2 2 Patriarchy 3 12 0 8 7 0 4 5 7 12 Division of labor 0 9 0 2 12 7 2 2 0 1 Culture/Gender ideology 3 19 0 0 9 0 2 5 11 4 Systems of inequality & oppression 7 8 4 8 19 10 9 2 1 1 Power inequality 3 18 3 5 4 3 1 3 4 6 No power/Powerlessness 0 0 0 0 1 0 0 3 0 0 Disempowerment 0 1 0 0 0 0 1 1 1 1 Violence 2 7 1 1 5 1 1 0 5 1 Poor material conditions 1 2 0 0 8 7 1 2 5 0 Prognostic Codes Status in community 0 2 0 0 2 1 1 0 0 3 Well-being 1 1 0 0 11 5 1 0 1 2 Resources 1 5 0 0 4 4 1 4 0 4 Participation 0 5 0 0 12 0 1 1 10 0 Women's (human) rights 0 5 0 0 7 6 9 3 6 1 Freedom 0 2 2 0 0 0 1 0 1 2 Agency/ Personal autonomy 0 23 1 1 11 2 0 2 2 8 Expanding possibilities 0 9 1 0 1 0 1 0 0 1 Gender equality/Gender equity 1 0 0 2 6 0 2 0 0 1 More power-over 1 10 0 3 7 2 1 6 8 8 Power-to/Power transformation 8 23 4 2 16 3 5 2 0 0 Power-within/Crit. consciousness 8 18 0 0 7 3 0 4 1 3 Transforming home relations 0 9 0 0 2 0 0 1 0 2 Cultural change 1 4 0 2 7 3 6 4 9 2 Motivational Codes I For development 0 0 0 0 0 0 0 0 0 0 Patriarchal bargain 0 1 0 0 0 0 0 0 0 1 Emic value 4 8 1 1 7 3 7 2 6 2 Alternatives for women 1 6 0 0 0 1 1 1 0 0 Alternatives to live more equitably 0 0 0 0 1 3 1 7 3 2 Alternative gender organization 6 2 3 5 8 2 8 5 6 3 Motivational Codes II People with power/privilege 5 9 0 0 9 3 4 2 0 10 Power-with/Collective solidarity 6 29 0 0 19 8 11 6 15 5 Women being empowered 1 13 1 0 8 0 2 3 7 4 Development orgs/NGOs 0 13 0 0 11 1 3 4 1 2 Governing structures 0 1 0 0 10 9 3 1 0 1 Private/Free market 0 0 0 0 0 2 0 0 0 0 Unclear 0 0 2 2 0 0 0 0 0 0 Motivational Codes III Women (no other distinction) 2 3 0 0 0 0 0 0 0 0 Women in the Global South 0 4 0 0 3 1 1 1 1 1 Rural women 0 6 0 0 0 2 0 0 0 2 People in poverty/"The poor" 2 2 0 0 7 3 0 3 0 0 Gender & sexual minorities 0 3 4 1 0 0 0 0 0 0 71 Table 1.3.1. Equal participation and rights collective action frame heat maps, part 1of 2 Human Rights & Gender Mainstreaming Gender Parity 9 31 41 43 52 7 30 18 17 33 39 24 46 48 49 51 Diagnostic Codes . . . . . . . . . . . . . . . . Gender gap 0 0 11 4 4 22 13 2 4 1 2 2 8 1 5 5 Gender discrimination 18 3 9 9 1 22 3 0 7 14 5 0 6 2 9 2 Gender inequality 5 5 3 4 2 5 1 0 3 9 2 2 8 1 6 5 Patriarchy 0 0 0 1 0 0 1 0 0 0 0 0 2 1 2 6 Division of labor 6 2 7 3 3 8 3 0 0 11 0 0 4 1 4 0 Culture/Gender ideology 3 1 2 2 0 1 3 0 1 5 5 2 4 1 3 3 Systems of inequality 1 3 3 4 10 0 0 3 0 11 0 0 7 1 1 0 Power inequality 2 0 0 1 4 0 0 2 1 2 2 0 4 0 0 2 No power/Powerlessness 0 0 0 0 0 0 0 4 0 0 1 0 0 0 1 0 Disempowerment 0 1 0 0 0 0 0 0 0 0 2 5 2 1 1 0 Violence 6 3 3 5 0 4 0 0 1 0 1 0 9 0 8 0 Poor material conditions 7 2 8 1 0 5 10 0 3 3 6 1 2 1 8 0 Prognostic Codes Status in community 2 0 0 0 0 0 0 3 1 0 0 0 0 3 7 9 Well-being 8 0 1 0 2 5 2 14 1 4 5 1 14 11 6 2 Resources 22 7 11 3 1 19 5 7 12 6 10 3 14 14 12 16 Participation 15 9 6 3 0 24 0 1 10 2 0 2 8 0 3 6 Women's (human) rights 18 10 8 7 3 3 1 0 3 1 7 0 31 1 7 0 Freedom 7 2 0 0 0 0 0 0 2 0 0 0 5 0 3 2 Agency 11 1 2 1 0 8 1 9 0 0 7 6 16 1 5 11 Expanding possibilities 0 0 1 0 1 0 0 0 1 3 0 1 0 1 1 1 Gender equality/equity 20 3 5 4 4 18 11 1 3 5 3 4 14 1 7 3 More power-over 8 0 0 0 1 0 0 0 4 0 0 1 2 0 0 0 Power-to/Power trans. 1 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 Power-within/Crit. con. 1 0 0 2 0 0 0 0 0 3 0 0 0 0 0 0 Transform home relations 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Cultural change 3 6 0 2 0 0 0 0 1 2 2 0 0 2 3 0 Motivational Codes I For development 12 5 4 0 1 12 1 1 4 2 4 5 5 1 0 3 Patriarchal bargain 0 0 0 0 0 1 1 0 0 0 1 0 4 4 1 0 Emic value 3 0 1 1 0 0 5 0 0 0 2 0 1 1 2 0 Alternatives for women 0 3 1 0 0 0 0 0 0 1 0 0 0 0 2 0 Alt. to live equitably 5 2 3 2 0 4 5 0 0 0 1 0 2 0 2 0 Alternative gender org. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Motivational Codes II People with power/ privilege 6 0 0 0 1 0 0 0 0 1 2 0 0 0 1 0 Power-with/Collective 0 1 0 7 0 1 0 0 2 1 0 0 6 2 4 0 Women being empowered 4 0 0 1 0 2 0 0 0 0 2 0 2 2 2 0 Development orgs/NGOs 22 1 3 3 0 3 0 0 4 3 6 4 3 0 2 0 Governing structures 25 4 3 3 5 9 7 2 9 2 6 2 2 9 7 1 Private/Free market 10 1 0 0 7 4 0 0 1 1 0 1 0 0 1 1 Unclear 5 0 0 0 1 1 0 2 0 0 0 0 0 0 0 0 Motivational Codes III Women (no other) 7 2 1 0 1 5 0 0 0 0 4 0 6 0 0 0 Women in the G.S. 0 0 1 1 0 2 1 0 1 3 0 1 1 1 1 2 Rural women 0 0 3 2 0 0 0 0 0 1 0 3 3 1 1 2 People in poverty 2 0 1 0 0 0 2 0 3 0 0 0 2 0 0 0 Gender & sex minorities 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 72 Table 1.3.2. Equal participation and rights collective action frame heat maps, part 2 of 2 Participation Women's Rights Gender Minorities' + Emic Starts with Emic Rights as Freedom 19 37 40 44 3 29 20 38 47 Diagnostic Codes . . . . . . . . . Gender gap 0 0 0 0 1 0 0 0 0 Gender discrimination 0 0 0 0 6 9 0 0 16 Gender inequality 0 0 5 2 3 1 2 0 1 Patriarchy 0 4 2 1 4 1 2 3 4 Division of labor 0 0 1 0 1 1 0 1 0 Culture/Gender ideology 3 3 8 10 5 3 1 0 10 Systems of inequality & 0 3 4 0 8 4 0 6 9 oppression Power inequality 5 7 5 2 4 3 0 4 7 No power/Powerlessness 0 0 0 0 0 0 0 0 0 Disempowerment 0 0 0 0 17 0 1 0 1 Violence 1 1 0 10 2 0 0 1 14 Poor material conditions 1 1 1 3 15 4 0 3 1 Prognostic Codes Status in community 0 2 0 0 0 0 0 0 0 Well-being 2 0 0 2 11 3 1 1 8 Resources 8 7 2 7 22 4 5 1 19 Participation 5 13 1 13 19 0 4 3 1 Women's (human) rights 3 0 0 2 11 9 3 5 15 Freedom 0 2 0 0 2 0 0 0 2 Agency/Personal autonomy 2 3 0 7 8 3 9 3 17 Expanding possibilities 0 0 0 0 4 0 1 0 0 Gender equality/Gender equity 4 4 2 0 2 3 2 1 1 More power-over 2 3 0 0 6 0 0 1 0 Power-to/Power transformation 1 0 0 4 5 0 0 2 0 Power-within/Crit. consciousness 1 0 0 4 0 0 0 3 1 Transforming home relations 0 0 1 0 0 1 1 0 0 Cultural change 1 9 10 4 0 0 1 3 Motivational Codes I For development 2 0 0 0 0 2 3 0 2 Patriarchal bargain 0 0 0 0 5 0 0 0 0 Emic value 4 1 1 9 21 2 4 2 4 Alternatives for women 0 0 0 3 6 0 0 0 2 Alternatives to live more equitably 0 2 1 0 7 0 0 0 0 Alternative gender organization 0 0 2 0 1 2 0 3 3 Motivational Codes II People with power/privilege 0 6 2 11 2 2 0 1 1 Power-with/Collective solidarity 9 11 1 8 24 2 2 3 10 Women being empowered 1 2 0 5 9 2 4 1 3 Development orgs/NGOs 5 1 1 6 6 3 1 3 5 Governing structures 13 7 0 4 14 4 2 4 6 Private Free market 1 0 0 0 5 0 0 0 0 Unclear 0 0 0 0 0 0 1 0 2 Motivational Codes III Women (no other distinction) 2 0 0 1 4 0 1 0 0 Women in the Global South 0 1 2 3 2 1 0 2 1 Rural women 0 0 0 0 0 1 0 0 0 People in poverty/"The poor" 4 0 0 0 8 1 0 2 0 Gender & sexual minorities 1 0 0 0 0 0 0 0 5 73 Table 1.4.1. Agency collective action frame heat maps, part 1 of 2: agency heavy frames Beyond Strategic Emic > Alternatives Alternatives Balanced > Emic Strategic for Women for Women 12 2 21 5 50 35 42 45 54 Diagnostic Codes . . . . . . . . . Gender gap 0 0 1 0 0 0 0 2 0 Gender discrimination 1 1 2 3 0 1 3 0 0 Gender inequality 4 5 2 2 0 4 9 4 0 Patriarchy 0 3 2 7 0 8 5 28 0 Division of labor 0 16 0 5 0 4 14 9 0 Culture/Gender ideology 8 3 11 2 1 9 14 38 21 Systems of inequality & oppression 0 1 1 6 0 3 21 10 11 Power inequality 1 1 7 2 3 8 9 13 0 No power/Powerlessness 0 0 2 1 2 1 0 0 0 Disempowerment 3 0 3 0 0 0 0 0 0 Violence 0 1 1 3 0 0 2 7 2 Poor material conditions 0 3 1 1 0 0 6 9 0 Prognostic Codes Status in community 0 0 0 0 0 5 0 2 0 Well-being 1 1 1 1 0 0 3 19 2 Resources 7 3 1 1 0 3 9 19 3 Participation 0 0 0 0 2 2 10 27 0 Women's (human) rights 0 2 0 0 0 0 4 0 0 Freedom 0 0 1 0 0 0 2 17 0 Agency/Personal autonomy 10 2 9 3 2 5 18 94 7 Expanding possibilities 3 0 3 1 1 1 1 8 2 Gender equality/Gender equity 0 1 0 2 0 1 10 2 0 More power-over 0 1 1 1 4 1 7 6 0 Power-to/Power transformation 0 4 8 3 6 0 22 4 2 Power-within/Crit. consciousness 1 4 6 4 3 0 0 9 6 Transforming home relations 0 2 1 0 0 0 1 11 0 Cultural change 0 1 2 0 1 3 6 12 7 Motivational Codes I For development 0 0 0 0 0 0 0 0 0 Patriarchal bargain 0 0 0 2 0 0 3 0 0 Emic value 7 9 4 5 1 2 15 5 6 Alternatives for women 1 2 1 6 0 8 13 23 9 Alternatives to live more equitably 1 4 0 1 2 1 1 4 0 Alternative gender organization 6 4 5 1 2 1 5 6 5 Motivational Codes II . . . . . . . . . People with power/privilege 0 2 2 4 0 0 5 10 1 Power-with/Collective solidarity 1 5 8 3 4 3 12 34 16 Women being empowered 2 4 4 2 1 2 9 9 5 Development orgs/NGOs 0 2 6 6 0 1 8 9 3 Governing structures 0 1 0 0 0 0 6 2 0 Private/Free market 0 0 0 0 0 0 0 0 0 Unclear 0 0 0 1 0 0 0 0 0 Motivational Codes III Women (no other distinction) 2 0 4 0 1 0 0 4 1 Women in the Global South 0 3 0 2 0 2 3 5 2 Rural women 0 0 0 1 0 0 0 6 0 People in poverty/"The poor" 0 3 0 0 0 0 4 5 0 Gender & sexual minorities 0 0 0 0 0 0 0 0 0 74 Table 1.4.2. Agency collective action frame heat maps, part 2 of 2 Agency & Agency & Participation Structure 6 10 8 27 13 14 15 16 22 23 Diagnostic Codes . . . . . . . . . . Gender gap 8 2 1 4 1 1 0 1 0 2 Gender discrimination 19 1 6 8 2 1 3 2 3 1 Gender inequality 18 1 1 0 0 3 3 7 0 1 Patriarchy 5 0 5 1 0 0 0 2 0 0 Division of labor 12 0 2 1 0 2 0 1 0 0 Culture/Gender ideology 12 1 3 7 5 1 2 15 4 15 Systems of inequality & oppression 7 1 0 0 0 0 2 0 5 0 Power inequality 9 1 5 0 0 0 2 0 1 7 No power/Powerlessness 1 1 0 0 0 1 1 2 5 0 Disempowerment 6 0 0 1 0 0 2 0 1 9 Violence 11 0 1 0 1 1 2 9 2 1 Poor material conditions 15 1 0 1 3 0 8 0 0 14 Prognostic Codes Status in community 5 0 6 1 0 1 2 0 0 0 Well-being 16 1 2 9 2 1 2 2 11 3 Resources 29 6 15 14 14 8 13 10 17 27 Participation 22 15 3 3 1 0 4 2 10 14 Women's (human) rights 35 2 2 3 1 0 1 0 1 0 Freedom 0 1 0 0 1 0 0 8 1 2 Agency/Personal autonomy 22 12 9 5 23 3 5 10 8 51 Expanding possibilities 1 1 1 3 0 0 1 0 0 7 Gender equality/Gender equity 10 1 2 0 4 1 9 0 0 8 More power-over 2 5 0 0 0 0 3 4 0 2 Power-to/Power transformation 1 0 0 0 0 0 0 0 0 1 Power-within/Crit. consciousness 8 4 1 1 0 0 0 0 1 1 Transforming home relations 0 0 0 0 0 0 0 0 0 0 Cultural change 4 2 3 1 1 0 0 1 0 0 Motivational Codes I For development 0 0 0 2 4 2 20 5 12 31 Patriarchal bargain 0 0 9 1 0 0 1 2 0 0 Emic value 22 11 5 4 0 1 0 0 0 0 Alternatives for women 4 1 2 2 0 1 1 0 0 0 Alternatives to live more equitably 13 2 1 0 0 0 8 0 0 3 Alternative gender organization 7 3 2 0 0 0 0 0 0 0 Motivational Codes II People with power/privilege 7 1 0 0 0 0 1 0 0 0 Power-with/Collective solidarity 10 11 1 3 0 1 6 0 13 8 Women being empowered 7 5 0 1 0 0 5 1 2 5 Development orgs/NGOs 20 6 5 8 1 7 5 3 6 13 Governing structures 3 1 4 4 0 1 5 4 14 8 Private/Free market 0 3 1 1 0 2 5 0 6 4 Unclear 2 0 0 0 1 0 2 0 1 1 Motivational Codes III Women (no other distinction) 1 1 0 0 0 1 0 0 0 0 Women in the Global South 1 0 1 3 2 0 0 1 0 0 Rural women 0 0 2 4 0 0 0 0 0 0 People in poverty/"The poor" 3 1 2 4 0 2 3 0 5 2 Gender & sexual minorities 0 0 0 0 0 0 0 0 0 0 75 APPENDIX E: POWER CODES 1. Power inequality (diagnostic code): reflects “power over” conceptualizations of power (involving either or relations of domination and subordination/obedience) and is used when power inequality or dependency is identified as the major or sole problem in the context of empowerment. 2. More power over (prognostic code): also reflects “power over” conceptualizations of power (implying a zero-sum model of power) but as a solution; it is used when empowerment (or the goal) is social autonomy, or women/minorities gaining/controlling/having more power over sources of power that create structures of subordination. This can imply either redistribution and/or women gaining the power over others - not just their own- decisions. 3. Culture/gendered ideology (diagnostic code): captures the more hidden (non-decision- making) and internalized oppressions women face when overt use of “power over” is no longer necessary. This code is used when ideology, norms, and beliefs are identified as creating a culture/system where women/minorities have low social value. 4. Power within/critical consciousness (prognostic code): comes from a feminist understanding of women undoing negative social constructions by recognizing the structures/ideology that perpetuate their oppression. It is used when gaining critical consciousness or increasing "power within" is identified as important to women's empowerment. Also includes other people's consciousness raising of the systems that oppress women. 5. Power to (diagnostic code): used when the author refers to a Foucauldian understanding of power as relational/embodied experience and/or rejects the idea that power is either finite or binary and accepts the idea that there are multiple forms of power including transformative, generative, or productive forms - which is the solution to empowerment.21 6. Power with/collective mobilization (motivational code) is used when the author recognizes the importance of collective solidarity in achieving empowerment/fixing the issue. Women’s movements, organizing, and society at large are identified as having an important role to play in creating the conditions for change. 21 I do not automatically code “power to” when authors’ mention this phrase, as I find that some authors use it when they mean the power to do something (e.g., agency or resistance), rather than describing the way power operates relationally or generatively. 76 CHAPTER 3: IMPLEMENTING EMPOWERMENT FOR WOMEN IN INDIA: THE CASE OF SELF-HELP GROUPS AS A MEANS TO IMPROVE NUTRITION Introduction More than 194 million people in India suffer from food insecurity today, where the tangible consequences are disproportionately felt by rural women compared to men and those living in urban areas (FAO et al. 2020; IIPS and ICF 2021). Across the country, women in rural areas are often targets of development programs and policies, frequently in the name of “empowering” them (Cornwall and Whitehead 2007; Sanyal 2014). In particular, women’s self- help groups (SHGs) have come to dominate this development landscape (Brody et al. 2017; Kumar et al. 2021). Since the early 2000’s, the Indian government has promoted SHGs in their policies as a platform for empowering women, developing rural communities and economies, and strengthening grassroots democracy (Guérin and Kumar 2017). Today, more than 83 million households across India include SHG members involved in the government’s national initiative National Rural Livelihood Mission (NRLM) (Ministry of Rural Development 2023). NRLM’s mandate is to enroll one woman from every rural poor household into an SHG and reach 100 million households by 2024 (Ministry of Rural Development 2018).22 Moreover, NGOs have, and continue to play, a leading role in forming and managing SHGs (Sanyal 2014). The majority of SHGs consist of groups of 8 - 20 women engaged in activities for the purpose of improving livelihoods and empowering women, but NGOs have expanded SHGs’ potential to also act as a platform for supporting women’s leadership, creating 22 NRLM prioritized the inclusion of schedule castes (SC), schedule tribes (ST), particularly vulnerable tribal groups (PVTGs), single women and women headed families, landless and casual laborers, elderly persons, person with disabilities (PwDs), minority groups, and people living in remote geographies and in uncertain political environments (Acharya et al. 2020). 77 health and nutrition awareness, and addressing an assortment of social problems (Kumar et al. 2021; Nichols 2021a; Scott et al. 2022). At the same time, programs and policies claiming to “empower” women, including those that involve SHGs, are rarely scrutinized or comparatively analyzed to understand the way empowerment is conceptualized and then implemented (and evaluated) in practice. Even more concerning, conceptions, implementations, and measures of women’s empowerment in the Global South are often proposed by powerful stakeholders inside and outside of their countries, with the most influential sources coming from the Global North (Cornwall and Whitehead 2007; Rowlands 1997; Syed and Ali 2011). In essence, what is considered empowering for all women, including Indian women, has historically been decided by powerful groups in the Global North.23 In chapter two, I demonstrated the various meanings of women’s empowerment in the development discourse using frame analysis. I found that the “Gender Parity” collective action frame is not only the most popular women’s empowerment frame in the development discourse, but is also one dominated by authors in the Global North. One case that falls into this frame is the Women’s Empowerment in Agriculture Index (WEAI). In the context of agricultural change, the WEAI is the most known and used empowerment framework. Initially, the WEAI was created by individuals in the Global North - as a monitoring and evaluation tool for USAID’s (United States Agency for International Development) Feed the Future Initiative projects - to examine differences across countries and track changes in women’s empowerment (Alkire et al. 2013; Meinzen-Dick et al. 2019). A combination of the WEAI’s increased popularity, as well as 23 Some scholars point to the 1997 Micro-Credit Summit, held in Washington, DC, as the turning point for many development NGOs and governments (including India) to implement savings- and credit- oriented SHGs into their policies; reflecting Western ideas of economic empowerment and U.S. feminist movements in the late 1960s (Guérin and Kumar 2017; Sanyal 2014). At the same time, the formal formation of women’s SHGs in India can be traced back to 1972 - with the establishment of the Self-Employed Women’s Association (SEWA) founded by feminist Ela Bhatt in the state of Gujarat, India (Bhatt 1989). 78 criticisms of the instrument in recent years, led to the creation of the project-level Women’s Empowerment in Agriculture Index (pro-WEAI) in 2019. The WEAI “family of tools”24 has gained tremendous attention by academics and practitioners working on women’s empowerment in rural regions throughout the Global South: As of March 2022, the WEAI’s methodology has been used by at least 230 organizations in 58 countries (Myers 2022). This chapter continues to examine the ways women’s empowerment is interpreted by revealing how empowerment frames are conceptualized and then implemented into particular solutions on the ground. Specifically, this chapter offers a detailed study of three NGOs working together on a development project in India that combines two particular strands of development: nutrition and self-help groups (SHGs). This project was selected due to its pertinence in involving the creators of the pro-WEAI and its implementing organizational partners in India. An analysis of content related to the development project is supplemented with the interpretations and perspectives of practitioners from the three NGOs to present a rich empirical picture of the way women’s empowerment is interpreted at all stages of the development process (i.e., conceptualization, implementation, and evaluation). Analytical research that pays close attention to the ways women’s empowerment is constructed in organizational policy and implemented on the ground into programs is needed to better understand development interventions’ potential impact on women’s lives and gender relations (Ferguson 2010). Thus, the present study investigates how frames of gender empowerment lead to particular actions aimed to empower women in India. Specifically, to investigate these processes, I formulate the following three research questions: 1. How is empowerment conceptualized? 24 The WEAI family of tools includes the WEAI, Abbreviated WEAI (A-WEAI), pro-WEAI, and pro-WEAI+MI (project-level WEAI for Market Inclusion). The pro-WEAI+MI is the new name for WEAI4VC. 79 2. How is empowerment implemented? 3. Does a certain concept of empowerment lead to particular solutions on the ground? To empirically answer these research questions, I draw from the frame approach (outlined by Snow and Benford) to analyze content and interviews relevant to the case study.25 I use this methodological approach for two reasons. First, the frame approach is sufficient in answering the research questions because it is analytically capable of demonstrating the contested ownership behind the apparent consensus on women’s empowerment by drawing attention to the shared understandings and conflicts over meanings (Benford and Snow 2000; Gamson 1992; Wright and Reid 2010). Second, because the frame analytical approach allows for various frames to be operationalized and studied as both properties of organizations (located in their records, websites, and other content) and individual cognitive structures (located in the minds of individuals) (Snow 2004), I analyze both content published by the case study’s NGOs and interview stakeholders involved in the formation, implementation, and evaluation of the case study’s program. Thus, this methodological approach allows me to examine the shared and contested interpretations amongst various claims-makers in the way empowerment is conceptualized and implemented into practice. In the following sections, I demonstrate the shared and sometimes conflicting claims over the ways empowerment is conceptualized, operationalized, and implemented in the development project, and present these findings in a cogent way that answers the research questions posed above. I begin by describing the case study from which I draw these findings from, to provide context for this study’s analyses, followed by a description of my methodological approach to 25 The framing perspective was outlined in a sequence of papers by David Snow and Robert Benford in the social movements literature (Snow et al. 1986; Snow and Benford 1988, 1992). 80 answering these questions. I conclude with a discussion of the potential issues that can arise when empowerment frames do not align. Case Study This section provides context by describing the development projects (WINGS and FAAM), and three organizations (IFPRI, PRADAN, and PHRS) that make up this case study. It should also be noted at the onset, that this case study is chosen to understand the processes of framing and implementing women’s empowerment because of its pertinence in the context of India’s investment in women’s empowerment and SHGs, the dominance of the “Gender Parity” frame of empowerment in the development discourse, and the growing use of the WEAI family of tools in recent years. To begin, this case study is situated within a larger international project known as the GAAP2 (Gender, Agriculture, and Assets Project, Phase 2); consisting of 13 development projects led by the NGO IFPRI (International Food Policy Research Institute) (Malapit et al. 2019). IFPRI is explicitly a research and evaluation organization, but they partner with implementing organizations across the world (Meinzen-Dick and Quisumbing 2018). Thus, each of the 13 GAAP2 projects consisted of a collaborative effort between evaluator organization IFPRI and an implementer organization located in Asia or Africa. As part of the GAAP2 portfolio, the 13 projects were selected to develop, implement, and pilot the pro-WEAI, and use it to determine if each of the project strategies worked to empower women (IFPRI 2022). Precisely, they were selected based on their convincing strategy to empower women, their ability to assess women’s empowerment, and additional focus on either nutrition or agriculture (Johnson et al. 2018:6). One of these projects is WINGS (Women Improving Nutrition through Group- 81 based Strategies); representing the only GAAP2 project in India, with its “additional focus” centered around nutrition (IFPRI 2022; Meinzen-Dick et al. 2019). To be clear, WINGS is the name of the evaluation portion of this case’s project, conducted by IFPRI, and the reason why I initially selected this particular case study.26 Moreover, the WINGS evaluation assessed the impact of an intervention known as FAAM (Facilitated Action Against Malnutrition), an intervention implemented by IFPRI’s partner NGO, PRADAN (Professional Assistance for Development Action). PRADAN has decades of experience creating and strengthening over 65,000 SHGs across marginalized communities (low caste and particularly those with high tribal populations) with the aim of improving livelihoods (Kumar et al. 2021; Nichols 2021b; Prasad 2016; Scott et al. 2022). Further, PRADAN supports the institutionalization of SHGs in creating a representative democratic governance structure in which 1) the formation of 12 – 13 women living in the same village meet in their SHG, 2) representatives from these SHGs meet in larger village, panchayat,27 organization meetings (known as VO meetings), and 3) representatives from VOs meet at a block level federation (BLF) – where issues discussed in the SHGs are communicated to the BLF and acted upon (see Figure 2.1) (Nichols 2021a; Scott et al. 2022). Along with SHG formation, PRADAN’s standard activities aim to engage women in livelihoods and agriculture (new technologies, market linkages), leadership training, and discussions around gender issues (Kumar et al. 2021; Scott et al. 2022). FAAM is a nutrition intensive (NI) behavior-change communication (BCC) intervention that used PRADAN’s SHGs as a platform for information dissemination. It is housed under 26 I always admired IFPRI’s gender work, especially in their developments of measuring women’s empowerment in agriculture. And while I shared concerns with other feminists about the WEAI, I was excited to learn about the application of the new, project-level, WEAI (Malapit et al. 2019). 