VOICES AGAINST VIOLENCE: EMPOWERING WOMEN TO ACCESS INFORMAL JUSTICE IN RURAL INDIA by JESSICA BROWN CAVAS A THESIS Presented to the Department of International Studies and the Graduate School of the University of Oregon in partial fulfillment of the requirements for a degree of Master of Arts September 2013 ii THESIS APPROVAL PAGE Student: Jessica Brown Cavas Title: Voices Against Violence: Empowering Women to Access Informal Justice in Rural India This thesis has been accepted and approved in partial fulfillment of the requirements for the Master of Arts degree in the Department of International Studies by: Dr. Anita M. Weiss Chairperson Dr. Sunil Khanna Member Dr. Colleen Chrisinger Member and Kimberly Andrews Espy Vice President for Research and Innovation; Dean of the Graduate School Original approval signatures are on file with the University of Oregon Graduate School. Degree awarded September 2013 iii Copyright 2013 Jessica Brown Cavas iv THESIS ABSTRACT Jessica Brown Cavas Master of Arts Department of International Studies September 2013 Title: Voices Against Violence: Empowering Women to Access Informal Justice in Rural India This thesis explores the effectiveness of a community-led response to domestic violence in tribal communities within Udaipur district, Rajasthan, India. Situated within an Indian Non Government Organization’s comprehensive women’s empowerment program, this research shows that the Women’s Resource Centers provide mediation to stop domestic violence and potentially prevent future violence through challenging cultural norms that perpetuate violence. Using a program evaluation approach, I compare how the program is intended to operate to my observations and interviews with community implementers, primarily at one Women’s Resource Center. This comparative lens serves to demonstrate the inevitable shifts and challenges that occur throughout implementation. I argue that development interventions addressing violence against women continually experiment with monitoring and evaluation tools, such as an outcome map, to capture their successes and setbacks to foster organizational learning and increase accountability to the intended beneficiaries of the program. v CURRICULUM VITAE NAME OF AUTHOR: Jessica Brown Cavas GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: University of Oregon, Eugene, USA Ohio University, Athens, USA DEGREES AWARDED: Master of Arts, International Studies, 2013 University of Oregon Master of Public Administration, Planning, Public Policy and Management, 2013, University of Oregon Graduate Certificate of Non-profit Management, Planning, Public Policy and Management, 2013, University of Oregon Bachelor of Science, Environmental Geography, 2004, Ohio University AREAS OF SPECIAL INTEREST: Empowerment, gender equality, non-formal education, participatory development, social change in South Asia PROFESSIONAL EXPERIENCE: Teaching Assistant, Department of International Studies, University of Oregon Eugene, 2013, 2010-2012 AmeriCorps VISTA (Volunteers in Service To America) Leader, Corporation for National and Community Service, Prescott, Arizona, 2007-2008 AmeriCorps VISTA (Volunteers in Service To America) Member, Corporation for National and Community Service, Prescott, Arizona, 2006-2007 Research Volunteer, CERCOPAN (Centre for Education, Research, and Conservation of Primates and Nature), Calabar, Nigeria, 2005 – 2006 Volunteer Coordinator, Inti Wara Yassi Wildlife Refuge, Villa Tunari, Bolivia, 2004-2005 vi GRANTS, AWARDS, AND HONORS: Graduate Research Grant, Centre for the Study of Women in Society, University of Oregon, 2012 Thurber Award for overseas research, Department of International Studies, University Oregon, 2012 vii ACKNOWLEDGMENTS I would like to thank my advisor and committee chair, Professor Anita Weiss, for her unending support, understanding and confidence throughout my thesis process. I would also like to thank my committee member Professor Colleen Chrisinger for her insightful advice and invaluable encouragement along the way. Thanks to Dr. Sunil Khanna for joining my committee at the last minute to contribute invaluable feedback and support. Warm appreciation goes to Hope Marston, Brooke Cagno and Julie Gray for always answering and anticipating my questions. Amidst deadlines and procedures, you gave more than answers, but listened and left me smiling after each interaction. Thank you to the Centre for the Study of Women in Society and the Thurber Grant from the Department of International Studies for the financial support that made this project possible. To my parents for their everlasting support and enthusiasm, despite minor reservations, for all of the choices I have made over the years that led me to this project. To the individuals, families, and communities that I stayed with during my fourteen months in Nepal and India, you welcomed me with such warmth and generosity I can hardly convey my gratitude. More than food and water, warmth or shade, you gave me a feeling of quiet interconnectedness that continues to inspire me. My immense thanks goes to the supportive and determined staff with whom I worked with at Seva Mandir, especially Laxmi Thakur and Swati Khardekar. Likewise, my deepest gratitude goes to my amazing translator and field assistant, Lokesh Kalal. Without you, none of this would have been possible. viii This thesis is dedicated to the men and women working with the Women’s Resource Centers in Balicha and Kharadiwara. Thank you for sharing with me your stories, your successes, and your challenges. Most of all, thank you for sharing with me your courage that makes for a future full of possibilities. ix TABLE OF CONTENTS Chapter Page I. INTRODUCTION .................................................................................................... 1 Violence Against Women: Public Outrage, Private Shame ................................... 1 Research Rationale and Methodology ................................................................... 5 Summary of Fieldwork and Research Questions ................................................... 7 Limitations of the Methodology ............................................................................ 9 Plan of the Thesis ................................................................................................... 15 II. DOMESTIC VIOLENCE IN INDIA, DEVELOPMENT AND MONITORING AND EVALUATION .................................................................................................. 17 Violence Against Women: The Case of Domestic Violence in India ................... 17 Conceptual Frameworks Guiding Intervention Strategies ..................................... 19 Violence Against Women and Development ......................................................... 22 Help-Seeking Behavior and Women’s Access to Justice ...................................... 23 Community Led Informal Justice Responses ........................................................ 27 Monitoring and Evaluating Violence Against Women Interventions .................... 32 III. SITUATING AND CREATING SEVA MANDIR’S PROGRAM THEORY ..... 38 Domestic Violence and Tribal Women in Udaipur District .................................. 38 Research Locations: Kherwara Block, Balicha and Kharadiwara WRCs ............. 39 Outcome Mapping ................................................................................................. 42 IV. COMPARING THEORY WITH PRACTICE: STRENGTHS AND CHALLENGES ........................................................................................................... 49 The WRC Team ..................................................................................................... 49 x Chapter Page Location ................................................................................................................. 53 Fees ........................................................................................................................ 54 Stipend and Bank Account ..................................................................................... 57 New WRC Leaders ................................................................................................ 58 Cases – What Prevents Women From Coming to the WRC? ................................ 59 Tolerance of Violence ............................................................................................ 61 Documentation ....................................................................................................... 62 Balicha ............................................................................................................. 62 Kharadiwara ..................................................................................................... 65 Gaps in Information: Who Does the WRC Serve? When Are Cases Registered? ............................................................................................................. 67 V. PERSONAL EMPOWERMENT, HELP SEEKING BEHAVIOR AND IMPACT INDICATORS ............................................................................................................. 68 Personal Empowerment of WRC Leaders ............................................................. 68 Help Seeking Behavior .......................................................................................... 71 Impact Indicators ................................................................................................... 78 VI. CONCLUSIONS AND RECOMMENDATIONS ................................................ 82 Key Findings .......................................................................................................... 