27 A panchayat is a small administrative unit comprising of several villages or one large village (Nichols 2021a). 82 PRADAN’s PoWER (Partnerships for Women’s Empowerment and Rights) project. FAAM was implemented in eight rural project sites across five states in eastern and central India (Jharkhand, Madhya Pradesh, Chattisgarh, Odisha, and West Bengal) (Kumar et al. 2021; Nichols 2021b). The NI BCC curriculum of FAAM was facilitated though a set of nine perspective-building (or story-telling) “micro-modules” in two phases, with the addition of a third review phase that consolidated earlier learnings into three modules (see Figure 2.1). While PRADAN was IFPRI’s official partner organization in the WINGS project, the FAAM intervention (BCC curriculum) was co-developed with PHRS (Public Health Resource Society), an NGO with expertise in building capacities for public health and nutrition action (Prasad 2016). Curriculum development for FAAM began with a community needs assessment between 2013 - 2015 in a few of the intervention sites to determine which topics around nutrition should be focused on (Nichols 2021b). To standardize and implement the BCC curriculum, PRADAN’s existing block program officers (one per block) were supplemented with a PHRS professional and 5 - 8 paid “mentors.” The block program officers and mentors would receive a three-day training (for each phase) by the PHRS professionals. The mentors would then supervise and provide three-day trainings (for each phase) and monthly block-level meetings to a larger volunteer cadre of VO-level SHG member women known as poshan sakhis28 or CV didis29 (2 - 3 per village). The mentors and poshan sakhis would then deliver the micro-modules to SHG members once a month at their regular weekly SHG meetings (see Figure 2.1) (Nichols 2021b; Scott et al. 2020). 28 Poshan sakhi is Hindi for “nutrition friend”. 29 CV is short for “change vector,” and didi is a colloquial word for “older sister” in Hindi. 83 Figure 2.1. PRADAN SHG institutions and FAAM intervention structure Source: Scott et al. 2022 Put simply, WINGS is an IFPRI evaluation project, while FAAM is the intervention (created and implemented by PRADAN and PHRS) that was being evaluated. The entire project (WINGS/FAAM) officially spanned from 2015 to 2020: For WINGS, IFPRI conducted three rounds of quantitative data collection and one round of qualitative data collection between October 2015 and March 2020; and the FAAM intervention was implemented in three phases between September 2016 and March 2020 (see Figure 2.2) (Kumar et al. 2021; Nichols 2021b; Scott et al. 2022). 84 Figure 2.2. Timeline of FAAM and WINGS activities Sources: Kumar et al. (2021); Nichols (2021b); Scott et al. (2022) Methodology Sample To demonstrate the ways women’s empowerment is interpreted (i.e., conceptualized and implemented) in this case study, I draw on content related to both WINGS and FAAM as well as semi-structured interviews with stakeholders involved in the conceptualization, implementation, and/or evaluation of the FAAM intervention (whether or not they currently work for the organization they were affiliated with at the time they were involved). Specifically, I examined peer-reviewed articles, conference papers and presentations, reports, newsletters, blogs, and YouTube videos publicly available and published by those affiliated with IFPRI, PRADAN, and/or PHRS pertaining to either WINGS or FAAM. In addition, I analyzed content that was not readily available to the public that was shared with me from interviewees, including WINGS evaluation plan documents and the modules implemented/making up the FAAM intervention created by PRADAN and PHRS. 85 To supplement and triangulate the data in this case study, I also conducted 10 semi- structured interviews (nine respondents, one follow-up) with stakeholders between December 2022 and February 2023. The interviews took place on Zoom with participants geographically located in either the United States (four respondents) or India (five respondents). Table 2.1 provides a snapshot of the final interview sample with the organization the respondent is most closely affiliated with professionally and the country where they were located at the time of the interview. I began selecting respondents using a stratified purposeful sampling method, identifying stakeholders from each of the three organizations involved in the WINGS/FAAM case study, to reflect a range of perspectives along the development process. I identified and contacted 11 potential interviewees via email and LinkedIn; if I did not hear back from a respondent (either accepting or kindly refusing an interview) I reached out a second time. Six out of the 11 potential interviewees were interviewed. I then supplemented this purposeful sampling procedure with a snowball procedure, interviewing suggested stakeholders involved in the development process (who were not initially identified). Out of the five potential snowball interviewees, I interviewed three.30 Table 2.1. Interview sample Organization Respondents in India Respondents in the U.S. IFPRI 1 4 PRADAN 3 PHRS 1 Employing semi-structured interviews, I began by asking participants to describe their role in the development project. While each interview differed slightly (given participants’ 30 The other two were suggested stakeholders involved in the funding selection process, and for fear it might be interpreted as a conflict of interest, these two were not interviewed. 86 positionality in relation to the development project),31 I used a general set of questions asking participants the following: to describe the goals of the program (including identifying the problem/issue being addressed); if these goals changed at all between the conceptualization stage and during implementation; how empowerment was defined or understood; if the intervention aimed to empower women, and if it succeeded from their perspective; what the criteria for success is when evaluating the intervention; what criteria constitutes indicators of women’s empowerment; who the target population was and why; and if any differentiation among women recipients were taken into account when evaluating the impacts of the program. I ended all interviews asking if they have preferred gender pronouns (if so, what they are), and if there was anything else they thought I should know that I did not ask. Each of the 10 interviews ranged between 44 - 94 minutes, with the majority lasting 60 minutes. All 10 interviews were recorded on Zoom and transcribed verbatim to ensure respondents’ words were accurately captured and analyzed. The aim of qualitative research is to discover relevant themes that emerge from the data. Thus, the most important and commonly used factor for determining the sample size in qualitative research relies on the concept of saturation. Saturation is defined as the point at which data collection offers no new or relevant information. Social scientists suggest theoretical saturation occurs within the first 12 interviews and that an adequate sample size for interviews is between 5 - 50 participants (Dworkin 2012; Guest, Bunce, and Johnson 2006). I found that between the supplemental use of the content and in-depth interviews of nine varying perspectives, I reached theoretical saturation at 10 interviews in this case study. 31 Interviews varied because of each person’s role and organizational affiliation. For example, a research assistant at IFPRI that solely worked on the WINGS evaluation was given a different set of questions than a senior-level implementer of the FAAM program, because of their varying abilities in answering certain questions due to their differing involvement in the program. 87 Analytical Approach In this study, I am interested in the framing processes of women’s empowerment in India, including capturing the ways conceptualizations of empowerment lead to particular solutions (e.g., development initiatives) aimed to empower women on the ground. To empirically investigate these processes, I use frame analysis to guide the formulation of my research questions, what questions to ask interviewees, and how to code this case study’s qualitative data. The framing perspective was born out of a need to examine and discuss processes of aligning (or fitting) individuals’ lives within social movements (Snow et al. 2014). Foremost, the framing perspective understands frames as an outcome of social interactions that is created over a shared meaning and reading of social life and situations (Snow and Benford 1992). Because frame analysis prevents researchers from treating meanings as a given, it is a suitable analytical approach for navigating the contested interpretations of empowerment in this case study. Furthermore, frame analysis provides an analytical approach to understanding the ways language translates to mobilizing people to action. Benford and Snow (2000) suggest that the way problems are framed (diagnostic frame) determine which solutions are possible or even considered over others (prognostic frame) (Benford and Snow 2000). In this respect, frames can be used to explain why certain programs and policies are implemented and others are not, including how framing women’s empowerment at the conceptual level shapes possibilities for program design and outcomes at the practical level. Therefore, to answer my research questions, I use the frame analytical approach to guide the creation of my interview guide. Because the frame analytical approach understands that the diagnostic framing task involves problem identification and forms the basis for the prognostic framing task of solution identification (Benford and Snow 2000), research question one (how 88 empowerment is conceptualized) should align with the diagnostic framing task, and research question two (how empowerment is implemented) should align with the prognostic framing task. To inform research question one, I asked interviewees questions related to the identified problem or issue that the project (e.g., WINGS, FAAM) was focused on addressing, the goals/objectives of the project, and how empowerment was assessed. Secondly, to answer research question two, I asked questions related to the prognostic framing task. This included asking if and how the project tackled the identified problems, including questions about if the intervention aimed to empower women. I then probed (if they did not already address) the ways the program did empower women and how empowerment was assessed. Lastly, the third research question (how frames of empowerment lead to certain solutions and possible outcomes) is informed by drawing connections to all these interview questions. Interview transcriptions and all additional content were uploaded to MAXQDA for analysis using inductive coding. The qualitative data were read and notes were taken on emerging themes guided by the diagnostic and prognostic functions of the frame analytical approach. Particular attention was paid to how empowerment is conceptualized and operationalized, how empowerment was implemented in the FAAM intervention, and how frames of women’s empowerment (at the conceptual level) lead to certain solutions implemented (at the practical level). The findings presented next consist of summarizing the dominant themes that emerged from this process. Findings In this section, I begin by answering this study’s first research question by reporting the ways empowerment is conceptualized across stakeholders in this case study. While I initially selected this case study because WINGS was touted as one of the GAAP2 projects that had a 89 “convincing strategy to empower women,” tasked with implementing and piloting the pro- WEAI, with only “an additional focus” on nutrition (Johnson et al. 2018), some interviewees involved in the project suggest that maternal health and nutrition was the project’s main focus and women’s empowerment was conceptualized as a means to this end. I describe this discovery through my interviews and content analysis to further elaborate on the contestation over ideas that women’s empowerment may be something separate (and sometimes inferior) to women’s nutrition. I then turn to the second research question by describing how empowerment was implemented as a means to an end into the FAAM intervention (i.e., solution), specifically in the form of women’s SHGs. Lastly, because empowerment is found to be a platform for delivering goals centered around nutrition, the final research question (i.e., does a certain concept of empowerment lead to a particular solution), draws attention to the factors that informed stakeholders’ solution to women’s empowerment. Research Question 1: How is Empowerment Conceptualized and Operationalized? One of the most important findings in this case study is that empowerment is interpreted as a pathway to achieve goals related to women’s nutrition. In other words, empowerment is a means to an end. Specifically, empowerment is conceptualized as a process that is experienced through women’s involvement being with other women, that leads to certain outcomes; the end outcome (or goal) in this case, is changing women’s nutrition behavior. Despite this shared conceptualization, I also discover contention around whether or not women’s empowerment may be something separate and secondary to women’s nutrition and consequently, differences in the way empowerment is defined and operationalized by evaluators and implementers of the development project. In the following section, I provide examples of WINGS evaluators (IFPRI affiliates) and FAAM implementers (PHRS and PRADAN affiliates) 90 communicating the same diagnostic frame (i.e., problem identification) and the way empowerment is conceptualized as a means to fixing the problem – while also addressing the caveats in their operationalizations and interpretations by addressing the separation and priority of the “means” (i.e., empowerment) in the context of the FAAM program. Conceptualizing empowerment as a means to an end is identified throughout the interviews in various ways. To illustrate this point, I first need to explain how the diagnostic framing of the project was similarly discussed by implementer and evaluator interviewees. Regardless of whether the interviewees were talking specifically about FAAM or WINGS, everyone identified malnutrition as the problem the project was focused on addressing. This finding is important because it is also made clear that the two projects (FAAM and WINGS) are different in their objectives due to FAAM being the intervention, and WINGS being the evaluation of the intervention. Nonetheless, I find that both parties often refer to FAAM as a “nutrition intervention” or “nutrition program,” similarly identify malnutrition as the problem, and share the intended outcome of improved nutrition. For instance, implementers describe that the goal of FAAM was to add BCC into women’s SHGs with the hopes of changing women’s practices or behaviors, including nutrition- related practices like consuming a diverse diet. A PRADAN affiliate describes the intention of both FAAM and WINGS precisely by saying that in FAAM, “the focus was on implementation of nutrition interventions and the goal of the WINGS project was to do research on the implementation.” She then describes the goals of the FAAM intervention, referring to it at as a nutrition program, saying, “In the nutrition program, the aim was to lay behavior change communication as part of an input in women's self-help groups. To bring about more awareness about the importance of 91 consuming a balanced diet, which actually leads to adoption of various kinds of practices, which leads to women actually adopting a more balanced and diverse diet. We were looking at - the final outcomes were consumption of a diverse diet particularly by women, through letting BCC interventions in self-help groups” (Aditi, line 37).32 IFPRI affiliates similarly describe that the main goal of the WINGS project was to understand what the added impact of layering NI BCC onto women’s SHGs had on nutrition outcomes. At the same time, evaluators’ explanations often included language that implied a certain hierarchy to the outcomes of the program. Sometimes interviewees would say nutrition was “primary” and women’s empowerment was “secondary”; other times they would say women’s nutrition was “primary” and children’s nutrition was “secondary” (without mentioning empowerment at all). By and large, evaluators made it known that women’s nutrition was the primary outcome and separate from children’s nutrition and women’s empowerment. Swapna, an evaluator of the WINGS program uses this language in explaining WINGS: “The main question I think we were asking is, what is the added impact of these nutrition- intensification activities layered onto a standard savings and credit micro-finance-based self-help group model? What is the added impact of those activities on health and nutrition outcomes? We had a certain set of key primary outcomes, both at the level of the woman, because PRADAN was very clear that the woman was kind of the intended beneficiary and recipient of their programming. So they are very centered on women members of SHGs. But then you also have secondary outcomes that looked at health and nutrition outcomes within the same household, of younger children, and so we measured things like dietary diversity of the child, anthropometry of other members of the household” (Swapna, line 9). 32 All interviewee names are pseudonyms. 92 To further understand the diagnostic framing tasks of the project, I asked interviewees what the main problem their organization was focused on fixing in the context of the FAAM/WINGS project. In an interview with an IFPRI affiliate, she found my question “strange” namely because “The WINGS project sort of has its own objectives, and its objectives were really to understand the factors underlying maternal and child nutrition” (Maya, line 34). While Maya explains that WINGS had separate objectives from those of FAAM, she ends up highlighting the way evaluators came to identify nutrition as the end goal and empowerment as a means to this end when I further prompted her by asking what she saw as the problem that was being fixed in the intervention: “The reason why I say it's a very odd question is that - I mean, of course the problem is the high rate of child malnutrition in India, and how the disempowerment of women contributes to that. But I think what was important in the WINGS project is that it's also focused on maternal nutrition, because a lot of nutrition projects only focus on the child but not the woman. We know a lot about child nutrition and how to prevent it. We know very little about the role of women's empowerment in her own nutrition and how that contributes to child nutrition as well” (Maya, line 39). The example above highlights two mirroring assumptions that are made in the WINGS evaluation: women’s disempowerment contributes to the problem of malnutrition - and conversely - women’s empowerment contributes to nutrition. As a result of this diagnostic framing of the problem, women’s empowerment is conceptualized as a means to improving nutrition. In other words, women’s empowerment is a method for achieving nutrition outcomes. Evaluators’ conceptualization of empowerment as a means is routinely made in their descriptions of the project, leading to logical steps suggesting that the project is first and 93 foremost about nutrition, and empowerment is a means to this end. First, evaluators tend to use women’s empowerment and SHGs synonymously to describe the way it is ultimately understood as a platform or mediator affecting nutrition outcomes. Second, because women’s SHGs are a normal part of PRADAN’s development activities, the added nutrition component of FAAM (i.e., NI BCC and nutrition outcomes) led evaluators to suggest that the project was ultimately targeted at improving nutrition – not necessarily empowerment. Thus suggesting that empowerment is ultimately a process that leads to improved nutrition. “The core piece of the WINGS project was not empowerment, it was this set of health and nutrition interventions that were being developed, and testing whether or not they work. But like I said before, empowerment was an underlying cross-cutting theme across the whole thing. We understood by virtue of using women's self-help groups, by trying to engage women in community decisions, like the whole project is in a sense on a bedrock of empowerment. The women's SHG movement is designed to empower women …. Like you can't say it has nothing to do with empowerment. But on that base, which is very empowerment centric, and understand that empowerment is both an outcome as well as a mediator and a pre-requisite, like it's everywhere! The specific interventions being studied were health and nutrition interventions. There was nothing about the programming that PRADAN was doing that was targeted at improving women's empowerment. The programming was targeted at improving health and nutrition” (Swapna, line 91). While implementers’ descriptions do not suggest empowerment is conceptualized as solely a means to end, they do share similar diagnostic frames and come to conceptualize empowerment as a means but through a different process. To begin, PRADAN and PHRS affiliates share similar diagnostic framing of the project as IFPRI by first identifying the problem 94 as malnutrition among women and children: things like lack of consumption of a balanced diet and less than optimal weight of women. At the same time, while they do not explicitly name women’s disempowerment as the source of the problem, implementers stress the ways gender norms contribute to women’s malnutrition. For example, Aditi, a PRADAN affiliate, mentions contributing factors of the problem including women doing nearly 70 percent of farm labor, women eating last in the family, early marriage, early and repeated pregnancy, and gender taboos like