83 Considerations of Future Research ........................................................................ 87 Recommendations ................................................................................................. 89 Significance for Development ............................................................................... 93 APPENDICES ............................................................................................................. 96 xi Chapter Page A. WOMEN AND CHILD DEVELOPMENT UNIT LOGICAL FRAMEWORK ANALYSIS ............................................................................. 96 B. SO THAT CHAINS .......................................................................................... 97 C. THEORY OF CHANGE OUTCOME MAP .................................................... 105 D. INTERVIEW GUIDES .................................................................................... 106 Seva Mandir Staff ........................................................................................... 106 WRC Leader ................................................................................................... 108 WRC Team Member ........................................................................................ 110 Case .............................................................................................................. 111 REFERENCES CITED ................................................................................................ 115 xii LIST OF FIGURES Figure Page 1. Summary of Field Work .............................................................................................. 8! 2. Number of Receipts, 2007- Oct 2012 ...........................................................................63 3. Balicha WRC Receipts by Month, 2007-2012 .............................................................67!! xiii LIST OF TABLES Table Page 1. Balicha Receipts 2007 – Oct 2012 ............................................................................ 64! 2. Kharadiwara Receipts 2007 - Oct 2012 .................................................................... 66! 3. Case Studies – Help Seeking Behavior ..................................................................... 71! 4. Case Studies - Outcomes .......................................................................................... 78! 1 CHAPTER I INTRODUCTION Violence Against Women: Public Outrage, Private Shame In December 2012, the brutal gang rape and death of a young woman on a bus in New Delhi sparked an unusual public outrage that has jolted the issue of violence against women in India into the spotlight, for the entire world to see. Her death serves as a tragic reminder of the global epidemic of violence against women, including rape, domestic violence, mutilation, murder and sexual abuse that affect women’s everyday lives. Worldwide, women aged 15-44 are more at risk from rape and domestic violence than from cancer, motor accidents, war and malaria (UNDPI 2009). While the protests in Delhi that followed the December rape case successfully held issues of violence against women in the forefront of the media, drawing attention from the Indian government and the international community, violence occurring within the so-called private sphere did not enter the public discourse. Heise (1994), who declared gender-based abuse a global epidemic in 1993, argued that the most pervasive form of gender violence is abuse of women by intimate male partners (p.136). Worldwide, at least one in three women are beaten, coerced into sex or otherwise abused by an intimate partner throughout the course of her lifetime (UNDPI 2009). In India, 40 percent of women of reproductive age have experienced physical, sexual, or emotional violence by their husband (IIPS 2007). Considering such a high prevalence of domestic violence, this issue remains a priority for development 2 practitioners, especially those working to address gender inequalities and women’s status (Ackerson and Subramanian 2008). This thesis focuses on domestic violence among tribal1 communities in Udaipur district in the state of Rajasthan, India. In particular, it looks at the effectiveness of a comprehensive women’s empowerment program that responds to domestic violence with support and conflict mediation, which is supported by a local Non Government Organization (NGO). Seva Mandir is an Indian NGO that “focuses on creating the social, institutional, and livelihood base for a democratic and participatory approach to development that benefits and empowers the poorest sections of society” (Seva Mandir 2012). Since 1982, Seva Mandir has been working to empower women socially, economically and politically in order to foster women’s participation in development. Seva Mandir’s participatory approach believes in building women’s capacities so that they can become agents of their own development. This is primarily done through the formation and strengthening of women’s associations at various levels. Building on over thirty years of experience, Seva Mandir’s Women’s Resource Centers (WRCs) were implemented in 2006 to address the high prevalence of domestic violence within the organization’s work area (Thornhill 2004). This thesis focuses on the WRCs as an intervention to address domestic violence. In 2003, Seva Mandir conducted a household survey with 300 women from majority tribal areas to assess the magnitude of domestic violence within their work area as well as to better understand the help-seeking behaviors of these women. Two-thirds of respondents reported experiencing domestic violence, defined as having been slapped, 1 In 1950, the Constitution of India recognized the British list, or schedule, of castes or tribes. Sometimes referred to as adivasis, “original inhabitants”, tribal communities in Udaipur district include the Bhil, Meenas and Garsaias. 3 beaten or hit by someone in their home (Thornbill 2004). Fifty-seven percent of respondents reported seeking help from a family member, and 38 percent reported seeking help outside the family (Thornbill 2004). At the time of this survey, Seva Mandir was doing little to directly confront the issue of domestic violence, outside of staff members providing support to individual women keeping help. However, one women’s group reacted to a case of domestic violence by sending a group of women to the home to demand that the husband stop beating his wife. With the success of this incident, Seva Mandir began process of conceptualizing a comprehensive response to domestic violence to implement throughout their work area. In 2006, Seva Mandir created Women’s Resource Centers (WRCs) as a space for women to share their problems and find solutions to conflict, including violence within their household or between households. The WRCs directly address a critical gap in resources available to tribal women living within Udaipur district. Tribal women face numerous barriers in accessing the traditional justice system, the caste panchayat2, as well as the state-led formal justice system. The WRCs offer an alternative informal justice system3, backed by Seva Mandir’s comprehensive women’s empowerment programming. Following Seva Mandir’s participatory approach and emphasis on the capacity building of the women who are members of the women’s associations, Seva Mandir structured the WRCs to be run by three local women leaders. These women leaders are elected by the members of the women’s groups and then receive extensive trainings on leadership, counseling, legal 2Traditional or customary justice system at the village level made up of elders who hear cases in an open forum where all members of the caste group are entitled to take part 3 “The resolution of disputes and the regulation of conduct by adjudication or the assistance of a neutral third party that is not part of the judiciary as established by law and/or whose substantive, procedural or structural foundation is not primarily based on statutory law” (UN, 2012, p.8) 4 issues and government schemes. They are also trained to deal with police and government officials so that they can connect women with resources in their community. The WRCs are open three days a month, and since 2006, records indicate that the WRCs have provided support to at least 400 cases of violence against women across Udaipur district (Seva Mandir 2011). As an initiative led by women for women, the WRCs are uniquely situated to not only address the immediate needs of women, but also to challenge the long-standing discriminatory cultural norms that perpetuate violence against women. The issue of violence against women holds great importance for the development community, as an impediment for growth, a hindrance to participation and a fundamental obstruction to women’s agency (Sen 1999). However, without consistent and accurate monitoring to enable rigorous evaluations, a lack of data persists making it impossible to agree upon recommendations for programming based on lessons from current projects. Consequently, the international community is unable to strategically direct resources toward this critical issue. Today, most programs working to prevent and respond to violence against women lack monitoring systems that lay the groundwork for evaluating outcomes in the intermediate and the long term (Bloom 2008). This is often the case with community-led initiatives, like Seva Mandir’s, which is being implemented by women living within the communities that they are serving. Considering the added challenges of incorporating monitoring within community-led programs, these programs have rarely been documented, let alone evaluated. As a