the following: “During menstruation women are not allowed to enter the kitchen, and they actually stay outside of the house in a separate … not really fully constructed room, sometimes sharing it with a pair of cows, buffalos, or goats, et cetera. So that kind of a thing … coupled with everything leads to a drain on her health and well-being” (Aditi, line 45). In addition, implementers similarly describe empowerment and SHGs synonymously as a process that leads to results. In other words, empowerment is again conceptualized as method for achieving goals. In this case, empowerment is described by implementers as embodying a process of collectivization – that leads to action. Specifically, empowerment is embodied in FAAM’s feminist approach of using participatory learning and enacting a transformational process where women come together in learning the sources of oppression in their lives and acting on those social, economic, and political structures collectively to make a difference in their own lives. In describing how FAAM worked to empower women, Gita, a PHRS affiliate, says: “In terms of a vision and the comprehension, it's all meant to be [empowering] and that's why the method is participatory action, because it's meant to be transformational. It's not just a kind of transference of information to get people to do something better. It goes far 95 beyond that. It's really a conscientization of women's groups to understand the discrimination they're facing as a collective, not just a women's collective, but maybe a rural or tribal, all those intersectionalities that apply to the communities we're working with. They're all poor, they're tribal, there may be caste issues, and they're women, and … historically … there has been a degree of social injustice in the sense of very poor access to resources, and so on. The curriculum was very much informed by not just the technicalities of nutrition or just, "okay let me tell you what to do to get better nutrition," but it's really to understand the group causes, which is a social, economic, or political causes of the situation and the context we were finding ourselves in, and then to be able to take action on all those things. The action was comprised of actions at the level self, action at the level of community, and action at the level of systems, or the state. The program was really intended to mobilize women. To be able to work with women to get some sense of understanding of health and nutrition from the point of view of being able to improve things for themselves, improve things for their community, but also to create pressure upon the government and the state to perform better on their behalf” (Gita, line 23). Throughout her interview, Gita explains that ultimately, the processes of women coming together in their SHGs around NI BCC creates an environment for women to mobilize and collectively make a difference. This transformational process, she says, is the most important part and declares, “The process is always much more important to me than necessarily the ends. I mean, the means are very important” (Gita, line 174). Her account highlights the shared conceptualization between implementers and evaluators of understanding empowerment as a means by which women come together as a collective; while at the same time, begins to show 96 some of the differences in placing “the means” – or empowerment – as the primary focus of the project. Differences in Defining and Operationalizing Empowerment So far, I have highlighted the major similarities in how empowerment is conceptualized as a means and the shared diagnostic framing between implementers and evaluators. I now turn to findings that showcase some differences in the way that PRADAN and PHRS vary from IFPRI in the way they operationalize empowerment. Namely, I describe how 1) PHRS and PRADAN do not see a clear distinction between nutrition and women’s empowerment like IFPRI does, and 2) IFPRI operationalizes empowerment in terms of agency, whereas PHRS and PRADAN operationalize it as achievements. When asked how empowerment was defined or understood in the project, all IFPRI affiliates gave consistent answers, all pointing to the use of the WEAI and its associated indicators related to agency. As an empowerment analyst in the WINGS evaluation, Maya defines empowerment in her description of what empowerment must look like in a program: strategies that lead to women improving or acting on strategic decisions, and not by simply including women’s nutrition into the program objectives. She adds that in order to actually evaluate empowerment, you need indicators specific to empowerment: “To be able to evaluate whether projects achieve their empowerment objectives, you need to have indicators which are empowerment indicators” (Maya line 75). All IFPRI affiliates, including Maya, said that the evaluations on empowerment for WINGS used the A-WEAI (Abbreviated Women’s Empowerment in Agriculture Index) at baseline and the pro-WEAI at midline and endline. As a result of using the WEAI, empowerment 97 is operationalized in terms of agency. Maya explains why empowerment is measured as agency using the WEAI, stating: “The WEAI and its associated family of indicators was inspired by Naila Kabeer's idea of empowerment, where it’s agency, resources, and achievements. Our motivation was that there are so many measures of resources and many measures of achievements, so we focused on measuring agency. It was agency that was more difficult to measure, so that's why the WEAI focuses on agency. But there were of course other indicators collected in WINGS and all the projects in GAAP2 because we have to know, is it a nutrition project? We have to know whether it affects nutritional outcomes.” (Maya, line 80). As a secondary outcome to women’s nutrition in the WINGS evaluation, I find that women’s empowerment, measured using the WEAI, is operationalized separately from IFPRI’s primary interest in evaluating women’s body mass index (BMI). All IFPRI interviewees clarify that women’s nutrition is considered a separate outcome and not part of women’s empowerment. For example, when I asked an IFPRI affiliate if this was the case, she answered: “Yea, yea, yea! Absolutely. It's an anthropometric outcome. Women's BMI, and then child's anthropometry, height, weight, z-scores, weight-varied z-scores, and then dietary diversity, both at the level of the woman as well as the level of the child. Those are all health and nutrition related outcomes” (Swapna, line 17). Thus, among evaluators of the project, women’s empowerment is understood as something separate from women’s nutrition: where women’s empowerment is operationalized as agency and effectively separated from women’s nutrition - operationalized as anthropometric measures including BMI and dietary diversity. 98 At the same time, this distinction and separation of women’s empowerment and nutrition is not something that is shared with implementers of the program (PHRS and PRADAN). For example, in one interview with an implementer, she shares that she does not make this conceptual distinction stating, “Even just looking after your own health, or understanding your own food and your own health a little bit better, doing better by your child, that's empowering. I don't make this distinction between talking directly about gender discrimination for example … being empowering … [and] … health and nutrition work not being empowering. I think they're all empowering” (Gita, line 123). In addition to not sharing the same ideas about empowerment being separate from nutrition, I also notice that among the implementers, there does not seem to be a consistent definition of empowerment like IFPRI affiliates have. There are two individuals (one from PHRS and one from PRADAN) that mentioned “decentralizing power” in their definition of empowerment, but they are the exception. In general, PRADAN staff were not as confident nor consistent in suggesting there was one definition of empowerment used in the FAAM intervention. Despite this, implementers similarly operationalize empowerment in terms of achievements (or end goals) – rather than agency. For instance, in an interview with an implementer of the FAAM project I asked, “How was empowerment defined or understood in these programs?” We both paused for two seconds, so I continued to ask, “Was it defined?” The interviewee began, “I'm not very sure whether this was defined or not. But the way the modules were being run … I was majorly involved in that aspect. What was happening was these didis were somewhere taking charge. Each module had a certain output. For example, the first module 99 talks about Sony and Madhu, which talks about the disadvantages or consequences of child marriage. It had an outcome saying, now the SHG members or the VO members need to be very careful and need to take certain steps where child marriage should not happen in the villages” (Rajesh, line 26). Rajesh continued to describe the other ways women of the village were empowered by the programs including didis stopping child marriages, women getting to know about the entitlements the state provides them, women coming into dialogue with panchayat (VO-level) about nutrition issues, crop diversity, and women having teen rangon ki thali33 (or diet diversity). This interviewee, like other PRADAN affiliates, operationalize empowerment in terms of the end goals the program intended to have; in other words, achievements. For instance, in a separate interview with another PRADAN affiliate, Disha explained the same examples as Rajesh did when I asked her what the criteria for success was in evaluating the FAAM program. “We were interested in three to four things, which we identified [as] critical. End of early marriage … or community taking action against early marriage. Another very important part was diet diversity. Number three was nutrition sensitive agriculture, because we were intervening in livelihoods. So what kind of crop diversity? … How [is the] community producing or cultivating diverse food from their own farm? And fourth indicator was how many … SHGs at village level … [are] taking up agenda of nutrition and health in their meeting … at continuous basis?” (Disha, line 122). What becomes clear in the interviews with the implementers of the FAAM program is that despite not having a shared definition of empowerment, all of them identify certain achievements as indicators of women’s empowerment. Specifically, they were all achievements 33 Teen rangon ki thali was a focus of a FAAM module that translates to food which is of three colors. In the module it specifically taught women that the three colors are green, white, and yellow. 100 that break norms impeding women’s nutrition. Aditi, a PRADAN affiliate, explains that while the usual definition of empowerment aligns with agency, in the context of the FAAM program (i.e., nutrition context) empowerment was more specifically focused on examining the norms that impact women’s health and nutrition - like early marriage, early pregnancy, and even violence: “The overall construct of women's empowerment is of course enhancing agency of women to determine whatever her destiny, and being able to exercise more control over her life. In the nutrition program, it was more specific around examining norms related to things that impact her health and nutrition status. Things like examining early marriage, early pregnancy, even violence, and how those impact nutrition. In everything we were coming with a construct of relating it to one's nutrition and health status.” (Aditi, line 69) In effect, by defining empowerment in terms of women breaking norms that impede their health and nutrition, empowerment is operationalized in terms of the achievements that showcase this social change related to women’s nutrition. Consequently, for implementers of the FAAM program, the line between women’s nutrition and her empowerment is blurred and the two concepts become one in the same in their operationalization. Notwithstanding, the findings thus far have shown the ways implementers and evaluators similarly identify the problem being addressed as nutrition and correspondingly conceptualize empowerment as a means to this end. I now turn to the ways these concepts of empowerment were implemented in practice. Research Question 2: How is Empowerment Implemented? The findings thus far have shed light on the diagnostic frame of the project, highlighting the ways empowerment is ultimately conceptualized and operationalized. In this section, I examine and describe the particular ways women’s empowerment was implemented (or a part 101 of) the FAAM program. In doing so, I shift the analytical focus from the diagnostic frame to the prognostic frame of this project by answering the following question: as a solution to the problem of malnutrition, how was empowerment implemented on the ground? As I described in the case study section of this chapter, the FAAM program itself consisted of many modules and activities. Both WINGS evaluators and FAAM implementers similarly say that the primary intervention of the project was the NI BCC modules layered onto SHGs: “The BCC communication in the SHG, now that was the primary intervention in the FAAM program or the WINGS program. So was there adequate layering in the SHG meetings of the different modules?” (Aditi, line 55 in follow-up interview). While the intervention itself consisted of two things: modules and SHGs, evaluators and implementers suggest that the “empowerment part” of the program came largely from the SHGs themselves. When asked if (and how) the intervention aimed to empower women, interviewees acknowledged that there was a module specific to women’s empowerment that included micro- modules on women’s rights and entitlements. However, their answers largely centered on the function of the SHGs. Thus, evaluators and implementers understand that in the FAAM program, empowerment was primarily implemented in the form of SHGs. Specifically, I find that interviewees describe the process of coming together (sometimes referred to as “capacity building”) as empowerment in practice. For example, an implementer describes the ways the modules acted as an entry point to the practice of empowerment – describing as the process of coming together in their SHGs – and using this participatory learning process (i.e., capacity building) to take liberating actions in their lives. “Empowerment doesn't come from just the modules. I'm seeing the empowerment comes from the practice, the action. The modules are there to engender action. It's that process that 102 is empowering, the process of coming together on health and nutrition. Health and nutrition is just one entry point. It could have been education, it could be livelihoods, it could be land, it could be anything. The empowering thing is to collectivize, to mobilize, to take action, and to win! To make things happen, and to see your own power. To be able to make things happen, make things happen within your own self. We've had examples of that. Within your own families, within your own community, meaning helping each other, and also forcing the system to work for you. That's what's empowering. It's not just the curriculum, it's just a starting off point. Because the curriculum is for what? It's for participatory learning and action” (Gita, line 129). Evaluators shared similar recollections on the ways empowerment – in the form of SHGs - were part of the FAAM program by describing the ways SHGs are a process of women coming together. For instance, Priya, an IFPRI interviewee, discusses evaluators’ ideas regarding the empowering effects SHGs have as a platform (and process) for bringing women together. In this interview, like many other IFPRI interviews, there seems to be a shared understanding that SHGs translates to empowerment – or at least has empowering effects: “The intervention that we were evaluating was a nutrition intervention. It was basically just giving information around nutrition behaviors. We did not expect it to [empower women] because it was primarily providing information. But! The women's self-help group platform itself, we had thought that that the basic platform itself would be empowering for women. Coming together in a group, even if you're not just doing the saving and lending activities, you do get access to credit, which is not something that is easy to get in the rural setting anyways, let alone for women. In that sense … it would lead to some empowerment, but not to the intervention per say that we were evaluating. Even in the final impact assessment 103 we don't find any impacts on empowerment of the intervention per say. But we did do some analysis looking at women who are members in a women's group, compared to those who are not, and we find that women who are group members are more likely to be empowered” (Priya, line 44). Priya’s description above ends by highlighting IFPRI’s research using the WEAI and pro- WEAI that found empowerment did not increase based on women’s involvement in the FAAM program (measured as agency), and that women in SHGs were more empowered (i.e., had more agency) compared to than those who are non-members (Kumar et al. 2021). Their finding suggests that IFPRI’s belief that women’s empowerment (measured as agency) would not be affected by the BCC curriculum alone was confirmed in their analysis of women’s empowerment using the WEAI methodology. I suspect this finding may have led IFPRI affiliates to come to the consensus that SHGs were the only part of the program that embodied empowerment in practice. While IFPRI affiliates show consensus that empowerment was solely implemented in the form of SHGs, the perspectives from implementers suggest a slight caveat. Interviewees from PRADAN and PHRS describe that while SHGs are indeed the main way empowerment was implemented in the program, they also describe the modules as being empowering as well. What is interesting is that it is not only the module specific to women’s rights and entitlements that is described as being empowering, but the way each module resonated with women’s stories as making the modules empowering. One PHRS professional that worked on creating the modules describes this, while hinting at the contested idea that SHGs are the only part of the program that made FAAM empowering: “I mean, IFPRI was always saying to us … “Why are you making this story?” That's where you'll see the empowerment bit, the stories. “So why are you making them so complex? 104 Why you bringing all these elements? Why so much nuance?” But then the women, they really felt, “This is my story. How is it that you're telling my story?” Because they were built out of that ground experience and the identification with that story” (Gita, line 143). I believe that implementers’ understanding that empowerment was also implemented as modules stemmed from their operationalization of empowerment as achievements and not something separate from women’s nutrition. Namely, because empowerment is operationalized as achievements that illustrate social change related to women’s nutrition, NI BCC that resonates with women’s lives are also considered empowering in practice. This point is made full-circle in an interview with a PRADAN affiliate, that describes some of the factors impeding women’s nutrition and empowerment, including women eating last in the family and not having nutrition information that is relatable to them: “In many households, women tend to eat last in the family. Lack of awareness of what needs to go into a balanced diet. Although, the government has been doing a lot of promotion, but still … poor households may not be able to relate to much of that information, thinking that kind of consumption is possible only by affluent households. They kind of take it as something the government is doing but not really possible to be done by them” (Aditi, line 45). PHRS’ earlier account highlights the ways the modules in the FAAM addressed the problem of women not having relatable nutrition information by telling stories that resonated with them. Implementers often describe how modules would include names familiar to them in story-telling issues related to women’s nutrition and would allow women SHGs to select the specific action items to address. In addition, women eating last in the family is mentioned by all PRADAN affiliates as a norm that was tackled in one of the modules (apart from women’s rights 105 and entitlements). I highlight this particular issue, of women eating last, not only because it is mentioned by all PRADAN affiliates as an important issue to tackle but perfectly highlights how implementers blur the lines between women’s nutrition and empowerment in practice. I now turn to the final research question to fully understand if a certain conceptualization of empowerment in this development project led to the solutions mentioned in this section. Research Question 3: Does a Certain Conceptualization Lead to Particular Solutions? To evaluate the final research question of this study, (i.e., does a certain conceptualization of empowerment lead to particular solutions), I tie together the findings presented thus far – and add a few clarifying points – to present the way empowerment’s conceptualization led to the particular solution (implementation) of empowerment in practice. Specifically, I argue that when empowerment is conceptualized as a process or means, it is implemented as a medium for specific goals to emerge. I begin this section by presenting the major findings of this research question by illustrating the relationship between the diagnostic frame (research question one) and the prognostic frame (research question two) of the FAAM/WINGS project. I do this by first drawing from the interview findings with WINGS evaluators, followed by FAAM implementers. I then turn to address the influential factors (e.g., diagnostic framing, resources, funders) that informed this particular solution/implementation of women’s empowerment. In doing so, I hope to inform the realities of why certain conceptualizations of empowerment may (or may not) neatly align with solutions when implemented on the ground. As I showed in research question one, evaluators conceptualize empowerment as a means to an end (i.e., nutrition). In several interviews with WINGS evaluators, empowerment was explicitly named as a pathway to the primary outcome (i.e., improved nutrition) and not 106 necessarily something that was thought to improve due to the intervention. For instance, when Priya, an IFPRI affiliate, suggested empowerment was secondary to the goals of nutrition in her interview, I first echoed, “So it kind of sounds like women's empowerment may have been secondary to the outcomes of women's BMI, dietary diversity.” When I asked her if this was always the case or if she saw a shift at some point throughout the project she replied, “No, I think we started out with the nutrition outcomes as the primary focus of the intervention. They were always the primary outcome, and women's empowerment was kind of part of an intermediate outcome or something that would also be a pathway. Given that it's being delivered through women's groups, and women's groups do have positive implications for women's empowerment, we thought it would help in getting to our final goals of the improved nutrition outcomes. It was thought of as a pathway to the primary outcome, rather than something that we thought would move because of the intervention” (Priya, line 56). Priya’s description perfectly summarizes some of the major findings of this study by demonstrating the relationship between the diagnostic and prognostic frame of the project. First, she highlights the diagnostic frame by revealing that WINGS evaluators conceptualized empowerment ultimately as a pathway to the primary end goal of improved nutrition. She does this by first identifying nutrition outcomes as the primary focus and outcome of the intervention. And while she initially identifies empowerment as both an intermediate outcome and a pathway, she concludes saying it was thought of as a pathway to the primary outcome (i.e., nutrition) – adding that empowerment was not something they expected would improve due to the intervention. Like the findings presented in research question one, the principal 107 conceptualization of empowerment is identified: women’s empowerment functions first and foremost as pathway to the primary end goal of nutrition. Equally important, her account reiterates the findings related to the prognostic frame in research question two, while simultaneously demonstrating the correlation between the diagnostic frame (i.e., conceptualizing empowerment as a means) and the particular use of empowerment in practice (solution to the problem of malnutrition). Earlier, I revealed evaluators’ identification of empowerment as solely being implemented in the form of SHGs. Here, SHGs are described as having the same function as empowerment: they are pathways to improving nutrition outcomes. Taken together, empowerment and SHGs are not only conceptualized as a pathway or means to improving nutrition, empowerment literally took the form of SHGs in practice. Thus, I find that the diagnostic frame of empowerment (i.e., as a means) and the prognostic frame of empowerment (i.e., SHGs) are neatly aligned in this case. In a similar vein, FAAM implementers also viewed SHGs as the solution based on their conceptualization of empowerment as a means or process. At the same time, they also identified the story-telling modules that made up the BCC as also being a part of the solution, or way empowerment was implemented. Despite this slight caveat between implementers and evaluators identifying BCC as having the potential to empower - and representing an empowerment process in practice - implementers’ diagnostic frame still neatly aligns with their prognostic frame. In the next section, I explain how, like evaluators, implementers’ diagnostic frame informs (and aligns) with their prognostic frame. I now turn to this analysis and describe the other factors stakeholders used to inform their framing of women’s empowerment in this case study. 108 Factors Informing the Prognostic Frame Compared to evaluators that identified SHGs as making up the “empowering part” of the intervention, implementers had an additional component they identified as being empowering in practice, namely the modules. Notwithstanding, like evaluators, implementers’ diagnostic frame informed (and aligned with) their prognostic frame. Namely, their broader conceptualization of empowerment led to the alignment and identification of the modules as a solution (or an example of empowerment in practice). First, implementers’ framing of empowerment as a means was more broadly understood as capacity building (that would lead to collective mobilization). Secondly, empowerment as a means was identified as being as important as the ends itself. Relatedly, implementers did not conceptually nor operationally separate nutrition from empowerment. Rather, empowerment and nutrition were one in the same, given equal importance; leading to the two to be operationalized together as achievements (representing transformational actions). This conceptual difference may also explain implementers’ accounts (see Gita’s quote in research question two) of IFPRI not seeing eye-to-eye on the role the complexity of the modules had in being empowering, and therefore leading to differences in identifying the modules/story-telling BCC as also being described as empowerment in practice. I find that because evaluators and implementers’ diagnostic frames neatly aligned onto their corresponding prognostic frames, alignment in their measurement of the impact of the FAAM on women’s empowerment is not aligned. This non-alignment can be seen in their differences in operationalizing women’s empowerment: where implementers’ operationalized empowerment in terms of achievements and evaluators’ operationalized empowerment in terms of agency using the WEAI. Implementers’ spoke about this concern in terms of the FAAM program not addressing/impacting certain indicators in the WEAI, while also suggesting 109 fundamental differences in the conceptual frame of empowerment that was absent in – and arguably opposed to – the WEAI WINGS evaluation. To illustrate this point, I draw from a follow-up interview with Aditi, a PRADAN affiliate. Multiple times throughout the interview she mentions things that point to the ways IFPRI (and the WEAI) did not align with PRADAN and PHRS frames (and the FAAM intervention). She mentions that not all dimensions of the WEAI were addressed in FAAM, specifically mentioning women’s asset ownership. She then describes their differing ideas about how women’s workload would be alleviated. Namely, she says how the WEAI turned this problem into something that is solved through mechanization; adding that there was not enough attention to the way women’s workload could be shared by other people in the family:34 “Our perspectives I think were different … Because even without introducing a machine you can see how drudgery or women's workload can be more shared. So that's why I'm saying we were not looking at it” (Aditi, line 133 in follow-up interview). PRADAN’s concern about women’s workload being relieved by sharing it with household members (including their husbands) might seem like a minor difference between evaluators and implementers, but it leads to larger implications on how opposed the WEAI evaluation may be in PRADAN’s understanding of FAAM’s empowerment potential. This is because the WEAI’s final assessment of women’s empowerment is measured against men’s empowerment. In the case of the WEAI, when women’s empowerment does not lead to men’s disempowerment, it is understood as successful empowerment (Malapit et al. 2019). It is a 34 Aditi describes how PRADAN was concerned about the potential negative outcomes of the FAAM intervention including the increased expectations women would have to do cash crops, maintain a nutrition garden, forage leafy green vegetables, and cook multiple things in a meal, “without any of the burden being shared … there's not enough focus on sharing of the increased workload by other people in the family. I don't think we considered it.” (Aditi line 45 in follow-up interview). 110 finding that IFPRI affiliates readily shared in their interviews as showing how SHGs are successful at empowering women, for example: “Within the household, we find that it's not that being a member in a group disempowers your husband or the primary male within the same household. It is positive for the women, but it's not negative for the men. This is something that we find” (Priya, line 44). Despite these implications, these conceptual and analytical differences are something all parties seemed to be aware of from the start, as Aditi mentions: “The schools of thought are quite different. IFPRI and PHRN *starts laughing* and we, had quite a lot of conceptual and approach related discussions. That issue was from the beginning kind of flagged … IFPRI would always flag this, that we've not really sharply defined, and urge to do that” (Aditi, line 63 in follow-up interview). Implementers explain that having the women beneficiaries take ownership of the program and the actionable items is crucial to their empowerment – while also admitting that development work is messy and that strictly defining things from the onset is not only difficult, but not something they can do without the women being involved. This brings us to my last finding, which simultaneously lends itself to explaining the factors (other than the diagnostic frame) that informed evaluators’ prognostic frame. The pro-WEAI was created as an improvement to its older sister, the WEAI. One of the improvements is its methodological addition of using qualitative analysis in understanding empowerment in specific project-level contexts (Malapit et al. 2019). Despite this improvement and IFPRI’s task of piloting the pro-WEAI, the qualitative analysis of the pro-WEAI did not take place in the WINGS project. Like implementers, evaluators also acknowledge the importance of qualitatively understanding what empowerment means to women. One interviewee states, “I 111 think it’s really important to understand what is it that the women think is empowerment? And we did not go into that in this study” (Priya, line 47). When I asked IFPRI affiliates why this did not occur, they said it was because of lack of funding and PRADAN’s direct funding of the project (from the Bill and Melinda Gates Foundation (BMGF)) after baseline data was collected (using the WEAI). It was explained to me by an IFPRI affiliate that the initial grant of the project was to do an impact evaluation of the pathways NI BCC in SHGs had, but ultimately grew from BMGF’s interest in promoting women’s groups as a means to empowerment and nutrition. “[PRADAN] was going to be rolling out this nutrition intensification model. They wanted to know whether it worked, what pathways were kind of triggered. If it didn't work, what were the barriers and challenges. The initial grant was just a grant to kind of conduct the impact evaluation of that, then it evolved into a much larger sort of learning agenda. I think this grew out of the Bill and Melinda Gates Foundation's interest in promoting women's groups as sort of a means to women's empowerment and also effecting change along other dimensions. In this case, it was health and nutrition” (Swapna, line 7). Thus, solutions of SHGs as taking the form of empowerment on the ground not only came from the diagnostic frame (conceptualizing women’s empowerment as a means and identifying the problem as malnutrition), but (lack of) resources and sources of funding also informed the prognostic frame of this case study. Conclusion The findings in this study suggest that when empowerment is primarily framed as a means, it is implemented as a medium for empowerment processes to occur. In the case of the FAAM/WINGS project, empowerment was framed as a means to improve the end goals of 112 improving nutrition, and was implemented as a medium in the primary form of SHGs as a source or embodiment of women’s collectivization – or “power with” others. Further, I find that when the end goal is operationalized in a way that separates it from empowerment, empowerment is solely framed as a means to an end. This example is found in the case of WINGS, where empowerment was operationalized as agency (using the WEAI methodology) and nutrition operationalized separately as BMI and dietary diversity; leading to accounts that ultimately landed on empowerment being a pathway – and not something intended to improve from a “nutrition intervention.” On the other hand, I also find that when the end goal is operationalized in a way that includes empowerment, empowerment has the potential to act as both a means to an end and an end goal. Implementers’ diagnostic framing demonstrated the way empowerment was equally important and synonymous with the goals related to nutrition. This conceptualization of empowerment was then operationalized as achievements, rather than agency, to embody the two – specifically, as a process for overcoming gender norms that cause women’s malnutrition. Thus, implementers saw the additional identification of story-telling modules (NI BCC) as also being a solution – or way empowerment was embodied – in practice. These differences between implementers and evaluators in identifying what constitutes empowerment in practice and therefore what indicates empowerment in analysis, led to the non- alignment in measuring or evaluating if empowerment “happened” as a result of the FAAM intervention. Using the WEAI to evaluate women’s empowerment as agency led evaluators to conclude that the NI BCC alone did not empower women, but SHGs did. Further, the WEAI results also concluded that SHGs empowered women without disempowering men. While evaluators found this result as a “positive” effect of the program, it may not align with the kind 113 of results implementers wanted to see. Namely, in terms of men sharing the burden of work with women in their household. At the same time, the WEAI may also not align with implementers because the tool operationalized and measured empowerment as agency and PRADAN operationalized empowerment in terms of achievements. Lastly, I argue that the WEAI may not be the best tool in evaluating empowerment in this case because empowerment indicators should align with the intended conceptualization. In this case, empowerment is ultimately conceptualized as a means – specifically as a collective process – and intended to be implemented in the form of SHGs. However, the WEAI operationalizes empowerment in terms of agency, with only two of these representing forms of collective agency (Kumar et al. 2021). Indicators that algin with the original conceptualization – like forms of “power with” – that women gain from collective mobilization in their SHGs would align better in evaluating the empowerment potential of FAAM. All in all, I find that evaluators and implementers were aligned in their diagnostic framing and prognostic framing of the development project. While IFPRI affiliates tried to address their differences early on, timing and lack of funding for qualitative understanding of empowerment in the FAAM project led to non-alignment between evaluators and implementers in operationalizing and identifying empowerment in practice. Notwithstanding, I hope the findings from this research help future development projects acknowledge the potential issues that can arise when these framing processes do not align between parties. 114 REFERENCES Acharya, Gayatri, Alreena Renita Pinto, and Deepti Kakkar. 2020. “Social Inclusion: Implementation, Outcomes and Key Learnings.” South Asia Agriculture and Rural Growth Discussion Note Series March 2020 (1). Retrieved June 2023: https://documents1.worldbank.org/curated/en/339661597075144917/pdf/National-Rural- Livelihoods-Project-Discussion-Notes-Compilation.pdf Alkire, Sabina, Ruth Meinzen-Dick, Amber Peterman, Agnes Quisumbing, Greg Seymour, and Ana Vaz. 2013. The Women’s Empowerment in Agriculture Index.” World Development 52:71–91. Benford Robert D. and David A. Snow. 2000. “Framing Processes and Social Movements: An Overview and Assessment” Annual Review Sociology 26:611-39. Bhatt, Ela. 1989. "Toward Empowerment." World Development 17(7): 1059-1065. Brody, Carinne, Thomas de Hoop, Martina Vojtkova, Ruby Warnock, Megan Dunbar, Padmini Murthy, and Shari L. Dworkin. 2017. “Can Self-help Group Programs Improve Women’s Empowerment? A Systematic Review.” Journal of Development Effectiveness 9(1):15– 40. Cornwall, Andrea, and Ann Whitehead. 2007. Feminisms in Development: Contradictions, Contestations and Challenges. Zed Books. Dworkin, Shari L. 2012. "Sample Size Policy for Qualitative Studies using In-depth Interviews." Archives of Sexual Behavior 41(6):1319-1320. FAO, IFAD, UNICEF, WFP and WHO. 2020. The State of Food Security and Nutrition in the World 2020: Transforming Food Systems for Affordable Healthy Diets. Rome, FAO. Retrieved January 10, 2021 (https://doi.org/10.4060/ca9692en) Gamson, William A. 1992. Talking Politics. New York: Cambridge University Press. Guérin, Isabeele, and Santosh Kumar. 2017. “Microcredit Self-help Groups and Dalit Women: Overcoming or Essentializing Caste Difference?” Pp. 158-186 in Dalit Women: Vanguard of an Alternative Politics in India, edited by Shanmugasundaram Anandhi and Karin Kapadia. Taylor & Francis Guest, Greg, Arwen Bunce, and Laura Johnson. 2006. "How Many Interviews are Enough? An Experiment with Data Saturation and Variability." Field Methods 18(1):59-82. IFPRI. 2022. “A Decade of the Women’s Empowerment in Agriculture Index (WEAI) [Video]. YouTube. Streamed February 16, 2022. https://www.youtube.com/watch?v=O5- 0MRUMNdM 115 IIPS (International Institute for Population Sciences) and ICF. 2021. National Family Health Survey (NFHS-5), 2019-21: India. Mumbai: IIPS. Jose, Sunny. 2011. "Adult Undernutrition in India: Is there a Huge Gender Gap?" Economic and Political Weekly: 95-102. Kuckartz, Udo. 2014. Qualitative Text Analysis: A Guide to Methods, Practice and Using Aoftware. Sage. Kumar, Neha, Kalyani Raghunathan, Alejandra Arrieta, Amir Jilani, and Shinjini Pandey. 2021. "The Power of the Collective Empowers Women: Evidence From Self-help Groups in India." World Development 146:105579. Malapit, Hazel, Agnes Quisumbing, Ruth Meinzen-Dick, Greg Seymour, Elena M. Martinez, Jessica Heckert, Deborah Rubin. 2019. "Development of the project-level Women’s Empowerment in Agriculture Index (pro-WEAI)." World Development 122:675-692. Meinzen-Dick, Ruth, and Agnes Quisumbing. 2018. “Response to Garncia and Wanner “Gender Inequality and Food Security: Lessons for the Gender-responsive work of the International Food Policy Research Institute and the Bill and Melinda Gates Foundation” Food Security (2017) 9:1091–1103." Food Security 10(2):247-251. Meinzen-Dick, Ruth Suseela; Rubin, Deborah; Elias, Marlène; Mulema, Annet Abenakyo, and Myers, Emily. 2019. Women’s Empowerment in Agriculture: Lessons from Qualitative Research. IFPRI Discussion Paper 1797. Washington, DC: International Food Policy Research Institute (IFPRI). Retrieved October 20, 2020 (http://ebrary.ifpri.org/cdm/ref/collection/p15738coll2/id/133060) Ministry of Rural Development. 2018. Mainstreaming Social Inclusion in DAY – NRLM - Volume 1 – Manual for Practitioners. Retrieved from: https://aajeevika.gov.in/sites/default/files/nrlp_repository/Vol%201%20- %20Manual%20for%20Practitioners.pdf __________. 2023. “SHGs Profile Entry Status” Ministry of Rural Development, Government of India. Retrieved July 4, 2023: https://nrlm.gov.in/shgReport.do?methodName=showIntensiveStateWiseReport Myers, Emily. 2022. “Lessons From a Decade of the Women’s Empowerment in Agriculture Index (WEAI).” IFPRI Blog: Event Post, International Food Policy Research Institute. March 3, 2022. https://www.ifpri.org/blog/lessons-decade-womens-empowerment- agriculture-index-weai Nichols, Carly E. 2021a. "Self-help Groups as Platforms for Development: The Role of Social Capital." World Development 146:105575. 116 __________ . 2021b. “Spaces for Women: Rethinking Behavior Change Communication in the Context of Women’s Groups and Nutrition-Sensitive Agriculture.” Social Science & Medicine 285:114282. Public Health Resource Society. 2017. PHRN Newsletter, 1(4):1-14. http://phrsindia.org/wp- content/uploads/2017/10/NewsletterDecember2017.pdf Prasad, Vandana. 2016. “Cyclical Negotiations Between Theory and Practice For Building Knowledge In Nutrition, With Intent To Action A Case Study of Collaboration.” Proceedings from the 2nd Symposium on Transformation for Rural Development: Collaboration and Co-Production of Knowledge. April 2016. https://www.researchgate.net/publication/301636489_Cyclical_Negotiations_Between_T heory_And_Practice_For_Building_Knowledge_In_Nutrition_With_Intent_To_Action_ A_Case_Study_of_Collaboration Rowlands, Jo. 1997. Questioning Empowerment: Working with Women in Honduras. Oxfam. Sanyal, Paromita. 2014. Credit to Capabilities: A Sociological Study of Microcredit Groups in India. New York, NY: Cambridge University Press. Scott, Samuel, Shivani Gupta, Purnima Menon, Kalyani Raghunathan, Giang Thai, Agnes Quisumbing, Vandana Prasad, Aditi Hegde, Avijit Choudhury, Madhu Khetan, Carly Nichols, and Neha Kumar. 2022. "A Quasi-Experimental Evaluation of a Nutrition Behavior Change Intervention Delivered Through Women's Self-Help Groups in Rural India: Impacts on Maternal and Young Child Diets, Anthropometry, and Intermediate Outcomes." Current Developments in Nutrition 6(6):nzac079. Snow, David. 2004. "Framing Processes, Ideology, and Discursive Fields." The Blackwell Companion to Social Movements 1:380-412. Snow, David A. and Robert D. Benford. 1988. “Ideology, Frame Resonance, and Participant Mobilization.” International Social Movement Research 1:197-217. __________ 1992. “Master Frames and Cycles of Protest.” Pp. 133-155 in Frontiers in Social Movement Theory, edited by Aldon D. Morris and Carol McClurg Mueller. New Haven, CT: Tale University Press. Snow, David, Robert Benford, Holly McCammon, Lyndi Hewitt, and Scott Fitzgerald. 2014. "The Emergence, Development, and Future of the Framing Perspective: 25+ Years Since Frame Alignment." Mobilization: An International Quarterly 19(1):23-46. Snow, David A., E. Burke Rochford Jr., Steven K. Worden, and Robert D. Benford. 1986. “Frame Alignment Processes, Micromobilization, and Movement Participation.” American Sociological Review 51(4):464-81. 117 Syed, Jawad and Faiza Ali. 2011. “The White Woman’s Burden: From Colonial Civilisation to Third World Development.” Third World Quarterly 32(2):349-365. Wright, Wynne and Taylor Reid. 2010. “Green Dreams or Pipe Dreams?: Media Framing of the U.S. Biofuels Movement.” Biomass and Bioenergy 35(4):1-10. 118 CHAPTER 4: EVALUATING EMPOWERMENT FOR WOMEN IN JHARKHAND, INDIA: A FEMINIST STRUCTURAL EQUATION MODELING APPROACH Introduction In chapter two, I presented the multiple ways empowerment is framed in the gender and development discourse. One of the major findings from chapter two’s study is that women’s empowerment is understood as either an intrinsic goal (an end in itself) or an instrumental goal (a means to an end). Most quantitative work in the development discourse frames women’s empowerment as an intrinsic goal, in which women’s agency is measured as the outcome of women acquiring resources (e.g., education, paid work, or saving and lending groups) (Mahata et al. 2017; Yount et al. 2016). At the same time, research in the Indian context that frames women’s empowerment as an instrumental goal, centers around a popular assumption that I also discovered in chapter three’s case study. Namely, that women’s empowerment leads to improved child health and nutrition (Kumar et al. 2018; Shroff et al. 2011). As a result, women’s individual-level nutritional status is rarely analyzed in relation to their own empowerment. This is concerning, not only because women in India are more likely to be undernourished compared to men, but the proportion of malnourished women in India is one of the highest in the world (Gupta et al. 2019; IIPS and ICF 2021; Jose and Navaneetham 2008). Moreover, the eastern state of Jharkhand has the highest proportion of underweight women (26 percent) in India (IIPS and ICF 2021). The disproportionately higher rate of undernutrition that characterizes women’s poor health can also be seen in India’s growing anemic population in recent years (IIPS and ICF 2021; WHO 2008). In Jharkhand, the proportion of anemic women is alarmingly high: 65 percent of the female population is anemic – more than double the 119 population of anemic men in Jharkhand (30 percent), higher than the national proportion of anemic women (57 percent), and well above the global threshold categorizing it as a severe public health issue (WHO 2008).35 Empowerment research to date has led to the investment of governmental and non- governmental programs in India aiming to empower women and address child malnutrition by increasing women’s access to resources and decision-making (e.g., Development of Women and Children in Rural Areas, the Indira Mahila Yojana, Swayamsidha). More recently, Indian scholars have pointed out that these programs are increasingly being delivered in disempowering ways that turn women into passive instruments to deliver development and economic ends, rather than keeping a feminist commitment towards improving women’s daily lives (Batliwala and Dhanraj 2007). Moreover, feminists have long suggested that analyses focused on the productive aspects of women’s lives tend to support the development of income-generating activities without addressing the overriding influence of patriarchy in women’s daily (Raju 2005; Rathgeber 1990; Rowlands 1997). In terms of women’s empowerment, daily practices within the household are sites where women often experience the greatest difficulty, because it is where women experience disempowerment in very negative and immediate ways (Rowlands 1997; Wieringa 1994). An important example of this in India relates to food and nutrition, in which women are disempowered from birth to adulthood. Common gender-inequitable taboos and practices among Indian households result in girl babies being breastfed for shorter periods compared to boys, girl children given less and worse food than boy children, and cultural norms that stipulate women eat their meals after men (Hathi et al. 2021). These daily practices have been cited as having observable negative consequences for women’s nutrition, especially because 35 According to the World Health Organization (2008), if the prevalence of anemia in a community is more than 40 percent, it is considered a severe public health issue. 