result, the invaluable learning that can emerge from capturing the progress of these initiatives, as well as the setbacks, is lost. This loss 5 reverberates globally as the international movement to address violence against women struggles to be confident which interventions should be supported. My research aims to fill this gap in the literature, adding to scarce efforts to capture how community responses are successful in addressing the needs of women facing domestic violence. Research Rationale and Methodology Considering the lack of documentation and understanding around the effectiveness of community-led responses to domestic violence, I decided to work with an NGO implementing one such project in Udaipur district, Rajasthan, India. Working primarily in rural areas with tribal women, Seva Mandir is engaged in efforts to comprehensively address gender inequality through social and economic empowerment initiatives. In 2006, Seva Mandir began implementing Women’s Resource Centers (WRCs) to address the high incidence of domestic violence within their work area (Thornhill 2004). To date, the WRCs have not been evaluated and Seva Mandir staff are looking for ways to capture the current functioning of the WRCs across sites, as well as to capture the impact of the WRCs on women who access the program. With this intention in mind, I focused my research on capturing the current functioning of the program with an eye to identifying potential strategies for determining the impact in the future. To investigate this program and ways of determining its effectiveness, I used qualitative methods; mainly, direct observation in the field and interview guides to speak with staff, community leaders implementing the program as well as women who approached the program for help resolving conflict. 6 I approached this research from a program evaluation perspective. Rossi, Lipsey and Freeman (2004) consider program evaluation as making assessments in one or more of five domains: needs assessment, program theory and design, implementation assessment, impact assessment, and efficiency assessment. The authors do place these domains in somewhat of an order reminiscent to the project cycle. Evaluations assessing program theory and design, for example, are considered foundational for later types of evaluation, such as implementation or impact assessment. Because program theory is rarely explicit, the first task of an evaluator is to articulate this implicit theory of how the program is intended to work (Rossi, Lipsey and Freeman 2004). Within the field of international development, the theory of change approach is increasingly being used to support development outcomes. Evidence from the use of theory of change tools, such as outcome mapping, in this field suggests that the approach clarifies concepts of impact within programs and can serve as a useful framework for monitoring, evaluation and long-term program assessment (Vogel 2012). I decided to use a theory of change approach in order to explicitly articulate the theory behind Seva Mandir’s violence intervention program, the WRCs. Since the WRCs are nested within Seva Mandir’s comprehensive women’s social empowerment initiatives, it is misleading to consider the program as acting in isolation to address domestic violence. The relationship between the WRCs and these empowerment initiatives is implied, and quite complex, making it difficult for someone other than staff directly working with the Women and Child Development (WCD) unit to understand the theory behind how the WRCs are intended to address the high prevalence of domestic violence in their work area. 7 The method of using a theory of change approach, creating an outcome map, serves to articulate the program’s implied theory, which can serve as a theoretical baseline to compare with my empirical observations. As a tool for my data analysis, the theory of change thinking essentially facilitates a process for comparing and contrasting what is occurring at my research sites with how the NGO envisioned the program to be operating. Summary of Fieldwork and Research Questions I spent a total of eleven weeks with Seva Mandir, three weeks of planning and research based in Udaipur, seven weeks in the field, and one final week analyzing and presenting my findings to the WCD unit. During the seven weeks of fieldwork I spent 32 days with my field assistant/translator. Figure 1 on the following page shows a summary of fieldwork divided between Balicha and Kharadiwara, located within Kherwara block. These two WRCs were selected from nine WRCs in operation within Seva Mandir’s work area as of August 2012. Kherwara Block was selected to be my research site based on there being three WRCs within this one area, Balicha, Kharadiwada, and Sagwada. Initially, I intended to work with all three to compare across sites, but dropped Sagwada WRC due to its remote location and it having only been implemented for one year. The other two WRCs, Balicha and Kharadiwada were both implemented in 2006 and are geographically closer together, making these sites more feasible considering transportation challenges, and a more logical comparison. Ultimately, the bulk of my data comes from Balicha because the women working here were much more responsive and available compared to the women working with the 8 Kharadiwada WRC. As much as possible I added observations and interviews from Kharadiwada to serve as comparison. Fig. 1: Summary of Field Work Summary of Field Work Balicha Kharadiwara • 21 recorded individual interviews • 2 group interviews • 3 WRC meetings • 1 SHG meeting • 1 cluster level meeting • Additional conversations unrecorded • 8 recorded individual interviews • 1 WRC meeting • Additional conversations unrecorded The research questions that I used to focus my fieldwork were: 1. What factors determine when a woman approaches the WRC and registers a case? What prevents women from coming to the WRC? 2. How does the current functioning of the WRC compare with what was intended initially for the program? 3. In what ways does Seva Mandir facilitate documentation and monitoring to track the impact of the WRC? 4. Overall, does the WRC effectively resolve family conflict and reinstate women’s rights within the family? These four research questions reflect my understanding of domestic violence interventions in rural India based on reports written by previous Seva Mandir volunteers and interns as well as limited academic literature that I had access to while based in Udaipur. This literature review occurred during my first three weeks while preparing a 9 new project proposal and making logistical arrangements for the field. The first question is more or less a direct response to Seva Mandir’s concern about the overall number of cases comes to the WRCs decreasing. The second stems from my lens of program evaluation, in looking at my work as being along the lines of a process evaluation. The third stems from my review of literature indicating that there is little documentation of community-level domestic violence responses as well as Seva Mandir’s interest in capturing the impact of their program. The fourth question comes from a study by the International Center for Research on Women (ICRW) in 1999, which found three separate community-led responses to domestic violence did in fact reinstate women’s rights within the family. Considering how the WRCs have the same objectives as these other responses, I was curious if the WRC had a similar outcome. In Chapter II, I discuss the ICRW study in further detail. Limitations of the Methodology There are several limitations that will be mentioned throughout the discussion of my findings, but this section will address some limitations associated with my methodology, including barriers and challenges while collecting data in the field. The most consistent and obvious limitation I experienced throughout my eleven weeks with Seva Mandir was the language barrier. When I arrived on site in Udaipur I had roughly early intermediate Hindi language proficiency. While my language level continued to improve, especially my listening comprehension, my speaking was not sufficient to communicate at the level required for interviewing. Regardless, only two people that I interviewed spoke Hindi, along with a few staff in the main office in Udaipur. With the exception of one woman, all women at my research sites could 10 understand Hindi but replied in Waagri, their local language. There was one women could speak fluent Hindi, which enabled us to have some superficial conversation without my translator. Otherwise, I could not understand when women responded in Waagri at all. This meant I could not actually have conversations with women without my translator. Talking through a translator is clearly not the same as interacting as I would on my own. This significantly changed the concept of building rapport. And because my translator was a young man, this altered the rapport building process in ways I cannot fully understand. He was from a nearby village and has worked with women and girls in the area discussing violence against women for his own NGO, which did give him a level of comfort and confidence that certainly helped us build rapport together. Because the language barrier hindered my communication, my fieldwork process became much more purely observational at times than I originally intended, which could be considered strength. During the meetings I attended it felt natural to more or less purely observe, since speaking with my translator at this time would have felt disruptive. Our procedure became one where we both were engaged, him listening and me observing, and then we