120 eating after men and boys often means that women eat leftover food of lesser quality and quantity compared to other members in their household (Chandran and Kirby 2021; Hathi et al. 2021). Understanding and identifying the ways women experience disempowerment – especially in their household - is important to properly diagnosing the sources that directly and indirectly limit women’s empowerment, including their nutritional status. Accordingly, one of the objectives in the present study is to quantitatively estimate the direct and indirect effects of the sources of women’s (dis)empowerment in rural Jharkhand. The ultimate goal is to create a feminist model of women’s empowerment that can be used as an example for future quantitative analyses evaluating and measuring women’s empowerment in the Global South. Abiding by the feminist commitment to improve women’s lives, I adopt a feminist methodology to quantitatively model women’s empowerment in this study. Most notably, this is done by using feminist theory derived primarily by Indian scholars to conceptualize women’s empowerment and contextualize it to rural Jharkhand conditions. By doing so, I frame empowerment as both a process of overcoming disempowering structures and sources, and the transformative result of that process. I also select quantitative methods that are analytically capable of capturing a feminist understanding of empowerment as a multidimensional and dynamic concept; namely, by utilizing confirmatory factor analysis (CFA) and structural equation modeling (SEM). Lastly, to shift the focus off women’s productive and instrumental role towards an intrinsic framing aimed at improving women’s lives, I change the focus of influential factors from resources toward socio-structural causes of women’s disempowerment 121 identified by Indian feminists including caste and patriarchy,36 and maintain women’s individual- level nutrition as the very last outcome of women’s empowerment. The current chapter proceeds as follows. I first describe the feminist approach and prior research that informs this study, in which I develop a structural model (path diagram) of the processes and outcomes of women’s empowerment. I then discuss the data and sample of the current study, followed by a description of the measures created and utilized in the models. Next, I describe the statistical methods, including CFA and SEM, followed by results and discussion of the analyses. The chapter concludes with a discussion of limitations, contributions, and policy recommendations. Framework Conceptualizing Women’s Empowerment Most quantitative analyses on women’s empowerment in the development discourse acknowledge empowerment as a process, which has led to the popular operationalization of empowerment as agency, most often measured using indicators of women’s household decision- making (Gupta and Yesudian 2006; Malapit et al. 2019; Miedema et al. 2018). Measuring empowerment in this way falls in line with theory that suggests power is demonstrated when a person, or set of people, prevails in decision making (Eade and Williams 1995). However, without the application of SEM (or path, quantitative analyses using these measurements end up treating agency as either a process (means to an end) or an end in itself (outcome of empowerment processes). Moreover, using this indicator alone assumes women’s role in household decision-making is empowering, without questioning if the measurement hides 36 I define patriarchy as a system of power-inequality that justifies male superiority and female subordination, consisting of gendered patterns of privilege and oppression. Thus, the subject matter of analysis considers the various forms patriarchy takes in both institutions and customs, including intimate partner violence (IPV) against women, men’s patriarchal values, and women’s internalized oppression. 122 women’s denial of choice, or covert forms of power that reinforce women’s disempowerment (Eade and Williams 1995; Kabeer 1999). In this study, I borrow from several Indian feminists’ interpretations of women’s empowerment to understand it as a multidimensional and dynamic term, and define it as both a process, and the outcome of that process, by which women gain transformative control over their lives (Batliwala 1994; Jejeebhoy 2000; Kabeer 1999; Raju 2005). The inclusion of the term transformative is critical to assessing if women’s lives reflect strategic empowerment outcomes, as opposed to outcomes reflecting women’s denial of choice stemming from women’s subordinate status (Batliwala 1994; Kabeer 1999). In other words, for women to obtain transformative (or strategic) empowerment outcomes, empowerment processes must address structures and sources of women’s disempowerment that are relevant to power. I further define women’s disempowerment as a condition of power inequality that manifests in the subordination and mistreatment of women in overt material ways (e.g., violence) and covert psychological ways (e.g., men and women’s gender values) (Batliwala 1994; Kabeer 1999; Rowlands 1997). In the next section, I describe the empirical model I create that considers women’s empowerment in this way - as a complex system of direct and indirect relations of women’s (dis)empowerment processes and outcomes. Modeling Women’s Empowerment in India Key issues of quantitatively measuring empowerment lie in its multidimensionality and are further complicated by issues of endogeneity (Alsop et al. 2006; Malhotra and Schuler 2005). To address these issues, I employ structural equation modeling (SEM) – a quantitative tool which is analytically capable of accounting for complex relationships between multiple endogenous (outcome) variables and addresses empowerment’s multidimensionality by utilizing multiple 123 latent variables in the model (e.g., women’s nutritional status - from multiple observed variables) (Liu et al. 2011). Using SEM, I construct a hypothesized conceptual path diagram modeling women’s empowerment in rural Jharkhand. To develop this model, I draw from feminist theory and past studies of women’s empowerment, intimate partner violence (IPV), and women’s nutrition in India. My model does not replicate any previous work, but rather contributes to the literature by creating a feminist-informed quantitative model of women’s empowerment as both a process and outcome of that process that leads to women gaining transformative control over their lives. Figure 3.1 shows my hypothesized model using drawing conventions of SEM (Schumacker and Lomax 2004). My model includes latent variables: constructs best represented by multiple conceptually interrelated observed indicators, which I describe in the methods section of this chapter. In the path diagram, latent variables are represented as ovals and observed variables are depicted as boxes. Lines directed from a predictor variable to an endogenous variable (i.e., outcome variable) denote a direct effect, or the influence of one variable to another variable. Exogenous variables (i.e., variables that are solely predictor variables) are presented on the left side of the model and are assumed to covary: this covariance is represented by the four- pronged-arrows to the far-left side of the model connecting household size, women’s caste, men’s gender values, and women’s age (Kline 2015). 124 Figure 3.1. Conceptual structural model of women’s empowerment in rural Jharkhand Kabeer (1999) posits that if there are systematic gender differences in universally valued functioning achievements, like proper nourishment and good health, they can be taken as evidence of underlying capabilities (i.e., disempowerment related to power) rather than differences in preferences. Applicably, women’s disproportionately higher rate of undernourishment in India can be taken as an indicator of women’s disempowerment – because it reflects an overt manifestation of power inequality. In the context of rural Jharkhand, I model women’s nutritional status sociologically, and informed by Indian feminist theory, as a manifestation of intersecting forms of power inequality including forms of patriarchy (i.e., men’s gender values, violence against women, women’s internalized oppression), caste, and age. I also control for household size to account for the potential impact women’s nutritional status may suffer from as a result of the cultural practice of eating after (many) household members. Given these sources of power inequality acting against women, women’s nutritional status is 125 conceptualized and treated as a strategic empowerment outcome in the model; where healthy nutritional status indicates women have gained transformative control over their life. As I noted earlier, women’s agency is often used either as an indicator of empowerment processes (means to an end) or as an outcome of empowerment processes (an end in itself). In my hypothesized model, agency is conceptualized as both. The benefit of using SEM is that it addresses issues of endogeneity by allowing variables to act as both predictors and outcomes of (dis)empowerment processes. For instance, past studies have found that women’s level of agency (measured as household decision-making) in India depends on several covert and overt sources of power inequality, including women’s age, caste, her experience of violence, and her and her husband/partner’s traditional gender values (Allendorf 2012; Bhattacharya and Banerjee 2012; Gupta and Yesudian 2006; Mondal and Paul 2021; Sahoo and Raju 2007). At the same time, a long line of research in development argues that women’s agency leads to improved nutrition (Kumar et al. 2018; Shroff et al. 2011). Many times, this argument is applied to studies and programs that aim to improve children’s nutrition, subsequently framing women’s empowerment as instrumental. In my model, I test this assumption by creating a hypothesized relationship between women’s agency and women’s nutrition. Rather than focusing on how women’s agency affects their children’s nutrition, I test the direct relationship between women’s agency and her own nutrition to determine if these frames really do treat women as instruments to developmental ends (if agency does not predict women’s nutrition). Moving to the center of Figure 3.1, women’s experience of intimate partner violence (IPV) and women’s internalized oppression are also treated as both predictors and outcomes of women’s (dis)empowerment processes. In a patriarchal context like India, one of the most recognized overt forms of women’s disempowerment is the prevalence of IPV against women 126 (Ackerson and Subramanian 2008; Mondal and Paul 2021). Nearly a third of ever-married women in India have experienced IPV; defined as psychological, physical, or sexual harm or the threat of such harm by a current or former partner or spouse (IIPS and ICF 2021; Yount et al. 2014). The prevalence of IPV in Indian homes proves to remain remarkably high, reflecting deeply embedded patriarchal ideologies that oppress women and threaten their health and well- being (Mondal and Paul 2021; Sahoo and Raju 2007). Moreover, feminists have pointed out instances where men’s overt use of “power over” – in the form of violence – may no longer be necessary, because such patriarchal control has become internalized by women. For example, a woman who is subjected to IPV when she expresses her own opinions, may start to withhold them and adopt patriarchal values as a result (Eade and Williams 1995; Rowlands 1997). In a context that socially conditions individuals to understand men’s superiority as inherent, an analysis of internalized oppression is not only important to feminists, but is imperative in evaluations of women’s (dis)empowerment in India. Accordingly, modeling women’s internalized oppression as a process and outcome of women’s (dis)empowerment can provide evidence on the active role women may have in maintaining and defending ideologies that perpetuate her own disempowerment (including her lack of agency and/or poor nutritional status). Based on this feminist theory (suggesting women subject to abuse from men may begin to adopt men’s patriarchal ones), I hypothesize a relationship from IPV to women’s level of internalized oppression (Eade and Williams 1995; Rowlands 1997). I do not estimate the reverse that assumes women’s internalized oppression predicts IPV, that is sometimes modeled by quantitative scholars (Mondal and Paul 2021), to avoid possible victim-blaming narratives that explain predictors of women’s experience of IPV. Rather, women’s experience of IPV is 127 modeled as a result of their husband/partner’s patriarchal gender values and predicted by women’s ascribed statuses of age and caste. Empowerment scholars are increasingly recognizing the important role men’s gender values have in influencing outcomes of women’s empowerment (Batliwala 1994; Sanyal 2014). For instance, past studies in India have found husbands that maintain more traditional (patriarchal) gender values tend to have wives that experience more IPV and have little autonomy, compared to wives married to men with more egalitarian values (Mondal and Paul 2021; Sanyal 2014). Despite the increasing acknowledgement amongst development scholars outlining the importance of involving men to empower women, they do not necessarily construct models that reflect the complexity of its direct and indirect effects on (dis)empowerment processes and outcomes. As a result, the indirect and direct effects of men’s gender values on both women’s agency and nutrition in India have yet to be modeled. Even more, there is very limited research that focuses on the interconnected linkages between men and women’s gender values, IPV, women’s agency, and women’s individual-level nutrition under one model. By using SEM, I can close the gap in the literature by modeling and estimating men’s values as having a direct and/or indirect effect on women’s agency or nutritional status through certain processes of women’s disempowerment (i.e., IPV, internalized oppression). In addition to the overarching influence of patriarchy that characterizes women’s lives, they are subjected to additional systems of power inequality that can either enable their empowerment or further disempower them. Feminist and development scholars have suggested that in the Indian context, women’s age and caste are two important sources of power inequality that directly influence women’s (dis)empowerment processes and outcomes. In the Indian family context, women’s age is an important ascribed status that determines her position in the 128 household hierarchy (Allenddorf 2012). Young women are likely to be disempowered living in joint families with their in-laws after marriage; subordinate not only to all men in the household, but also to more senior women - especially their mother-in-law (Kandiyoti 1988). Culturally, young married women are expected to not participate in household matters, face close and constant scrutiny in their movement and behavior, and are more vulnerable to IPV (Gupta and Yesudian 2006; Sahoo and Raju 2007). As married women grow older within the joint family system, transitioning from daughter-in-law to mother-in-law, they accrue social advantages and gain power (Allenddorf 2012; Gupta and Yesudian 2006). While women’s age theoretically creates harmonious enabling effects for empowerment processes and outcomes for women, their caste location in society presents inconsonant effects on women’s (dis)empowerment. The proverbial caste-class nexus posits that higher caste locations coincide with relatively better levels of wealth and nutritional status (Sahoo and Raju 2007). For instance, quantitative studies found that women in scheduled tribes had the highest rates of anemia and undernutrition, followed by women in scheduled castes and other backward classes; with women of higher castes exhibiting the smallest rates (Jose and Navaneetham 2008). At the same time, higher caste status can also translate to more disempowering processes and outcomes for women. For example, feminists using qualitative methods have documented the ways upper-caste-class values of status, honor, and respect breed patriarchal norms and behavior in Indian households, including women “keeping within their limits” (e.g., not going out), and when taken to the extreme, “provides the rationale and legitimation for violence against women” (Still 2017:208). Thus, the caste system may lead to upper caste women exhibiting healthier nutritional status, while simultaneously working against them in terms of violence, internalized oppression, and agency. 129 A major goal of this study is to improve an understanding of women’s empowerment processes that are important toward addressing the most immediate and negative sources of women’s disempowerment in India. This section just described these processes and sources, as well as their hypothesized effects. Taken together, I model the effects of women’s caste and age, her experience of IPV, and the gender values held by women and their husband/partner - on women’s agency and nutrition. I now turn to the dataset and sample I use to test this model followed by a description of the methods used to estimate the effects. Methods Data and Sample This study uses data from India’s National Family Health Survey 2019-21 (NFHS-5). The NFHS-5 includes national representative data from 29 states in India and the observable variables needed for the present study. There are seven separate datasets that make up the NFHS- 5. To obtain the relevant variables for the analyses of this study, I merge and use data from two of these seven datasets including the Persons/Household Members Recode (PR) and the Couples Recode (CR).37 The CR dataset contains data for married or cohabitating women and men who both declared their status to each other and completed individual interviews found in the Individual/Women’s Recode (IR) and the Men’s Recode (MR) datasets. The sample in the present study includes married women and their current husband/partner in rural Jharkhand. The sample for this study includes only married women because of the importance of testing the influence of men’s gender values and women’s experience of IPV on women’s agency and nutritional status. The decision to include the 37 Specifically, I conducted a one-to-one merge of the CR and PR data across the same women respondents. Household size as well as individual-level data on women’s nutrition came from the PR dataset; all other data (including men’s survey items) came from the CR dataset. 130 influence of men’s values in the model - over the decision to include single women in the model - was also chosen because of feminists and development practitioners’ insistence on understanding the role gender norms and family dynamics have in promoting more gender equitable outcomes (Allendorf 2012; Mondal and Paul 2021; Quisumbing et al. 2022). Furthermore, the sample only includes households in rural Jharkhand because of the great diversity in women’s lives, not only across India, but between rural and urban spaces. For instance, women in rural areas not only experience stricter caste norms, but also gender norms, where patriarchy tends to be more pronounced; such as the experience of physical violence that is more commonly reported among women in rural areas compared to women in urban areas (Anandhi and Kapadia 2017; IIPS and ICF 2021; Sahoo and Raju 2007). Studies also find that higher incidences of women’s poor nutrition outcomes are more prevalent in rural areas compared to urban areas (Jose and Navaneetham 2008). Lastly, the decision to focus on the state of Jharkhand was for theoretical and methodological reasons. First, national-level quantitative analysis can mask different contextual patterns that sub-national level analysis can provide. When it comes to nutrition, national level analysis is appropriate for identifying issues of supply or availability; but for identifying women’s nutritional challenges related to accessibility and utilization, sub-national level analyses are more appropriate (Denny et al. 2018). Jharkhand, in particular, is an important case study because it has the highest proportion of underweight women in the country, an above national rate of women experiencing IPV (34 percent), and demonstrates one of the largest inequalities in basic well-being outcomes between men and women (IIPS and ICF 2021).38 38 Rural Jharkhand was also one of the targeted areas in the WINGS/FAAM project. 131 Measures The goal of quantitative modeling is to balance model parsimony with data quality to create a model that explains women’s empowerment with as few predictor variables as possible (Agresti 2018; Denny et al. 2018). To accomplish this goal, I created latent variables to narrow the number of variables in the structural model. Latent variables are constructed as hypothesized measurement models that are empirically tested using confirmatory factor analysis (CFA). I use the processes and outcomes of empowerment identified in the framework section to guide the creation of five latent variables (measurement models), including women’s nutritional status, women’s agency, IPV, women’s internalized oppression, and men’s gender values. NFHS-5 data relevant to these five concepts were used to measure their hypothesized constructs. In addition to these five latent variables, I incorporate three observed predictor variables, including women’s age, caste, and household size. The indicators presented below were selected and measured based on previous studies using NFHS data and literature on women’s empowerment either specific to the Indian context or previously tested CFA models. Women’s Nutritional Status Nutrition is an abstract concept that embodies an essential role in human well-being, where an individual’s status of malnourishment indicates a person’s lack of freedom to lead a minimally healthy life (Jose and Navaneetham 2008). Given nutrition’s vital importance to human well-being and the implication malnutrition has for human development, scholars across disciplines in sociology to medicine study and measure this abstract concept using observable indicators (Bentley and Griffiths 2003; Shroff et al. 2011). Popular among observed variables capturing adult nutritional status include measures of body mass index (BMI), anemia levels, and dietary diversity (Bharati et al. 2019; Bhattacharya et al. 2013; Hazarika et al. 2012; Scott et al. 132 2022). Guided by these well-known measures, I created a latent endogenous variable, “women’s nutritional status,” characterized by observed data from women’s individual-level NFHS-5 results related to her BMI, anemia level, and dietary diversity. I now turn to the specifications of these three observed variables that measure women’s nutritional status. Scholars using NFHS data have noted that there could be differential factors impacting those who are underweight versus those who are overweight and obese (Bharati et al. 2019). To better understand the influential factors of women’s malnutrition caused by not getting enough nutrients, I use the World Health Organization’s (2023) BMI classifications to create an ordinal variable of women’s BMI that dropped overweight and obese observations (BMI >25), and coded women of healthy or normal weight (BMI between 18.5-24.9) as 2, underweight (BMI between 17-18.49) as 1, and moderately to severely thin women (BMI < 17) as 0. Currently pregnant women and women who gave birth in the last two months were also dropped in creating the observed variable of women’s BMI, following NFHS guidelines (Croft et al. 2018). The second observed variable used to capture the latent variable women’s nutritional status is women’s anemia. Women’s anemia levels were determined by using the concentration of hemoglobin in capillary blood. Blood specimens were taken from eligible women aged 15-49 using a finger prick and collected in a microcuvette for analysis using the HemoCue system (Croft et al. 2018). An ordinal variable was created to categorize women’s anemia levels by their hemoglobin count so that lower values reflected poor nutritional status: moderate to severe anemia (below 10.9g/dl) were coded as 0, mild anemia (11.0–11.9 g/dl) was coded as 1, and no anemia (above 12 g/dl) was coded as 2. The observed variable women’s dietary diversity derived from seven survey questions asking women about their frequency eating 1) milk or curd, 2) pulses or beans, 3) dark green 133 leafy vegetables, 4) fruits, 5) eggs, 6) fish, and 7) chicken or meat. Available responses were coded as dichotomous to reflect if a woman ate the food in question daily (coded as 1) versus women that ate it weekly, occasionally, or never (coded as 0). Further, the most recent NFHS-5 report found that 29 percent of women in India are vegetarians, defined as those who do not eat chicken, meat, fish, or eggs daily (IIPS and ICF 2021). Using this definition to account for vegetarianism among women, I used the seven binary responses to classify each woman into one of four groups to reflect their daily dietary diversity. Accordingly, the ordinal dietary diversity variable used in the nutrition measurement model reflected the following categories: women who eat zero out of the seven possible food types per day were coded as 0, one food type per day as 1, two food types per day as 2, and 3 - 7 food types per day were coded as 3. Women’s Agency The second latent endogenous variable modeled is women’s agency. To test the popular assumption that women’s agency leads to improved nutrition, I follow previous quantitative studies suggestions on measuring agency using indicators of women’s decision-making power (Gupta and Yesudian 2006; Hazarika et al. 2012; Richardson et al. 2019). Five items from the women’s NFHS-5 survey captured this recommendation. Specifically, women were asked about who has the final say in 1) making large household purchases, 2) respondents (own) health care, 3) decisions about visiting family or relatives, 4) what to do with money their husband earns, and lastly, 5) who mainly decides how the money earned by the respondent (wife/self) is used. Possible survey responses included wife alone (respondent/self), wife and husband, husband/partner alone, someone else, and other. Each of the five items were coded as an ordinal variable where: those that reported self alone were coded as 2 to reflect women with full 134 autonomy, joint decision making with their husband/partner were coded as 1, and those that reported husband alone, someone else, or other were coded as 0 to reflect no agency. Intimate Partner Violence Using a holistic definition of intimate partner violence (IPV), that includes modes of physical, sexual, and psychological (or emotional) abuse, I create a third latent variable using five observed variables (Ackerson and Subramanian 2008; Mondal and Paul 2021; Yount et al. 2014). A portion of the women surveyed in the NFHS-5 were given an additional set of questions related to violence, including IPV.39 The five items selected for this latent variable came from the portion of the “domestic violence module” asking currently married women about their experience of violence perpetrated by their current husband/partner. Specifically, women were asked if they have ever experienced the following forms of abuse from their husband/partner: 1) any severe physical violence (including being kicked, dragged, strangled, burnt, or threatened with a gun/weapon), 2) any “less” severe physical violence (including being pushed, shook, had something thrown at them, slapped, punched with fist, or hit by something harmful), 3) any sexual violence, 4) were humiliated (by their husband/partner), and 5) were threatened with harm (by their husband/partner). Possible responses to the two physical abuse items and one sexual abuse item were never (coded as 0) and yes (coded as 1). Possible responses to the two emotional abuse items were recoded in the same dichotomous way to reflect if women had ever experienced psychological abuse by their husband: where never was coded as 0, and often, sometimes, and yes but not in the last 12 months were coded as 1. 39 One woman per household (among eligible women in the household) was selected to participate in the domestic violence module. The module is also an optional portion of the NFHS-5 survey, in which those selected may not complete it due to lack of privacy or other reasons (Croft et al. 2018). 135 Women’s Internalized Oppression Five observed variables measured the latent variable women’s internalized oppression. The survey questions asked whether women justified wife beating under various conditions, that captured gendered beliefs about the subordination and mistreatment of women (i.e., women’s internalized oppression). Women were asked whether they believed a husband was justified in beating his wife if she 1) goes out without telling him, 2) neglects the children, 3) argues with him, 4) refuses to have sex with him, and 5) burns the food. Response options were yes, no, or “don’t know.” Those that responded yes were coded as 1, reflecting women’s internalized oppression about female subjugation. No, “don’t know,” and missing responses were coded as 0 to reflect women’s lack of internalized oppression, following DHS guidelines that missing values on attitudes towards wife beating are assumed to not be agreeing with the scenario in question (Croft et al. 2018). Men’s Gender Values The fifth latent variable, men’s gender values, is the only latent variable that is an exogenous variable in the model. The same five questions that were used to capture the latent variable women’s internalized oppression were used to capture men’s gender values – but with men’s survey responses. Specifically, I used responses from men who were linked to a woman (married, partner) in the NFHS-5 couples recode dataset with completed interviews in the men’s survey. Like women, men were asked whether they believed a husband was justified in beating his wife if she 1) goes out without telling him, 2) neglects the children, 3) argues with him, 4) refuses to have sex with him, and 5) burns the food. Yes, no, and “don’t know” responses were recoded in the same dichotomous way, with yes responses coded as 1 to reflect men’s patriarchal 136 gender values, and no, “don’t know,” and missing responses coded as 0 to reflect men possessing values which do not align with the ideals about women’s subjugation (Croft et al. 2018). Observed Predictor Variables In addition to the five latent variables in the structural model, I also include three observed exogenous variables: women’s age, caste, and household size. Women’s age was measured by asking women their date of birth and was recorded as a continuous variable in years completed. Household size is also a continuous variable, reflecting the number of all de jure household members (i.e., the number of usual household residents, whether or not they stayed in the house the night before the survey was conducted). This measure was used to reflect a definition of a household as a person or group of persons that normally reside together and share a cooking arrangement. Information on women’s caste was obtained by asking women if they belong to a scheduled caste, a scheduled tribe, other backward classes (OBCs), or none of them. Scheduled caste, scheduled tribe, and OBCs are official designations of the Government of India used to identify groups who have been historically (and continue to be) marginalized by their perceived low status in order to provide them reservations (i.e., affirmative action in education and government employment). Only those marginalized by the Hindu caste system (outside of the four major castes – Brahmin, Kshatriya, Vaishya, and Shudra) are deemed a scheduled caste according to The Constitution (Scheduled Castes) Order 1950; they are also known by their self- referential term dalits (Anandhi and Kapadia 2017; Das Gupta 2019). The socio-economic position of rural OBCs and dalits are similar, but OBCs are primarily made up of the lowest caste within the caste system (specific castes are classified as socially and economically backward) and do not face the same social isolation and violence based on caste that dalits experience 137 (Anandhi and Kapadia 2017; Das Gupta 2019). In addition, compared to dalits, those in scheduled tribes face geographical isolation, and continue to have the lowest levels of education and make up the majority of the lowest level of relative wealth in Indian society (Gopinath 2018; IIPS and ICF 2021). I therefore code the survey item on women’s caste as an ordinal variable to reflect a scale of relative oppression and privilege on the basis of caste: with women identifying with a scheduled tribe as the most oppressed group and coded as 0, those in a scheduled caste coded as 1, those in other backward classes as 2, and those in none of these as 3 (to reflect their relative privilege in society). Those that responded “don’t know” to the survey question were coded as missing. Statistical Analyses The data was first checked for outliers to identify abnormal BMI and anemic values based on a categorized box plot comparing the distribution of BMI scores over ordinal anemia levels (Agresti 2018). Two observations were removed from the analysis because their data may have affected the results (Bharati et al. 2019; Stevens 1996).40 Recoding of all the variables in the model, as well as the merging of the NFHS-5 datasets, were done using STATA/SE 17.0. Descriptive statistics for all observed variables of the final sample are presented in Table 3.1 with their corresponding measures that were described in the previous section. 40 The two observations removed from analysis happened to also be the lowest recorded BMI scores in the data (e.g., 12.55 and 13.27), with corresponding levels of anemia that indicated these two women had “no anemia.” 138 Table 3.1. Descriptive statistics of observed indicators used in models41 Item Measurement Mean SD Exogenous variables Household size Number of de jure household members 5.35 2.08 Women's age Years since birth 32.64 8.52 Women's caste 0 = scheduled tribe 0.38 1 = scheduled caste 0.16 2 = other backward classes 0.40 3 = none of these/upper caste 0.06 Men’s gender values (latent) Justifies wife beat 1: goes out w/o telling husband 0 = no, 1 = yes 0.09 0.29 Justifies wife beat 2: neglects the children 0 = no, 1 = yes 0.09 0.29 Justifies wife beat 3: argues w husband 0 = no, 1 = yes 0.12 0.33 Justifies wife beat 4: refuses sex w husband 0 = no, 1 = yes 0.07 0.26 Justifies wife beat 5: burns the food 0 = no, 1 = yes 0.06 0.24 Endogenous variables Women’s internalized oppression (latent) Justifies wife beat 1: goes out w/o telling husband 0 = no, 1 = yes 0.12 0.32 Justifies wife beat 2: neglects the children 0 = no, 1 = yes 0.11 0.31 Justifies wife beat 3: argues w husband 0 = no, 1 = yes 0.12 0.33 Justifies wife beat 4: refuses sex w husband 0 = no, 1 = yes 0.06 0.24 Justifies wife beat 5: burns the food 0 = no, 1 = yes 0.09 0.29 Women’s experience of IPV (latent) Emotional abuse 1 (humiliated) 0 = never, 1 = ever experienced 0.07 0.26 Emotional abuse 2 (threatened with harm) 0 = never, 1 = ever experienced 0.07 0.25 Sexual abuse 0 = never, 1 = ever experienced 0.07 0.25 Severe physical abuse (e.g., strangled, burnt) 0 = never, 1 = often, sometimes, ever 0.32 0.47 Less severe physical abuse (e.g., pushed, slapped) 0 = never, 1 = often, sometimes, ever 0.09 0.28 Women’s agency (latent) Makes large household purchases 0= husband alone, someone else, other 0.17 1 = wife and husband (jointly) 0.79 2 = wife (respondent) alone 0.04 Responsible for own health care 0= husband alone, someone else, other 0.17 1 = wife and husband (jointly) 0.81 2 = wife (respondent) alone 0.03 Deciding to visit family or relatives 0= husband alone, someone else, other 0.17 1 = wife and husband (jointly) 0.79 2 = wife (respondent) alone 0.04 How to use husband’s money 0= husband alone, someone else, other 0.17 1 = wife and husband (jointly) 0.79 2 = wife (respondent) alone 0.04 How to use own money 0= husband alone, someone else, other 0.13 1 = wife and husband (jointly) 0.78 2 = wife (respondent) alone 0.09 41 Categorical variables are interpreted as percentages; ordinal variables reflect each level's proportional count to the variable, and binary variables are provided with standard deviations. Sample sizes are slightly smaller for caste (n = 1567), violence variables (n = 1320), BMI (n = 1292), anemia (n = 1522), decisions about husband's money (n = 1559) and decisions about own money (n = 338), compared to full sample (N = 1597). 139 Table 3.1. (cont’d) Women’s nutritional status (latent) Woman’s BMI (kg/m2) 0 = moderate-severely thin ( < 17 kg/m2) 0.10 1 = underweight (17-18.49 kg/m2) 0.18 2 = healthy/normal (18.5-24.9 kg/m2) 0.71 Woman’s anemia status via hemoglobin 0 = moderate-severe anemia ( < 10.9g/dl) 0.37 ...concentration in blood (g/dl) 1 = mild anemia (11.0–11.9 g/dl) 0.27 2 = no anemia ( > 12 g/dl) 0.36 Woman’s dietary diversity (eating seven types of 0 = zero food types/day 0.18 ...foods per day: milk/curd, pulses/beans, dark green 1 = one food type/day 0.26 ...leafy vegetables, fruits, eggs, fish, and 2 = two food types/day 0.40 ...chicken/meat) 3 = three to seven food types/day 0.16 Structural equation modeling (SEM) was used to test the hypothesized empowerment model that specifies the relationship among women’s caste, age, household size, men’s gender values, women’s internalized oppression, IPV, women’s agency, and women’s nutrition status. This analysis involved a two-step process. The first stage evaluated the validity of the latent constructs (measurement models), and the second stage tested the path analysis (structural model). Confirmatory factor analysis (CFA) was used in the first stage to estimate/validate the relationships between the latent constructs and their indicator variables (measurement model). CFA tests how well the hypothesized measurement model fits the data (Kline 2015). Each of the five latent variables were estimated separately to assess the validity of each measurement model. That is, I completed CFA analyses for each latent construct on its own to check the reliability of each measurement model. Because all five of the latent variables are constructed with categorical indicators (e.g., ordinal and binary observed variables), mean- and variance-adjusted weighted least squares (WLSMV) (i.e., robust WLS) was used to estimate each measurement model in Mplus 8. WLSMV is the appropriate estimation method when indicators of a latent variable are ordinal or dichotomous, as it makes no distributional assumptions and adjusts for the nonlinear association between the latent response variable and observed indicators (Kline 2015). 140 Best fitting measurement models were identified using four goodness-of-fit (GOF) indices appropriate when using WLSMV estimation: the root mean square error of approximation (RMSEA), the Bentler Comparative Fit Index (CFI), the Tucker Lewis Index (TLI), and the standardized root mean square residual (SRMR) (Bentler 1990; Kim et al. 2021; Liu et al. 2011; Tucker and Lewis 1973). Scholars have suggested standard benchmarks for these indices to serve as indictors of good model fit: CFI and TLI values of 0.90 or above are considered to have acceptable fit and values close to or above 0.95 indicating good fit; RMSEA less than 0.05, and SRMR less than 0.08 (Bentler 1990; Hu and Bentler 1999; Kim et al. 2021). Standardized factor loadings, an assessments of construct validity, were also analyzed to ensure all factor loadings were significantly different from zero (i.e., p < 0.05) and had the expected sign (i.e., positive loadings), along with positive residual variances for each observed categorical variable (Kline 2015; Schumacker and Lomax 2004). Once the measurement models were validated, the second stage of estimation (structural model) proceeded. Structural equation modeling analysis was performed to test the hypothesized relationships among all five latent constructs and three observed exogenous variables and obtain the GOF indices with the same assessment criteria as in CFA. The structural model was also estimated using WLSMV in Mplus 8. Results Measurement Models Numerous measurement models were tested, and some were re-specified or reduced to a single indicator due to poor measurement model fit. Specifically, measurement models that resulted in both poor GOF indices and poor local fit statistics - including observed variables that were either not significantly different from zero (p > 0.05), had factor loadings below 0.2, above 141 one,42 or were negative, and/or had negative residual variances43 – were re-specified (e.g., added measurement error covariances between similarly asked survey questions, re-defined observed variable categories, removed the variable(s) from the measurement model) or were reduced to a single indicator (Kline 2015; Richardson et al. 2018; Schumacker and Lomax 2004). Results of the final measurement models making up the five latent variables in the full model are presented in Table 3.2. Appendix A provides details on two cases that resulted in the reduction of observed variables. Table 3.2. Measurement models fit indices, observed indicators, and factor loadings Construct N df RMSEA CFI TFI SRMR Observed indicator Standardized factor loading Men’s 1597 5 0.019 0.999 0.999 0.019 Man justify wife beating 1 0.912*** gender Man justify wife beating 2 0.935*** values Man justify wife beating 3 0.882*** Man justify wife beating 4 0.779*** Man justify wife beating 5 0.854*** Women’s 1597 5 0.037 0.998 0.997 0.018 Woman justify wife beating 1 0.837*** internalized Woman justify wife beating 2 0.942*** oppression Woman justify wife beating 3 0.946*** Woman justify wife beating 4 0.894*** Woman justify wife beating 5 0.878*** Women’s 1320 3 0.032 0.998 0.995 0.018 Emotional violence 1 0.839*** experience Emotional violence 2 0.846*** of IPV Sexual violence 0.879*** Severe physical violence 0.776*** Less severe physical violence 0.866*** Women’s 1597 3 0.042 0.999 0.996 0.018 Makes large purchases 0.874*** agency Responsible for own healthcare 0.863*** Deciding to visit family 0.697*** How to use husband’s money 0.727*** How to use own money 0.451*** Women’s 1597 0 0.000 1.000 1.000 0.000 Woman’s BMI 0.544** nutritional Woman’s anemia status 0.227** status Woman’s dietary diversity 0.262** Note: *p < 0.05; **p < 0.01; ***p < 0.001. 42 In cases where measurement models had poor global fit statistics and factor loadings above one, the variables were removed. In cases like these, high factor loadings around one can signal a high degree of multicollinearity in the data leading to poor model fit (Jöreskog 1999). 43 Negative residual variances indicates the model is inadmissible; specifically the observed variable in question cannot be used in the model (Kline 2015). 142 Structural Model Presented in this section are the results of the structural model (i.e., path analysis). The analysis provides structure for representing and estimating assumed causal effects between the exogenous variables and endogenous variables in the model (Agresti 2018). The hypothesized structural model tested is identical to the one presented in Figure 3.1. The results of the structural model confirm a good fit: df=274, RMSEA=0.014 (90% confidence for RMSEA from 0.010 to 0.018), CFI=0.994, TLI=0.993, SRMR=0.053. The results of the structural model are illustrated in Figure 3.2: showing standardized WSLMV coefficients of significant direct path coefficients (bold black lines) and insignificant direct path coefficients (dashed grey lines). The six assumed covariances between the four exogenous variables (e.g., household size, women’s caste, men’s gender values, and women’s age) were estimated in this model (Kline 2015); this estimation is depicted in Figure 3.2 as the non-bolded four-pronged-arrows on the far-left side of the model. Table 3.3 in Appendix B provides details of standardized WSLMV coefficients of the direct and selected total indirect effects estimated in the structural model. 