would catch up as small discussion occurred or after the meeting. Occasionally I had concerns that I was not capturing what was going on in detail, but I reminded myself that observation of the current functioning of the program was certainly an important task that aligned with the aim of my project. Finally, the language barrier emerged as a limitation most evidently during interviews. Women would often not answer questions directly, going on long tangents to tell stories. During a story it was difficult to find a balance between me feeling completely lost without any translation and interrupting the flow of the interviewee’s 11 story. Our style tended to avoid interrupting at all times, which felt uncomfortable for me initially, but in the end I learned to trust that the long story could be summarized succinctly. However, this made keeping the conversation on topic a challenge for both my translator and myself, and difficult to ask follow up questions due to my delayed response receiving the translation. Fortunately, all participants agreed to my recording the interviews so we were able to go back and transcribe interviews together, typically the evening immediately following the interview. In this way we could capture any areas that required follow up as well as discuss what we learned as a team in doing the interview. Undoubtedly, this project would not have been possible without a local translator available nearly full-time throughout my seven weeks in the field. Without a local translator familiar with the cultural norms of the research site, there would have been an additional barrier in terms of two outsiders working together to adjust to this particular cultural context. Because he had previously worked with Seva Mandir, as well as currently runs his own NGO addressing youth development and violence against women in a nearby village to my research site, this mitigated the language and cultural barrier to a great extent. His personal interest in learning more about the WRCs, their challenges and effectiveness, as well as how the program intersects with his own work, clearly contributed to his ability to build rapport. Logistically, transportation in the field created limitations related to whom I was able to meet and interview. Transportation in Kherwara block is time consuming and somewhat unpredictable. Traveling from the block office to the research site could take up to three hours depending on jeep timings. Traveling from Karawada, where I stayed overnight while in the field, to the research sites was quicker and less costly, but meant 12 being disconnected from Seva Mandir staff at the block office. In the end, I decided to stay in Karawada and therefore had little interaction with block staff. This in itself may be considered a limitation, as I did not have many in depth conversations with staff. Therefore, the bulk of my interviews were in the field with the women implementing the program. The villages in Kherwara block are geographically spread out, with individual houses being set apart sometimes quite far from paved roads. Even with staying close to the research sites, it was required that we visit various people in their homes for interviews. The WRC is only open three days a week, and women with cases often could not leave their homes to come to the WRC for an interview. Walking to people’s homes in September heat was not practical, so a motorcycle became more and more crucial as we began our fieldwork. Using public transportation and walking meant only reaching one or two people in a day, and often people were not there or sick when we would arrive. Eventually, we arranged renting a motorcycle for the last two weeks of fieldwork so that we could meet four or five people in a day. Another limitation occurred due to seasonality. The months of September and October immediately follow the monsoon season. During this time the temperature becomes quite hot and dry again after the more cool and wet months of July and August. The changing weather is considered the cause of many illnesses during this time. Throughout my time in Balicha all of the women leaders were sick with colds and flu- like symptoms. Sometimes they could not even speak with me when I arrived at their home because they were so sick, and they definitely could not travel. Illnesses also sometimes resulted in death for the elderly. When there is a death, family members are 13 mourning for at least a week, and it is inappropriate to approach family regarding an interview at this time. In addition, by mid-October several people travel to visit relatives for festivals. These reasons combined meant that on several occasions we could not visit people’s homes or talk to people that I would have liked to interview. Another limitation that continually came up was a concern regarding women’s safety. Ultimately, this prevented me from doing more case studies, specifically domestic violence cases. As I spent more time in the village I began to question the impact of our interviewing women who had been experiencing violence. I continually asked staff and the women leaders about the best location to do these interviews. We agreed it would be best to bring people to the WRC, or perhaps even a more neutral location if possible. I was concerned about interviewing women in their homes because other family members could be nearby, preventing women from speaking freely about the case or their current situation. Because few women had cell phones, this again became a transportation issue simply to set up an interview. From the perspective of the women leaders and cluster level workers, they initially did not seem to have any concerns about us going and talking to women who had cases in their homes. However, they did express that it was better if one of them came with us. And this brought us back to issues in finding a healthy woman leader who was available to join us. In the end, I spoke with only four resolved cases, which were all people who were physically able to come meet me near the WRC in Balicha. These limitations in language, transportation, seasonality and the mobility of participants reflect Chambers’ (2008) discussion of the biases that impede outsiders’ contact with rural poverty. Our challenges to talk to people due to seasonality and 14 transportation inevitably led to us relying on those who were most accessible via paved roads or public transportation as well as healthy and available for interviews. While Chambers (2008) discusses seasonal bias regarding researchers doing field visits when it is most convenient or pleasant, such as following harvest, this was not the case for my situation. The barriers we faced were more due to people being busy in their fields or suffering from seasonal illness. The bias of paved roads certainly impacted the data collection process. The days we ventured onto dirt roads in search of a house by foot often resulted in a sense of lost time because the person was not home. The four case studies interviewed, for example, were all with people who came to meet us in the central area of the Village, on the paved road. While I worked to off set biases as best I could, I am critically aware that these biases distort my findings and reinforce the need to encourage future documentation efforts to address these biases more directly. Finally, there are limitations in using the theory of change approach to explicitly articulate a program theory as an external evaluator. The theory of change thinking can be used for many different purposes, at various levels, ranging from exploring types of implementation options to critically reviewing an established program either as a stand alone effort or within a broader sector-wide strategy. For my purposes, the theory of change thinking is similar to how an evaluator would use the approach. An evaluator tends to highlight differences between the design-theory stage and the implementation model. Stern and colleagues (2012) states that this is a rapidly developing technical area of evaluation literature. This way of using the theory of change thinking is different from how an implementing agency would use this approach. 15 Implementing agencies can use this thinking to support their decision-making and underpin performance management or evaluation frameworks. The agency incorporates extensive detail regarding the context, including power relationships and networks between actors. This systems perspective illuminates how change in one area contributes to changes in another area, emphasizing links between attitude and behavior change and how these are aggregated to influence change at a macro-level and beyond. While I certainly attempted to this as best as I could based on my observations and review of organizational documents, it is important to note that my theory of change thinking is subjective and from an evaluator’s perspective. Plan of the Thesis This chapter introduces my thesis topic by framing my project within the critical global issue of violence against women and the challenges that the development community faces in addressing this issue. Having briefly described Seva Mandir’s community-led informal justice initiative, embedded within a comprehensive women’s empowerment program, this chapter also discusses my research rationale and methodology, summarizes my fieldwork, states my initial research questions and the limitations associated with my methodology and approach. Chapter II discusses literature regarding domestic violence, including key definitions, conceptual frameworks, development significance, help-seeking behavior and women’s use of informal justice systems. A final topic explored within my literature review is the challenges of monitoring and evaluation violence against women interventions. Chapter III describes the process of using a theory of change approach, particularly an outcome map to articulate the implied theory behind the Women’s Resource Centers (WRCs). In addition, 16 this chapter further situates the issue of domestic violence as it pertains to Seva Mandir’s work area, and specifically how the intervention evolved in my research site. This chapter also notes how staff defines and describes domestic violence in my research area. Chapter IV is a discussion based on comparisons between the original program theory and the current operation of the program over six years later. Within this chapter I integrate quotes and observations from my fieldwork to highlight key findings in terms of strengths and challenges. Chapter V returns to my original research questions and the needs of Seva Mandir to explore themes related to the impact of the WRCs and help- seeking behavior. Chapter VI reiterates and posits additional key findings and the implications of these findings, along with recommendations for Seva Mandir’s WRCs and broader women’s empowerment program. 