143 Figure 3.2. Structural model of women’s empowerment in rural Jharkhand showing standardized WSLMV coefficients of direct paths (N = 1597) Notes: *p < 0.05; **p < 0.01; ***p < 0.001. RMSEA = 0.014, CFI = 0.994, TLI = 0.993, SRMR = 0.053. Women’s nutritional status is found to be directly influenced by two exogenous variables and one endogenous variable. First, women’s caste and age have significant direct links to their nutritional status, with elevated caste status and older age associated with higher nutritional status. In addition, the latent endogenous measure of IPV also significantly predict women’s nutritional status: showing that women’s experience of IPV substantially reduce women’s nutrition. These findings are consistent with previous studies on the predictive effects higher caste, age, and IPV have on women’s nutrition outcomes in India (Ackerson and Subramanian 2008; Hazarika et al. 2012; Jose and Navaneetham 2008). Women’s agency does not have a significant effect on women’s nutritional status, providing evidence against the long-standing assumption that women’s agency leads to healthier nutrition (Kumar et al. 2018; Shroff et al. 2011). Men’s gender values also do not have a direct 144 effect on either women’s nutritional status or agency, but men’s patriarchal gender values have a significant negative total indirect effect on reducing both women’s nutritional status and her agency through IPV pathways. Results from the structural model also show that women’s agency is directly reduced by IPV and directly increased by women’s age and internalized oppression. The effects of age and IPV are consistent with feminist theory and previous findings in India, suggesting that husbands use violence to control their wives’ behavior, and that women become more involved in decision- making with respect to their husbands as they age (Allendorf 2012; Gupta and Yesudian 2006; Sahoo and Raju 2007). At the same time, previous studies found the opposite (negative) relationship between women’s internalized oppression and women’s agency than the one found in this model (Gupta and Yesudian 2006). Consistent with feminist theory, women’s internalized oppression is most significantly predicted by the direct effect of women’s experience of violence from her husband/partner (Eade and Williams 1995; Rowlands 1997; Sahoo and Raju 2007). In addition, men’s gender values have a significant direct and indirect effect on women’s internalized oppression. Namely, men’s patriarchal gender values predict heightened internalized oppression for women directly and through IPV. Lastly, women’s experience of IPV is most significantly predict by men’s gender values, followed by women’s age; consistent with other studies in India (Mondal and Paul 2021). As expected, women’s elevated caste status is associated with heightened levels of IPV, yet the association is not significant by traditional threshold standards (p = 0.056) (Sahoo and Raju 2007). Nonetheless, the standardized coefficients of these IPV findings suggest that, more than 145 women’s age or caste, men’s patriarchal gender values have the greatest influence predicting married women’s experience of IPV in rural Jharkhand. Discussion Using SEM to model women’s empowerment, I tested the popular instrumentalist frame of women’s empowerment that assumes women’s agency leads to improved nutrition, by turning the end goal into women’s nutritional status – rather than at the household level or that of her children’s. Results of the structural model indicate a non-significant positive effect from women’s agency to women’s nutritional status. I interpret this finding to suggest that women’s agency – as a measurement of women’s decision-making – is not a valid means (process) towards improving women’s nutritional status in rural Jharkhand. Thus, programs or studies that frame women’s decision making as a way to achieve development goals, are likely to use women as instruments in development schemes that disempower them (Batliwala and Dhanraj 2007). What deserves more attention when it comes to addressing the most immediate causes of women’s nutrition and agency are those related to caste, age, and patriarchy. First, the findings suggest that caste is the greatest predictor of women’s nutritional status in rural Jharkhand, followed by women’s experience of IPV, and her age. This is consistent with previous findings and theory suggesting that women in India experience intersecting forms of power-inequality that manifest in differential material outcomes on the basis of caste, age, and gender (Anandhi and Kapadia 2017; Gupta and Yesudian 2006; Kabeer 1999). Moreover, IPV is shown to be central to women’s (dis)empowerment. Findings from the structural model reveal that women’s experience of IPV is directly associated with higher levels of internalized oppression and lower levels of agency and nutritional status; the latter being consistent with similar quantitative studies, and the former contributing to the literature (Mondal 146 and Paul 2021). Relatedly, the indirect and direct effects of patriarchy - stemming from men’s patriarchal values - on women’s empowerment processes and outcomes suggest that women with husbands that hold patriarchal values predict heightened levels of IPV against women, which directly and indirectly predict women’s elevated internalized oppression, lack of agency, and poor nutritional status. From previous quantitative studies, it was hypothesized that all covert forms of patriarchy, including women’s internalized oppression, would predict lower levels of women’s agency (Gupta and Yesudian 2006). However, results from the structural model suggest that married women with internalized oppression have increased levels of agency. I offer two explanations for this finding informed by different lines of feminist theory. The first suggests that women may be adopting patriarchal gender values to gain more power in their household – a phenomenon known as a patriarchal bargain (Kandiyoti 1988). In highly patriarchal and abusive households, women may actively adopt patriarchal values as a practical tactic to avoid conflict and preserve energy (Raju 2005). Some argue that it may even be the only option for women in a context of extreme violence, suggesting that women adopt patriarchal values as a “survival mechanism” (Eade and Williams 1995; Samuel 2001). The findings indicate that this may be the case in rural Jharkhand. Nearly 32 percent of married women in the sample reported ever experiencing a form of severe physical violence by their husband, and results from the structural model suggest that men use violence to control women’s behavior. On the other hand, feminists including Kabeer (1999), warn to not simply accept women’s behavior (agency) as “real choice” (i.e., strategic empowerment), because it may instead reflect the denial of choice that stems from and reinforces women’s subordinate status. Thus, this line of feminist theory posits that women may not be actively choosing tactics 147 (patriarchal bargains) based on their positionality as oppressed (Mohanty and Miraglia 2012), but rather are unaware of the way their internalized oppression creates perceived limitations on what they can and cannot do (Kabeer 1999; Rowlands 1997). Nonetheless, either feminist explanation proposed here suggests that the measurement of women’s agency in the model may not represent women’s strategic empowerment outcomes. Instead, it could be more representative of women’s burdened agency, a patriarchal bargain, or simply a form of power that fails to generate transformative outcomes for women. More importantly, these findings suggest that women’s agency (decision making power) does not equate to women’s empowerment in rural Jharkhand; and future studies should be cautious of making this assumption. Furthermore, understanding agency in this light may also help in explaining the non-significant relationship between women’s agency and women’s nutritional status in the model. Conclusion In addition to discovering the sources that directly and indirectly limit women’s empowerment in rural Jharkhand, I offer some concluding observations that may prove useful for future evaluation, measurement, and modeling efforts around women’s empowerment. I begin with some of the limitations and suggestions for future research, then turn towards the contributions of the present study, and end with policy recommendations. The real world is, of course, not as simple as quantitative models portray. Modifications to the structural model presented in this study are possible, including additional variables and paths not included here that are relevant to the context of rural Jharkhand. For instance, future research should consider predictor variables such as women and men’s exposure to new ideas and consider including endogenous variables capturing women’s freedom of mobility (Sanyal 2014). These are just two suggestions, but we know that empowerment is a multidimensional 148 concept that can be operationalized and measured in countless ways. However, capturing every domain and potential influence of women’s empowerment was not the intended goal of this study. In this study, I attempted to create a model of empowerment that shifted the focus off instrumentalist frames of women’s empowerment towards an intrinsic feminist model of women’s empowerment. To do this, I modeled women’s empowerment as a process of women gaining transformative control over her life. By using SEM to account for the complex relationships and endogeneity empowerment processes entail, I treated multiple domains of empowerment as both predictors and outcomes of empowerment processes in the model. I also addressed structures and sources of women’s disempowerment relevant to power, including caste, age, and three forms of patriarchy that overwhelmingly characterize married women’s daily lives in India (i.e., husband’s patriarchal gender values, IPV, and women’s internalized oppression). In doing so, the structural model revealed men’s gender values and IPV as having a central role in shaping women’s (dis)empowerment processes in rural Jharkhand. The results of the model suggest development programs focused on improving women’s strategic empowerment, including her nutritional status, should focus more attention towards mitigating IPV by addressing men’s patriarchal gender values. Without addressing these forms of patriarchy, women’s nutritional status will remain lower to men’s, and the country will continue to have high rates of malnourished women. Furthermore, despite the Indian constitution outlawing untouchability and caste-based discrimination over 70 years ago, findings from the structural model suggest that women’s nutritional status in rural Jharkhand is still largely predicted by the direct effect of women’s caste identity. 149 The Indian government has historically proven to care about addressing its undernourished population and empowering rural women with its investment in programs that aim to remedy these issues. The present study suggests that the focus of these programs should be modified. Namely, in ways that begin to address caste and the prevailing forms of patriarchy that actively disempower women in their daily lives. First, Indian feminists have previously suggested that a legal remedy for addressing a structural wrong in society (like outlawing untouchability) is not a viable solution towards addressing caste-based violence and discrimination (Anandhi and Kapadia 2017). Rather, the need today is to abolish the material base of the system by outlawing caste-based identities (Anandhi and Kapadia 2017). Only then can we start to see the direct disempowering effects of caste on women’s nutritional status begin to dissipate. Second, programs aimed at empowering women need to acknowledge the central role that men’s gender values and interpersonal relationships at home, including their use of violence on women, have in affecting women’s internalized oppression, her level of agency, and ultimately her nutritional status. Thus, non-governmental and governmental programs should not interpret the need to involve men as a step away from empowering women, but as a reflection of their feminist commitment in ensuring women’s strategic needs are being addressed. Lastly, I hope that the findings related to women’s agency and internalized oppression shed light on the importance of including an analysis of internalized oppression in studies of women’s empowerment outcomes. The present study suggests that while women’s agency, measured as decision-making, has become the most widely-used indicator of women’s empowerment, it may not reflect feminist definitions of women’s empowerment as leading to transformative or strategic ends. Rather, it may more closely reflect women’s lack of choice 150 stemming from and reinforcing her disempowered status. By addressing women’s internalized oppression and men’s patriarchal values in programs aimed to empower women, the possibility of fostering forms of agency that lead to women gaining transformative control over their lives will become possible. 151 REFERENCES Ackerson, Leland K., and SV2789268 Subramanian. 2008. "Domestic Violence and Chronic Malnutrition among Women and Children in India." American Journal of Epidemiology 167(10):1188-1196. Agresti, Alan. 2018. Statistical Methods for the Social Sciences. Pearson Education. Alkire, Sabina, Ruth Meinzen-Dick, Amber Peterman, Agnes Quisumbing, Greg Seymour, and Ana Vaz. 2013. The Women’s Empowerment in Agriculture Index.” World Development 52:71–91. Allendorf, Keera. 2012. "Women’s Agency and the Quality of Family Relationships in India." Population Research and Policy Review 31:187-206. Alsop, Ruth, Mette Bertelsen, and Jeremy Holland. 2006. Empowerment in Practice: From Analysis to Implementation. Washington, DC: The World Bank Publications. Alston, Margaret, and Badi Akhter. 2016. "Gender and Food Security in Bangladesh: The Impact of Climate Change." Gender, Place & Culture 23(10):1450-1464. Anandhi, Shanmugasundaram, and Karin Kapadia, eds. 2017. Dalit Women: Vanguard of an Alternative Politics in India. Taylor & Francis. Batliwala, Srilatha. 1994. "The Meaning of Women's Empowerment: New Concepts form Action.” Pp. 127-138 in Population Policies Reconsidered: Health, Empowerment, and Rights, edited by G. Sen, A. Germain, and L. C. Chen. Batliwala, Srilatha, and Deepa Dhanraj. 2007. "Gender Myths that Instrumentalize Women: A View from the Indian Front Line." Pp. 21-34 in Feminisms in Development: Contradictions, Contestations and Challenges, edited by Andrea Cornwall and Ann Whitehead. Zedbooks. Bentler, Peter M. 1990. Comparative Fit Indexes in Structural Models. Psychological Bulletin 107(2):238 Bentley, Margaret E., and Paula L. Griffiths. 2003. "The Burden of Anemia among Women in India." European Journal of Clinical Nutrition 57(1):52-60. Bharati, Susmita, Manoranjan Pal, Soumendu Sen, and Premananda Bharati. 2019. "Malnutrition and Anaemia Among Adult Women in India." Journal of Biosocial Science 51(5):658- 668. Bhattacharya, Joysankar, and Sarmila Banerjee. 2012. "Women Empowerment as Multidimensional Capability Enhancement: An Application of Structural Equation Modeling." Poverty & Public Policy 4(3):79-98. 152 Bhattacharya, Joysankar, Sarmila Banerjee, and Montu Bose. 2013. “On Assessment of Women Empowerment at Individual Level: An Analytical Exposition.” Pp. 385-400 in Development and Sustainability: India in A Global Perspective, edited by Joysankar Bhattacharya and Anjan Chakrabarti. Springer. Hathi, Payal, Diane Coffey, Amit Thorat, and Nazar Khalid. 2021. "When Women Eat Last: Discrimination at Home and Women’s Mental Health." PLoS One 16(3):e0247065. Chandran, Vidya, and Russell S. Kirby. 2021. "An Analysis of Maternal, Social and Household Factors Associated with Childhood Anemia." International Journal of Environmental Research and Public Health 18(6):3105. Cornwall, Andrea, and Ann Whitehead. 2007. Feminisms in Development: Contradictions, Contestations and Challenges. Zed Books. Croft, Trevor N., Aileen MJ Marshall, Courtney K. Allen, Fred Arnold, Shireen Assaf, and Sarah Balian. 2018. Guide to DHS Statistics. Rockville, Maryland, USA: ICF. Das Gupta, Sejuti. 2019. Class, Politics, and Agrarian Policies in Post-Liberalisation India. Cambridge University Press. Denny, Riva CH, Sandra T. Marquart-Pyatt, Arika Ligmann-Zielinska, Laura Schmitt Olabisi, Louie Rivers, Jing Du, and Lenis Saweda O. Liverpool-Tasie. 2018. "Food Security in Africa: A Cross-Scale, Empirical Investigation Using Structural Equation Modeling." Environment Systems and Decisions 38:6-22. Eade, Deborah and Suzanne Williams. 1995. The Oxfam Handbook of Relief and Development. Vol 1, Approaches, Focusing on People, Capacity Building. Oxford: Oxfam. Gopinath, Vishnu. 2018. “Who Are the Scheduled Castes, Scheduled Tribes, OBCs and EBCs?” The Quint. Retrieved June 2022: https://www.thequint.com/explainers/scheduled-caste- scheduled-tribe-obc-ebc-sc-st-prevention-of-atrocities-act-explainer#read-more Gupta, Kamla, and P. Princy Yesudian. 2006. "Evidence of Women’s Empowerment in India: A Study of Socio-Spatial Disparities." GeoJournal 65:365-380. Gupta, Soumya, Prabhu Pingali, and Per Pinstrup-Andersen. 2019. "Women’s Empowerment and Nutrition Status: The Case of Iron Deficiency in India." Food Policy 88:101763. Hazarika, Jiten, Irani Saikia, and Partha Jyoti Hazarika. 2012. "Risk Factors of Undernutrition Among Women in the Reproductive Age Group of India: Evidence from NFHS-3." Am Eur J Sci Res 7(1):05-11. Hu, L. T., and Peter M. Bentler. 1999. "Cutoff Criteria for Fit Indexes in Covariance Structure Analysis: Conventional Criteria Versus New Alternatives." Structural Equation Modeling: A Multidisciplinary Journal 6(1):1-55. 153 IIPS (International Institute for Population Sciences) and ICF. 2021. National Family Health Survey (NFHS-5), 2019-21: India. Mumbai: IIPS. Jejeebhoy, Shireen J. 2000. “Women’s Autonomy in Rural India: Its Dimensions, Determinants, and the Influence of Context.” Pp. 204-238 in Women’s Empowerment and Demographic Processes: Moving Beyond Cairo, edited by H. B. Presser and G. Sen. New York: Oxford University Press. Jöreskog, Karl. 1999. “How Large Can a Standardized Coefficient Be.” Retrieved May 15, 2023: https://web.archive.org/web/20180430044906id_/http://www.ssicentral.com/lisrel/techdo cs/HowLargeCanaStandardizedCoefficientbe.pdf Jose, Sunny, and Kannan Navaneetham. 2008. "A Factsheet on Women's Malnutrition in India." Economic and Political Weekly:61-67. Kabeer, Naila. 1999. “Resources, Agency, Achievements: Reflections on the Measurement of Women’s Empowerment.” Development and Change 30(3):435–464. Kim, Minjung, Christa Winkler, and Susan Talley. 2021. "Binary Item CFA of Behavior Problem Index (BPI) Using Mplus: A Step-By-Step Tutorial." The Quantitative Methods for Psychology 17:141-153. Kline, Rex B. 2015. Principles and Practice of Structural Equation Modeling. Guilford Publications. Kumar, Neha, Samuel Scott, Purnima Menon, Samyuktha Kannan, Kenda Cunningham, Parul Tyagi, Gargi Wable, Kalyani Raghunathan, and Agnes Quisumbing. 2018. "Pathways from women's group-based Programs to Nutrition Change in South Asia: A Conceptual Framework and Literature Review." Global Food Security 17:172-185. Liu, Hui, Qiu Wang, Venessa Keesler, and Barbara Schneider. 2011. "Non-Standard Work Schedules, Work–Family Conflict and Parental Well-Being: A Comparison of Married and Cohabiting Unions." Social Science Research 40(2):473-484. Mahata, Dinabandhu, Amit Kumar, and Ambarish Kumar Rai. 2017. "Female Work Force Participation and Women Empowerment in Haryana." International Journal of Humanities and Social Sciences 11(4):1039. Malapit, Hazel, Agnes Quisumbing, Ruth Meinzen-Dick, Greg Seymour, Elena M. Martinez, Jessica Heckert, Deborah Rubin, Ana Vaz, Kathryn M. Yount, and Gender Agriculture Assets Project Phase 2 (GAAP2) Study Team. 2019. "Development of the Project-Level Women’s Empowerment in Agriculture Index (pro-WEAI)." World Development 122:675-692. Malhotra, Anju, and Sidney Ruth Schuler. 2005. “Women’s Empowerment as a Variable in International Development.” Pp. 71–88 in Measuring Empowerment: Cross-Disciplinary 154 Perspectives, edited by D. Narayan. Washington, DC: World Bank, Gender and Development Group. Mason, Karen Oppenheim, and Herbert Smith. 2003. Women’s Empowerment and Social Context: Results from Five Asian Countries. Washington, DC: Gender and Development Group, World Bank. Miedema, Stephanie Spaid, Regine Haardörfer, Amy Webb Girard, and Kathryn M. Yount. 2018. "Women’s Empowerment in East Africa: Development of a Cross-Country Comparable Measure." World Development 110:453-464. Mohanty, Chandra Talpade, and Sarah Miraglia. 2012. "Gendering Justice, Building Alternative Futures." Pp. 115-148 in Alternatives to Privatization, edited by D. McDonald and G. Ruiters. New York, NY: Routledge. Mondal, Dinabandhu, and Pintu Paul. 2021. “Associations of Power Relations, Wife-Beating Attitudes, and Controlling Behavior of Husband with Domestic Violence Against Women in India: Insights from the National Family Health Survey–4” Violence Against Women 27(14): 2530-2551 Quisumbing, Agnes R.; Meinzen-Dick, and Malapit, Hazel J. 2022. Women’s Empowerment and Gender Equality in South Asian Agriculture: Measuring Progress Using the Project-Level Women’s Empowerment in Agriculture Index (pro-WEAI) in Bangladesh and India. World Development 151:105396. https://doi.org/10.1016/j.worlddev.2021.105396 Richardson, Robin, Norbert Schmitz, Sam Harper, and Arijit Nandi. 2019. "Development of a Tool to Measure Women’s Agency in India." Journal of Human Development and Capabilities 20(1):26-53. Raju, Saraswati. 2005. “Limited Options-Rethinking Women’s Empowerment ‘Projects’ in Development Discourses: A Case from Rural India.” Gender, Technology and Development 9(2):253-271. Rowlands, Jo. 1997. Questioning Empowerment. Oxford: Oxfam. Sahoo, Harihar, and Saraswati Raju. 2007. "Domestic Violence in India: Evidences and Implications for Working Women." Social Change 37(4):131-152. Sanyal, Paromita. 2014. Credit to Capabilities: A Sociological Study of Microcredit Groups in India. Cambridge University Press. Schumacker, Randall E., and Richard G. Lomax. 2004. A Beginner's Guide to Structural Equation Modeling. Psychology Press. Scott, Samuel, Shivani Gupta, Purnima Menon, Kalyani Raghunathan, Giang Thai, Agnes Quisumbing, Vandana Prasad, Aditi Hegde, Avijit Choudhury, Madhu Khetan, Carly 155 Nichols, and Neha Kumar. 2022. "A Quasi-Experimental Evaluation of a Nutrition Behavior Change Intervention Delivered Through Women's Self-Help Groups in Rural India: Impacts on Maternal and Young Child Diets, Anthropometry, and Intermediate Outcomes." Current Developments in Nutrition 6(6):nzac079. Shroff, Monal R., Paula L. Griffiths, Chirayath Suchindran, Balakrishna Nagalla, Shahnaz Vazir, and Margaret E. Bentley. 2011. "Does Maternal Autonomy Influence Feeding Practices and Infant Growth in Rural India?." Social Science & Medicine 73(3):447-455. Stevens, James P. 1996. Applied Multivariate Statistics for the Social Sciences. Lawrence Erlbaum Associates: Mahwah. Still, Clarinda. 2017. “Dalit Women, Rape and the Revitalisation of Patriarchy?.” Pp. 189-217 in Dalit Women: Vanguard of an Alternative Politics in India, edited by Shanmugasundaram Anandhi and Karin Kapadia. Taylor & Francis. Tucker, Ledyard R., and Charles Lewis. 1973. "A Reliability Coefficient for Maximum Likelihood Factor Analysis." Psychometrika 38(1):1-10. WHO (World Health Organization). 2008. Worldwide Prevalence of Anaemia 1993-2005: WHO Global Database on Anaemia, edited by Bruno de Benoist, Erin McLean, Ines Egli and Mary Cogswell. World Health Organization. https://apps.who.int/iris/handle/10665/43894 __________. 2023. “Global Nutrition Monitoring Framework” Nutrition Landscape Information Systen (NLiS) Retrieved April 6, 2022: https://apps.who.int/nutrition/landscape/help.aspx?menu=0&helpid=392&lang=EN#:~:te xt=BMI%20%3C17.0%3A%20moderate%20and%20severe,BMI%20%E2%89%A525.0 %3A%20overweight Yount, Kathryn M., Kristin VanderEnde, Sarah Zureick-Brown, Hoang Tu Anh, Sidney Ruth Schuler, and Tran Hung Minh. 2014. "Measuring Attitudes About Intimate Partner Violence Against Women: The ATT-IPV Scale." Demography 51(4):1551-1572. Yount, Kathryn M., Kristin E. VanderEnde, Sylvie Dodell, and Yuk Fai Cheong. 2016. "Measurement of Women’s Agency in Egypt: A National Validation Study." Social Indicators Research 128:1171-1192. 156 APPENDIX A: REDUCED MEASUREMENT MODELS Caste I attempted to create a latent variable of women’s socio-economic status (SES) to capture the importance of class in India today (Das Gupta 2019). The original variable comprised of women’s caste, her level of education, relative household wealth, and a binary variable named “paid cash” representing if the woman in question earned cash for her employment in the last 12 months (as opposed to solely being paid in-kind or not at all for her formal employment). However, when I estimated this latent variable, the measurement model reported the latent variable covariance matrix (psi) was not positive definite: indicating a factor loading for the observed ordinal variable “wealth” was above one (1.067) and had a negative residual variance (- 0.139). This finding indicates that the observed variable wealth showed a high degree of multicollinearity in the data (suggesting that another observed variable in the measurement model measures the same thing) leading to poor model fit (Jöreskog 1999). The model also showed that the binary variable “paid cash” was well below the 0.2 cutoff with a factor loading of 0.075 and was not significantly different from zero (p=0.216). These findings indicate that the variable “paid cash” does not measure the latent variable SES well and should be removed from the measurement model. Thus, in removing both wealth and paid cash from the SES measurement model, I was only left with two indicators. However, two-indicator measurement models are considered “not identified” unless they are correlated with another latent variable in the model (Schumacker and Lomax 2004). Correlating SES (measured by caste and education) with a related exogenous latent variable was not applicable in this structural equation model, so I reduced the latent variable SES to a single indicator of caste given its importance in modeling women’s (dis)empowerment in India and its likelihood in causing multicollinearity issues with the variable household wealth. Agency Originally, I operationalized the latent variable capturing women’s agency to include not just women’s decision making but also three items regarding women’s freedom of movement. However, RMSEA and SRMR values indicated poor global fit statistics (i.e., poor latent model), and observed variables representing women’s mobility to health facilities and mobility to the market had factor loading approaching one - indicating multicollinearity between these two variable and others in the model. Once I dropped these two variables, the model still reported poor global fit statistics. This mobility variable was then dropped and correlated errors between decisions regarding household purchases and family visits, and household purchases and health were made (given modification indicators from Mplus and their methodological similarity in asking about decision making).44 Thus, the final latent variable of women’s agency included only the five items about women’s decision making. This resulting measurement model of women’s agency is also in line with previous studies using confirmatory factor analysis on women’s agency that suggest indictors of women’s household decision-making are conceptually distinct from measures of women’s freedom of mobility (Richardson et al. 2019). 44 In latent variable models, omission of correlated errors between observed variables with similar measures may not always harm fit, but their absence could lead to inaccurate results (Kline 2015). 157 APPENDIX B: DIRECT AND TOTAL INDIRECT EFFECTS IN STRUCTURAL EQUATION MODEL Table 3.3. Standardized WSLMV coefficients of structural equation model of women's empowerment, predictors of four endogenous variables in rural Jharkhand (N = 1597) Intimate Partner Internalized Agency Nutritional Direct effects Violence (IPV) oppression status Household size → -0.011 Women's age → 0.110* 0.120** 0.193** Women's caste → -0.082 0.046 0.010 0.329*** Men's gender values → 0.181** 0.162** -0.075 0.075 IPV → 0.315*** -0.301*** -0.243* Internalized oppression → 0.101* -0.072 Agency → -0.015 Total indirect effects Caste → IPV → -0.026 0.022 0.022 Caste → Internalized oppression → 0.002 -0.001 Men's values → IPV → 0.057* -0.049** -0.047* Men's values → Internalized oppression → 0.022 -0.016 IPV → Internalized oppression → 0.032 -0.023 IPV → Agency → 0.004 Notes: *p < 0.05; **p < 0.01; ***p < 0.001. Total indirect effects are the sum of all indirect effects on the endogenous variable that include the specific italicized mediator variable. 158 CHAPTER 5: CONCLUSION In this dissertation, I engaged with a variety of qualitative and quantitative methods, across three studies, to examine the contested terrain of women’s empowerment throughout all stages of the development process including conceptualization, implementation, and evaluation. My overall goal throughout these studies was to produce feminist research that aimed to improve women’s lives. In this concluding chapter, I synthesize these three studies into a cogent approach that summarizes my key findings and draws connections to its proceeding study. In doing so, I will explain each study’s contributions, including its feminist scope and relevant policy and program recommendations. I end by illustrating the cross-cutting findings that create a more holistic picture of women’s empowerment in the development discourse, what the implications are of claiming “empowerment” in the contemporary contested terrain, and recommendations for future feminist research. The goal of the first study (chapter two) was to conceptually clarify the way gender empowerment advocates (e.g., feminists and development practitioners) frame empowerment for women in the Global South. More specifically, I aimed to elevate the less acknowledged frames of empowerment by advocates from the Global South and identify the defining features that produce contested claims regarding whether a program or policy empowers women. Using the frame analytical approach to guide the inductive coding of 54 books and articles, written by advocates from all over the world, I revealed how empowerment frames are similar and different from one another in the way they elaborate on certain solutions (prognostic frame) to particular end goals (motivational frame) within a set of amplified problems (diagnostic frame). Through this frame analysis, I discovered and described 16 distinct collective action frames, or “stories,” 159 of women’s empowerment in the development discourse that derive from and elaborate on one of three master frames of empowerment: equal rights and participation, power, or agency. Despite Third World feminists offering the first master frame of empowerment for claims- makers to elaborate their grievances around their problem-solving schema of power, I found that most collective action frames are derivative of Kabeer’s (1999) master frame centered around agency. Notwithstanding, the most popular collective action frame used in the development discourse is the “Gender Parity” frame of empowerment: which tends to frame empowerment as instrumental based on essentialist assumptions and derives from the equal rights and participation master frame. The “Gender Parity” frame of empowerment is not only an elaboration of the mainstream empowerment frame credited with the institutionalization of the term (e.g., 1995 Beijing Platform for Action), but the frame itself consists of the most widely used measures today, including the Women’s Empowerment in Agriculture Index (WEAI) that uses quantitative indicators to evaluate women’s empowerment in the Global South. A combination of the WEAI’s increased popularity, as well as criticisms of the instrument in recent years, led to the creation of the project-level Women’s Empowerment in Agriculture Index (pro-WEAI) in 2019. Thirteen development projects across the Global South were selected to develop, implement, and pilot this new WEAI with evaluators from IFPRI (International Food Policy Research Institute); one of which was WINGS (Women Improving Nutrition through Group-based Strategies) with its implementing organizational partners in India. With the growing popularity of the WEAI family of tools in the development discourse, and my interest in further examining the ways empowerment is conceptualized and then implemented on the ground, WINGS offered a pertinent case study to showcase the multiple and possibly 160 contradictory ways in which women’s empowerment is interpreted and deployed by various actors in the implementation stage of the development process. Specifically, the second study (chapter three) provided a detailed study of the collaborative effort of the two Indian-based NGOs (non-governmental organizations) that partnered with IFPRI in the WINGS project. To study the contested terrain of empowerment in its conceptualization, operationalization, and implementation in this case study, I applied a grounded approach guided by frame analysis to 10 semi-structured interviews with NGO staff and triangulated this with content related to the project’s proposal, the FAAM (Facilitated Action Against Malnutrition) intervention materials, and WINGS evaluations. I found that the conceptualization of women’s empowerment led to the implementation of particular solutions (e.g., interventions) on the ground. In this case, implementers and evaluators similarly identified maternal (mal)nutrition as the ultimate problem and end goal of the project, and conceptualized empowerment as a means to this end. In practice, the concept of empowerment, as an instrument or platform for delivering nutrition behavior change communication (BCC), was implemented in the form of women’s self-help groups (SHGs). Notwithstanding, I also found contestation over empowerment’s operationalization and the idea that SHGs were the only part of the development project that made it empowering. While evaluators tended to solely point to SHGs as the way empowerment was implemented in practice, implementers suggested that the story-telling modules (i.e., BCC) played an important role in empowering women. I suggested that implementers’ understanding that empowerment also took the form of the modules stemmed from their operationalization of empowerment as achievements and not something separate from women’s nutrition; as opposed to evaluators operationalizing empowerment as agency and distinctly sperate from nutrition. One example 161 implementers described portrayed the indistinguishability of nutrition (and BCC) and empowerment (in SHGs) well: namely, the practice of women eating last in the family was an important issue identified that was addressed in the modules. Moreover, chapter three’s case study showed the importance of the research conducted in chapter two. First, the findings in the first study (chapter two) showed that the diagnostic frame largely informed the range of “reasonable” solutions (prognostic frame), providing evidence to support the frame analytic approach (Benford and Snow 2000). Second, I found that WINGS used the WEAI and pro-WEAI to evaluate women’s empowerment in terms of agency; one of the most commonly used quantitative indicators in the field of gender and development today (Alkire et al. 2013; Malapit et al. 2019; Yount et al. 2016). Despite its popularity as a measure, I found that agency did not align with either the implementers’ operationalization of empowerment as achievements, nor with the original conceptualization of empowerment as a process of collectivization that leads to action. Thus, the collective action frames discussed in chapter two can be used as a reference for future programs in development to confirm their diagnostic frame provides a range of possible solutions and end goals they agree with, and to make sure their operationalization and evaluation of women’s empowerment aligns with their conceptualization of empowerment. Doing so will ensure that their interpretation of “success” aligns with their understanding of women’s empowerment. With this being said, one of the major contributions of chapter two’s analyses is to the empowerment and development discourse. Namely, by describing the 16 collective action frames and their defining features that lead to contested interpretations and outcomes regarding a policy or program’s “successfulness” in empowering women, scholars and organizations can worry less about adopting the “wrong” empowerment frame. Chapter two can help activists and institutions 162 understand the tradeoffs of promoting and excluding certain solutions over others and select the one that aligns with their movement’s aims. At the same time, the findings presented in chapter two and three similarly contribute not only to the gender and empowerment discourse, but also the field of social movement frame analysis where their findings support Benford and Snow’s (2000) claim that most frames do not vary based simply on the topics or issues identified, but by the way they are discussed by claims-makers (e.g., feminists and organizations). In the third study (chapter four), I created a feminist of model of women’s empowerment in rural Jharkhand that estimated the direct and indirect effects of men’s values, women’s caste, age, internalized oppression, and experience of violence on women’s agency and her nutrition status. Within this model, I tested the assumption that agency leads to nutrition. By using structural equation modeling (SEM), I was able to treat women’s agency as both instrumental (a process to) to women’s nutrition as well as intrinsic (an outcome of empowerment processes), as a way to test both theories proposed by implementers and evaluators in chapter three’s case study. I found that the non-significant effect from women’s agency to her nutrition and the significant positive effect from women’s internalized oppression to her agency suggested that increasing women’s agency (measured as decision-making) is not a valid means to improving women’s nutrition in rural Jharkhand. In effect, rather than representing a form of power that is transformational in improving women’s strategic needs and goals, agency is more likely a reflection of a patriarchal bargain or form of burdened agency within highly patriarchal (and violent) households. These findings contribute to the gender empowerment literature most notably in supporting Kandiyoti’s (1988) seminal piece on women’s patriarchal bargains in certain areas including India, Lentz’s (2018) study on women’s burdened agency in deciding between violence or eating in Bangladesh, and feminists’ calls for including an analysis of 163 internalized oppression in evaluations of women’s empowerment (Kabeer 1999; Rowlands 1997). Methodologically and methodically, chapter four also helped close the current gap in the contested discourse between feminists primarily utilizing qualitative methods and the bias towards quantitative analyses in the field of development. Guided by feminist theory derived primarily by Indian scholars to conceptualize women’s empowerment as both a process of overcoming disempowering structures and the transformative result of that process, I created a feminist quantitative model of women’s empowerment to capture this complex socio-cultural system in the context of married women’s conditions in rural Jharkhand. My hope is that my feminist approach and use of quantitative tools that are analytically capable of capturing this dynamic and multidimensional concept (i.e., SEM and confirmatory factor analysis) can serve as a model from which quantitative analysts in the field of international development can borrow from in the future. Lastly, the study in chapter four addressed the very real and disproportionate high rates of undernutrition and intimate partner violence (IPV) that characterize women’s disempowerment in India. Findings from the structural model suggested that caste is the greatest predictor of women’s nutritional status, followed by women’s experience of IPV. Moreover, the estimated direct and indirect effects from the model revealed men’s gender values and IPV as having a central role in shaping all of women’s (dis)empowerment processes in rural Jharkhand. Together these findings indicate that women’s empowerment (including her nutritional status) in rural Jharkhand cannot improve without involving those with relative power - including the Indian government and men. Thus, my hope is that my findings from my feminist SEM model of women’s empowerment can contribute in the way Indian feminist Bhatt (1988) described as 164 “political visibility,” or the process of making policy makers aware of women’s problems and perspectives so they can act in their favor. Implications of Claiming Women’s Empowerment Within a Contested Terrain I conclude this dissertation by answering the question I posed in the introduction: What are the consequences of claiming women are “empowered” if we mean different things? Informed by the findings across all three studies, I answer this question with the knowledge I have gained throughout this dissertation research. Theoretically and methodically, claims and “evidence” of empowerment can become weak if we are not transparent in how we operationalize the concept. This is most notably demonstrated in the first study (chapter two) with the revelation that the most popular frame of empowerment used today (i.e., “Gender Parity”) contains the problematic aspects that initiated Third World feminists to promote an “empowerment” approach in the first place. Namely, the criticized top-down development approaches that did not take the interests of women into account and treated women as instruments to achieve development goals (e.g., population control) (Sen and Grown 1987). While the specific development goals have changed from population control to mostly those centered around income and nutrition, women and their empowerment today tends to be framed as instrumental to achieving development for all. Thus, the application of certain assumptions embedded in instrumental frames of empowerment may end up falling into the same trap as the problematic development schemes that Third World feminists called oppressive over 30 years ago. Practically, the findings presented in the third and fourth chapter of this dissertation highlight the dangers of conflating “widely used” frames and measurements of women’s empowerment as “the best” methods for improving women’s lives in the Global South. The 165 second study (chapter three) found that it was not just the diagnostic frame that influenced the implementation of women’s SHGs as a platform for delivering nutrition behavior change communication but was initially guided by the source of the project’s funding. Namely, the interviews from this case study revealed that the implementation and application of women’s empowerment as an instrument to achieve development goals came from the funder’s (Bill and Melinda Gates Foundation) interest in promoting women’s groups as a means to women’s empowerment and other goals (in this case those related to health and nutrition). Because of this, many evaluators acknowledged that one of the major limitations related to the project’s operationalization and evaluation of empowerment was that it did not conduct qualitative interviews at any point to better understand or measure empowerment from recipients’ perspectives as women living in rural India. Moreover, the third study (chapter four) tested not only the popular instrumental framing of women’s empowerment but also the most widely used quantitative indicator of empowerment used today, agency. Findings from my feminist-informed structural model indicated that those applying instrumental frames of women’s empowerment (to achieve nutrition for all) in rural Jharkhand are just that – instrumental. By re-directing the focus back on women, I found a non- significant relationship between women’s agency and her nutritional status, as well as a significant positive relationship between women’s internalized oppression and her level of agency. These findings revealed that improving women’s agency (measured as decision-making power) is not a valid means (or process) to achieve healthy nutrition nor is it a feminist indicator of women’s strategic empowerment in rural Jharkhand. By emphasizing relevant sources of power inequality relevant to women’s daily lives into the model, including women’s internalized oppression, her experience of IPV, and her husband/partner’s patriarchal values, I found that 166 women’s agency most likely reflects women’s burdened agency or a patriarchal bargain in a context of extreme physical violence that characterizes many (32 percent) married women’s lives in rural Jharkhand. Thus, programs that assume women’s decision-making agency will lead to strategic empowerment and/or other development goals, are likely to use women as instruments in development schemes that actually disempower them (Batliwala and Dhanraj 2007). While the findings presented in chapter four provide a generalized understanding of the dangers in applying popular assumptions and measures of empowerment in rural India, it provides a feminist framework by which quantitative analysts evaluating women’s empowerment in other regions can work from. This is important because as individuals claiming to support women’s empowerment, we have a responsibility to take its application seriously. This applies not only to the measurement and evaluation of women’s empowerment, but also in its implementation through programs and policies that aim to empower women. More research (that is similar to chapter three’s case study) regarding the contested terrain in which empowerment is implemented in practice is needed to explore the operationalization and application of the other 15 frames of empowerment found in chapter two, as well as across different contexts (e.g., government schemes, countries, and identities). I also want to encourage scholars to build on the groundwork I provided in the first study (chapter two), by continuing to use a grounded approach guided by frame analysis to expand our knowledge of empowerment frames for women around the world - across all languages - as well as those framed for women in the Global North. Ultimately, future research, implementations, and evaluations of women’s empowerment cannot lose sight of what women’s empowerment is all about: a feminist commitment towards improving women’s lives. Only through persistent feminist research that uncovers and respects the plurality that feminism embodies, will women experience empowerment for themselves. 167 REFERENCES Alkire, Sabina, Ruth Meinzen-Dick, Amber Peterman, Agnes Quisumbing, Greg Seymour, and Ana Vaz. 2013. The Women’s Empowerment in Agriculture Index.” World Development 52:71–91. Batliwala, Srilatha, and Deepa Dhanraj. 2007. "Gender Myths that Instrumentalize Women: A View from the Indian Front Line." Pp. 21-34 in Feminisms in Development: Contradictions, Contestations and Challenges, edited by Andrea Cornwall and Ann Whitehead. Zedbooks. Benford Robert D. and David A. Snow. 2000. “Framing Processes and Social Movements: An Overview and Assessment” Annual Review Sociology 26:611-39. Bhatt, Ela. 1989. "Toward Empowerment." World Development 17(7): 1059-1065. Kabeer, Naila. 1999. “Resources, Agency, Achievements: Reflections on the Measurement of Women’s Empowerment.” Development and Change 30(3):435–464. Kandiyoti, Deniz. 1988. “Bargaining with Patriarchy.” Gender & Society 2(3):274-289. Lentz, Erin C. 2018. "Complicating Narratives of Women’s Food and Nutrition Insecurity: Domestic Violence in rural Bangladesh." World Development 104:271-280. Malapit, Hazel, Agnes Quisumbing, Ruth Meinzen-Dick, Greg Seymour, Elena M. Martinez, Jessica Heckert, Deborah Rubin, Ana Vaz, Kathryn M. Yount, and Gender Agriculture Assets Project Phase 2 (GAAP2) Study Team. 2019. "Development of the Project-Level Women’s Empowerment in Agriculture Index (pro-WEAI)." World Development 122:675-692. Rowlands, Jo. 1997. Questioning Empowerment. Oxford: Oxfam. Sen, Gita, and Caren Grown. 1987. Development, Crises, and Alternative Visions: Third Word Women’s Perspectives. London: Earthscan Publications Yount, Kathryn M., Kristin E. VanderEnde, Sylvie Dodell, and Yuk Fai Cheong. 2016. "Measurement of Women’s Agency in Egypt: A National Validation Study." Social Indicators Research 128:1171-1192. 168