17 CHAPTER II DOMESTIC VIOLENCE IN INDIA, DEVELOPMENT AND MONITORING AND EVALUATION “A woman’s human rights framework equips women with a way to define, analyze and articulate their experiences of violence, degradation and marginality…and very importantly, the idea of women’s rights provides a common framework for developing a vast array of visions and concrete strategies for change.” - Bunch and Frost, 2000 p. 1 Violence Against Women: The Case of Domestic Violence in India In the 1990s, the women’s movement successfully harnessed the slogan “women’s rights are human rights”, to bring the issue of violence against women to worldwide attention. Women around the world experience violence in various forms during their lifetime. In South Asia, the cultural preference for boys results in sex selective abortions, nutritional neglect during childhood, lower educational levels and higher overall mortality rates for women and girls in comparison to their male counterparts (Heise 1989). Over the past two decades, women’s groups have organized to combat these acts, considered “gender based violence” or “violence against women.” The terms “gender-based violence” and “violence against women” are often used interchangeably throughout the literature and by practitioners; however, it should be recognized that men too can be victims of gender based violence. Gender based violence emphasizes that the violence occurred due to the victim’s gender, usually as a result of vulnerabilities due to subordinate status (UN 2012). Following the United Nations’ use of the term “women”, this refers to females of all ages, including girls, unless specified otherwise. This thesis focuses on violence against women committed by both men and 18 women. Research among women from various communities in South Asia, the Middle East, as well as North American immigrant communities (Latina, South Asian, East Asian) has documented the role of elderly family members, particularly mother in laws, as potentially exacerbating family-based violence against women (Chan et al., 2009 in Krishnan 2005). One of the problematic issues with research regarding domestic violence, especially in cross-cultural contexts, is the definition of “domestic violence.” There are numerous definitions used throughout the literature that more or less vary in the detail of the definition. The definition I use for this project is, “Any act of physical, sexual or psychological abuse, or the threat of such abuse, inflicted against a woman by a person intimately connected to her through marriage, family relation or acquaintanceship (Burton, Duvvury and Varia, 2000, p. 5). I selected this definition from an ICRW report, published in collaboration with the Centre for Development and Population Activities (CEDPA), who worked in partnership with ICRW on projects aimed to improve the lives of women in developing countries. I feel this definition represents what is commonly perceived as domestic violence within the international community. Of course, this does not mean that those women experiencing such acts of violence always consider themselves victims of domestic violence. Cutting across the boundaries of caste, class, religion, and region, domestic violence is the most common type of violence against women prevalent in India (Ghosh 2007). It is estimated that two of every five women in India experience physical abuse, regardless of her educational level (ICRW 1999). According to the most recent Indian National Family Health Survey (2007), 40 percent of women of reproductive age have 19 experienced physical (35 percent), sexual (10 percent), or emotional (16 percent) violence by their husband. Despite being widespread, the prevalence of violence is considerably high among socioeconomically disadvantaged women, 52 percent of women in the lowest wealth quintile compared to 21 percent of women in the highest (IIPS, 2007. In addition, a high prevalence of violence has been documented among women during pregnancy (Ahmed, Koenig and Stephenson, 2006). Conceptual Frameworks Guiding Intervention Strategies Domestic violence literature uses three interrelated conceptual frameworks of women’s empowerment, status, or autonomy, to investigate the occurrence and distribution of violence. Indeed, the presence of domestic violence can be considered an indicator of a woman’s relatively lower status, disempowerment, or lesser autonomy. The concepts of women’s “status”, “empowerment” or “autonomy” vary among authors, and are often used interchangeably. Each concept is useful in considering when and why women experience domestic violence. Malhortra and colleagues (2002) define status as, “the expansion in people’s ability to make strategic life choices in a context where this ability was previously denied to them” (p.6). Hobcraft defines autonomy as, “increasing the means and ability of females to choose and control outcomes, implying a shift towards greater individual agency” (Hobcraft in Presser and Sen, 2000, p. 161). It is widely accepted that empowerment encompasses both a process and an end product, although arguably never complete, at the individual and collective levels. Batliwala (1994) defines empowerment as, “the process of challenging existing power relations, and of gaining greater control over the sources of power” (p.130). Since indicators of empowerment processes are more 20 difficult to quantify, the concepts of autonomy or status lend to more measureable indicators, such as women’s mobility or decision-making authority in the home. Women’s status has been seen to have variable effect on the incidence of domestic violence due to the interaction between women’s status and different levels of patriarchy (Mogford 2011). Considering distribution among Indian states, for example, women’s experience of violence ranges from six percent in Himanchal Pradesh to 56 percent in Bihar, with 40 percent or more in Rajasthan, Madhya Pradesh, and Tripura (IIPS 2007). Koenig et al. (2003) argue that women in more egalitarian settings that achieve greater status enjoy less risk of domestic violence compared to women in highly patriarchal settings whose gains in status pose a threat to social norms that condone women’s subordination. As a result, what could be considered a protective factor against violence in certain contexts, may serve as a risk factor in a different context. Mogford (2011) considers the relationship between women’s status and domestic violence and illustrates that the effect of status on violence is irregular, varying both within and across the different locations of status that women occupy, such as within the household, the community or the workplace. Women’s employment, for example, is considered a protective factor that enables women to leave abusive situations since her earnings could not only give her self-sufficiency and access to services, but also connect her with social networks of support (Hindi and Adair, 2002). While this holds true in certain cases, multi-country studies on domestic violence have found that women’s contributions to household finances could place them at heightened risk (Kishor and Johnson, 2004). Other risk factors that have been explored relating to women’s status or 21 autonomy include women’s decision-making autonomy, women’s mobility, and women’s status relative to her husband. These factors fit within the microsystem level of the ecological framework for violence against women, first created by Heise, in 1998. The ecological framework for violence against women has become widely accepted as a tool to inform violence intervention strategies. Building from previous models, the ecological framework integrates factors that are associated with why men become violent, as well as factors that contribute to why women become targets of violence. Heise (1998) delineates four interdependent levels of risk that have guided intervention strategies: individual, microsystem, exosystem, and macrosystem. An example at the individual level associated with both men becoming perpetrators and girls becoming victims is witnessing abuse or experiencing abuse during childhood. At the microsystem level, male control over family wealth, or low women’s decision-making autonomy, places women at risk, along with educational and employment status disparities between men and women in an intimate relationship. At the exosystem level, which encompasses institutions and social structure that embed the microsystem, high male unemployment or underemployment can place women at risk of violence (Heise, 1998). Considering the macrosystem level, the broad set of cultural values and beliefs that inform the other layers, Heise (1998) argues that violence is more likely to occur within societies that have rigid gender roles and where the use of violence within the family to address conflict has been normalized. It is important to note that the ecological framework includes factors empirically related to rates of violence against women and 22 girls; however, they are based on an incomplete research base and do not suggest causality. The ecological framework illustrates the need for an integrated, coordinated response from multiple sectors to address violence against women. Violence Against Women and Development Similar to the ecological framework, the costs and consequences of violence against women can also be articulated by focusing either on the rights of the individual, the way individuals contribute to families and economies, or how public resources are utilized. Women’s rights advocates argue that in order for women to participate in the process of development or benefit from development efforts, women must be free from violence (Burton, Duvvury and Varia, 2000). In linking violence against women within the aims of development, this further urges commitment from both governments and the international community to address the issue of violence against women. Sen (1998) identified three different approaches to violence used by development researchers and practitioners that remain relevant today: (1) violence limits the efficiency of development; (2) violence hinders participation; (3) violence is contradiction to all human development (p.7). In the first approach, women are viewed as a resource for development; therefore, anything that prevents women from exercising their capacities to contribute is considered a loss economically. Under this approach direct and indirect costs of violence can be calculated, which is particularly effective in galvanizing the government’s involvement in coordinated response to violence in the interests of saving public resources. The second approach considers women’s participation. Sen (1998) problematizes the good intentions of participation within contemporary development, arguing that 23 practitioners often ignore the gendered difficulties of engaging women in development activities (p.10). Rao, Gupta and Weiss (1998) argue that practitioners must be aware that violence or the threat of violence can hinder women’s ability to make choices, whether this means using contraceptives, discussing her health needs with a development researcher, or being able to leave the home long enough to participate in a women’s empowerment program. The third approach builds on the second approach, but further states that the mere existence of violence is a contradiction to the goals of development. According to the United Nations Development Program (UNDP), development is defined as a, “concern with the enlargement of people’s choices” (UNDP 1990). Since violence against women is an act of control, against women’s choice, this renders violence against women a central issue to development. From this perspective, development efforts aim to improve women’s individual agency, contributing to women’s expansion of choice. All three approaches illustrate the significance of violence against women for the development community; hence, the demand for recommendations regarding effective ways to respond and prevent violence against women. Help-Seeking Behavior and Women’s Access to Justice Research is limited from less developed countries regarding how women, their families, and their neighbors respond to violence (Schuler, Bates and Islam, 2008). Domestic violence has remained largely unchallenged within Indian society due to an enduring perception that such violence is a private concern and not an issue of social responsibility (ICRW 1999). Throughout much of India it remains culturally acceptable for the male head of a household to commit acts of violence against members of the 24 household in order to dispense discipline and maintain rule of authority within the family, particularly when a woman does not follow her husband’s orders (Krishnan 2005). This is not to say that homes are not places of love, support and bonding; however, homes are frequently sites of violence where women are undeniably the primary victims (ICRW 1999). The National Family Health Survey (2007) found that 54 percent of women believe that it is justifiable for a husband to beat his wife. This is supported by the IndiaSafe study where 58 percent of women who had experienced violence did not leave their home because domestic violence is a “normal” part of married life (INCLEN 2000). Women’s persistent acceptance of abuse illustrates the level of women’s status or disempowerment as well as the limited resources available for help. Consequently, women often remain in abusive relationships out of shame, fear of retaliation or separation from their children, lack of emotional support or lack of economic alternatives (Schuler, 2008). By the late 1980s, women’s rural empowerment initiatives emerged in India where women organized and engaged in reflective processes, gaining increased awareness of their condition and position in society. As a result of this grassroots movement, rural women throughout India have become more involved in the public arena as well as gained entitlements to legal rights. However, these gains have also resulted in an increased rate of violence against women in both the private and public spheres (Iyengar, 2007). Especially in rural areas, where men are more likely to have control over resources within the family, women have limited resources available to address violence (Pinnewala, 2009). Referring back to Heise’s ecological framework (1998), the exosystem includes 25 both the formal and informal social structures within the community or the state that could serve to support women experiencing violence. Building off of Amartya Sen’s capabilities approach to social development, Pyles (2008) argues that women who experience violence with less freedom and access to institutions are especially vulnerable to experiencing fear, shame and isolation, which will prevent women from seeking help, thus remaining in violent relationships. Therefore, practitioners addressing domestic violence often work to address the barriers women face in accessing justice. Women in numerous developing countries throughout the world who have the courage and resources to approach the formal justice system are faced with insensitive attitudes and inadequate redressal mechanisms such as low conviction rates and irrational procedures requiring evidence of abuse (IDLO 2013). Although Section 498A of the Indian Penal Code prohibits cruelty and mental torture to women within a marriage, particularly linked with dowry4, this law remains narrowly interpreted and rarely enforced by police (Burton, Duvvury and Varia, 2000). Furthermore, the outcome of the formal justice system, whether divorce or conviction, is often socially and economically undesirable; rather, women prefer to end the violence within the relationship (Pinnewala, 2009). As a result of these cultural and structural barriers, women in India predominantly first seek assistance from family members, particularly their own family rather than their husband’s, and rarely seek help from external justice systems (INCLEN 2000). Rao and colleagues found this true throughout South Asia, where women most often seek support through informal networks such as friends, family, or community members prior to 4 The practice where the bride’s family gives money or valuables to the groom’s family as part of the marriage agreement, which widely practiced despite being outlawed. 26 accessing formal support (Rao et. al, 2000). Indeed, worldwide many women are unable to access or navigate through the formal justice system, resulting in four out of five cases in developing countries being solved by informal courts (IDLO 2012). The United Nations’ (2012) report on informal justice systems and access to justice and human rights identified five different types of informal justice systems: traditional leaders, religious leaders, local administrators with an adjudicative or mediation function, customary or community courts where the adjudicator is not a lawyer, and community mediators. The report also defined informal justice systems broadly as, “Encompassing the resolution of disputes and the regulation of conduct by adjudication or the assistance of a neutral third party that is not a part of the judiciary as established by law and/or whose substantive, procedural or structural foundation is not primarily based on statutory law.” (UN, 2012, p.8) Compared to the formal justice system, there are several attractive characteristics that lend to women accessing informal justice systems. These characteristics include: familiar procedures and language; limited costs of settlement procedures; short duration of case resolution; knowledge of the local context; and a more restorative process (Wojkowska 2011). Despite these attractive qualities, these systems are likewise embedded within the male dominated power structures that perpetuate the very norms that lead to violence against women in the first place. As such, the systems are prone to corruption, with leaders rarely held accountable to their communities or any higher authority (IDLO 2012). It is within this contradiction that the development community has recognized the potential to harness informal justice systems as vehicles to promote community harmony 27 and peace within the family, while upholding women’s right to be free from violence. Although these informal systems inevitably demand some degree of compromise to maintain credibility within the community, the sheer magnitude of women accessing informal justice systems warrants the attention of the international community to consider, how can these systems become more responsive women’s needs and serve as a platform to challenge discriminatory cultural norms? Community Led Informal Justice Responses While there is growing attention regarding the prevalence and risk factors associated with violence against women, as well as the inadequacies of responses from formal institutions to this violence, there is sparse literature addressing effective ways of fostering women’s rights within informal justice systems. Informal justice systems that originate from civil society organizations to use a human rights based approach with community mediators are the most rare type of informal dispute resolution system (UN 2012). Within India, there are currently three community-led responses to domestic violence of this type that have been documented. In 1999, the International Center for Research on Women (ICRW) used a participatory research approach to document three community-led responses: the shalishi in West Bengal, the nari adalat/mahila panch in Gujarat, and the nari adalat/sahara sangh in Uttar Pradesh. The study had three main objectives, to document the response, assess the impact of the response, and build the capacity of the organizations implementing the response. The third objective aims to strengthen the organizations supporting these responses so that they can produce their own research, process documentation and evaluation in the future. 28 These three examples of community-led responses to domestic violence are all supported by broader women’s empowerment efforts in the region. The shalishi, a village-level dispute resolution system, is a resilient and adaptive informal justice system that has been a feature of rural communities in West Bengal for centuries. The Samity shalishi discussed here, which I will refer to as the “shalishi” only, is an example of how a traditionally male forum has been transformed into a space for women. Two Indian NGOs, Jana Sanghati Kendra (JSK) and the Sharmajibee Mahila Samity (SMS), support this particular shalishi. Both NGOs are peasant activist organizations highly involved in building a mass movement in the region, and these activists, both men and women run the shalishi. The nari adalat/mahila panch and the nari adalat/sahara sangh both grew out of the Mahila Samakhya (MS), an autonomous registered society sponsored by India’s Department of Education. Women’s groups form the core of the MS program, organizing women at the village level, block level and district level. Primarily serving the interests of rural women, leaders of the MS women’s groups involved at the district level recognized the lack of access to justice and created the concept of a nari adalat, “women’s court”, to be run by women to address gender based injustices. Considering the critical need to create viable informal justice responses to violence in the rural Indian context, it is important to note a few key findings regarding these three case studies. First of all, Bhatla and Rajan (2003) defined community-led responses as having the following key elements (p. 1659): 1. It is situated within the community setting and implemented by members of the community 29 2. It derives its authority and acceptance from the community, as opposed to from a more formal codified system of law 3. The community owns the process and decision; it is their sanction and responsibility that validates and enforces the decision 4. The process aims to shape and change existing community norms. In addition to these characteristics, the three responses documented all involved women and women’s collectives in the initiation, conceptualization and implementation of the response. I argue that the WRCs also have these same four elements, making them a community-led response to compare with the three responses in this study. Second, the authors summarize main strategies of the justice processes that the three community-led responses share in common. Recognizing that the community serves as an important emotional reference point for the individual, the responses use open, participatory, democratic processes to resolve cases. As a result, this degree of involvement of the community during the process itself leads to the community taking responsibility for the resolution that is decided upon. Inevitably, issues of gender roles and gender stereotypes emerge throughout this process creating valuable opportunities to raise community awareness and engage in discussion on spousal relationships and the importance of mutual respect and trust. The shalishi, for example, resolves cases in central and accessible spaces where passersby can come and watch and can even choose to participate in the proceedings. The Nari Adalat/Mahila Panch also see cases in formal places near government offices, which are accessible as well as add to the legitimacy of the process. This strategy of public involvement makes for obvious social punishment for the perpetrators in the case. 30 In order to create such open, public processes, this must be balanced with the goal of maintaining a women-centered approach. The facilitators, or mediators, of the process must ensure that the woman involved with the case is supported and feels safe. To do this, the nari adalats in Gujarat and Uttar Pradesh have a large presence of women who are members of the local women’s collectives during the process. Throughout the process women may speak up as they see fit to support or advocate on behalf of the woman. In addition, all three responses open the process by hearing from the woman first, following with perspectives from the others involved in the case. This validates the woman’s choice to speak up and take the case public. Perhaps the most important element of the strategy to resolve conflict shared across the three responses is a consistent position that violence is never justified or acceptable. Although these responses are built on principles of neutrality, and each emphasizes listening to all voices involved directly or indirectly with the case, the mediators always defend a position of nonviolence and respect for women’s rights. At times this must be done subtly, from within the perspective of culturally prevalent beliefs and customs so as not to alienate the community from the process, which would threaten the process of creating agreements based on consensus to be upheld by the community. Third, Bhatla and Rajan (2003) illustrate the critical role of women who arbitrate or facilitate these community-led responses. The women who lead these responses require a high level of diplomacy and facilitation skills to handle these cases in a way that results in solutions that are both pragmatic and culturally appropriate as well as pushing the boundaries for increased women’s rights. A key factor that makes this balance possible is that the women are from the local socio-cultural context, and so they have an, “intuitive 31 cultural sense of beliefs, value and normative codes of that area” (p. 1663). Considering the impact of being in this position, the women leading these responses reported an increase in social and familial status, with a redefining of gender roles in the household to adjust for women doing this work (Bhatla and Rajan 2003). However, the report also noted that there were, “continual conflicts and upheavals in the family as the women negotiate their roles and identities within families and communities” (Bhatla and Rajan, 2003, p. 1664). Fourth, the ICRW study aimed to identify potential indicators for these community- led responses to use to capture the impact of their work. The organizations supporting these interventions, JKS and MS, work to address broader development issues, and did not isolate impact indicators for these programs. The following indicators were identified based on interviews and surveys with the shalishi response: growth of the intervention (number of cases); change in the woman’s condition (better off, definitely worse off, mixed); change in self-confidence; change in the ability to seek help (time from when incident occurred); change in the involvement in the movement against violence (Talway and Samiy, 1999). One important finding was that over 57 percent of women reported recurrence of the problem and 46.4 percent reported new problems had emerged after their case was handled by the shalishi (Talwar and Samity, 1999). This illustrates the complexity of resolving family conflict and the necessity to follow up with cases well after the response considers the case “resolved.” Finally, Bhatla and Rajan (2003) argue that these responses serve both as an intervention to violence, restoring women’s safety in the home, as well as prevention of future violence by opposing the norms that attribute to violence against women. When 32 the community members present for the decision process conclude that violence is unacceptable, this threatens long established discriminatory norms that have contributed to women’s tolerance of violence. The existence and acceptance of the community-led response itself serves as the most obvious norm that is being challenged, that of violence being a private concern. Furthermore, that women are leading in public forums dispensing on matters of justice marks a fundamental reshaping of norms regarding women’s role in the public arena. This brief review of the ICRW report highlights the characteristics and strategies used by these community-led responses to address domestic violence and the potential for these responses to reinstate women’s rights in the family as well as prevent future violence through challenging community norms. In an effort to add to this small body of literature related to the effectiveness of community-led responses, this thesis focuses on the Women Resource Centers (WRCs) in Rajasthan, supported by the NGO, Seva Mandir. Monitoring and Evaluating Violence Against Women Interventions The complexity of factors that contribute to women experiencing domestic violence can only be addressed with a complexity of responses aimed to both empower women collectively and provide women with support, services and justice so she can realize her rights. The issue of violence against women, including domestic violence, has been recognized as a critical issue within the international arena, with key development actors such as the United States Agency for International Development (USAID), the World Health Organization (WHO), and others dedicating funds towards preventing and responding to violence against women. While this is clearly success in terms of visibility and commitment through to the top-level policy arena, practitioners struggle to support 33 recommendations for best practices due to a dearth of rigorous evaluations and a lack of data to support programmatic efforts. Likewise, many programs lack monitoring and evaluation systems that lay the groundwork for evaluating outcomes in the intermediate and the long term (Bloom, 2008, p. 9). The challenge in obtaining reliable data from such programs aiming to prevent and respond to violence against women stems from the special ethical considerations required to collect such data. Regardless of consensus on the indicators, any indicator that involves women self-reporting their experience with physical or sexual violence raises ethical considerations. It is expected that such population-based follow ethical guidelines to provide a safe, sensitive setting for interviews; however, this along with other precautions does not compensate for the fact that women will under report incidents of violence for various reasons. Under reporting can occur when violence perpetrated by intimate partners or family members is perceived as normal, when a woman fears retaliation upon disclosure, or when a woman would rather avoid disclosure fearing the social stigma to either herself or her family (Bloom, 2008). Along these lines, researchers or staff should not be collecting data without establishing what actions should be taken or not taken when women reveal that they are experiencing violence, or have in the past. Thus, confidentiality is of great concern to ensure women’s identities and the information they share is protected. The WHO (2001) has developed documents with safety and confidentiality recommendations for research about violence against women, highlighting the importance of specialized training for research team members and the ethical obligation 34 of researchers and donors to help ensure that findings are properly interpreted and used to advance policy and intervention development. This is not to deter research, in fact the WHO states that studies can be conducted with full respect of ethical and safety considerations, and that when interviewed in a non-judgmental manner in an appropriate setting, many women will not only discuss their experiences with violence but find their participation beneficial (Center for Health and Gender Equity in WHO, 2007). A monitoring and evaluation (M&E) plan encompasses various tools to provide information to program managers about how the program is progressing towards the intended results. A good plan has various components in place to collect, report and analyze data in order to assess and improve implementation as well as demonstrate effectiveness. It is worth briefly defining the key components that make up an M&E plan. Monitoring refers to managers routinely tracking how a program is running on a daily, weekly, monthly or quarterly basis. Consistent monitoring shows how program inputs lead to expected program outputs. Changes in outputs over time can be considered one way of alerting managers to potential problems. Evaluation includes numerous types of evaluation that serve numerous purposes depending on when they are used within the program cycle. In general, program evaluation is a tool to determine effectiveness of a program with regard to stated objectives and desired outcomes. Data for evaluations comes from predetermined indicators, occasional surveys as well as special studies. When program evaluations are thoughtful, well designed and appropriately analyzed they can provide valuable insights into how programs are operating, the extent that they are serving the intended beneficiaries as well as reveal strengths and weaknesses of the intervention. Depending 35 on the timing, evaluations can set priorities, guide allocation of resources and identify areas where modifications can be made to better suit the needs of participations or respond to new circumstances. Despite these promising aspects, evaluations are costly and require intensive planning and time to execute. Due to the worldwide recession that began at the end of 2008, along with conservative regimes restricting access of civil society organizations to bilateral aid, the resources available for women’s rights work have dwindled (Batliwala, 2011). Since this time, the mainstream development discourse has been marked by efforts utilizing an efficiency approach to women’s development, whereby investing in women is advocated on behalf of the promised economic returns on this investment. This approach, Zuckerman (2007) argues, does not value women’s worth apart from their value as economic agents. The World Bank’s “smart economics” or the “girl effect” models are examples of this instrumentalist approach as compared to a human rights or social justice approach. It is important to note that these approaches do not need to be mutually exclusive, and that an economic argument for women’s empowerment can be a positive strategy as long as women’s human rights are not negatively impacted (Lauterbach and Zuckerman, 2013). The implications of increasing competition for funds along with women’s empowerment, gender equality initiatives that emphasize economic outcomes, and hence more quantitative indicators, is that women’s rights efforts are under more pressure to demonstrate results. The measuring of outputs, outcomes, and impacts has become the norm for development interventions, which has lead to increasingly complex measurement tools. Consequently, program evaluation is typically performed by staff or consultants far 36 removed from the project. Similarly, monitoring falls into the hands of staff with a necessary level of literacy. For gender equality programs that use a participatory approach, with local women leading at the grassroots level as implementers of the program, this creates challenges for documentation and monitoring. Parpart, Rai and Staudt (2003) argues that participatory approaches often underestimate the skills needed for full participation in all phases of development, particularly the skills needed for writing reports and assisting in program evaluation. Within the context of initiatives addressing violence against women, measurement tools that emphasize outputs produce indicators that can be manipulated and/or misleading due to simplified notions of attribution (Batliwala, 2006). This is not to say that numbers should not be counted, but that these numbers must be contextualized within the increasingly complex environments that contribute to, or could be attributed to, the outputs. This requires qualitative data to accompany quantitative results, which is more time intensive to collect and record, especially if there are limitations regarding literacy. When harnessed as a tool for organizational learning and downward accountability, to the participants and beneficiaries of a program, evaluation can provide valuable insights. These insights serve the actual organization but also serve as the basis for critical dialogue at regional, national and international levels regarding effective intervention strategies. Unfortunately, all too often NGOs are burdened by program planning and evaluation requirements that emphasize quantitative outputs and tools that require a particular type of critical thinking and literacy likely to be found only among top-level staff. Stretched for time and resources, NGOs using these quantitative 37 monitoring and evaluation tools are not capturing the key qualitative shifts or unintended consequences of the program. This lack of documentation and monitoring is a missed opportunity, which limits the effectiveness at the organizational level, but also for the larger community of practitioners working to achieve similar objectives. The challenges of monitoring and evaluating community-led responses to domestic violence have lead to a lack of data regarding the effectiveness of these responses. My research aims to fill this gap, by working to document Seva Mandir’s community-led initiative with an eye to identifying potential strategies for determining the program’s impact in the future. 38 CHAPTER III SITUATING AND CREATING SEVA MANDIR’S PROGRAM THEORY In this chapter, I describe the process of using the theory of change approach, specifically an outcome map, to define Seva Mandir’s social empowerment program theory as it relates to the issue of domestic violence. Prior to this, I situate the program theory by describing the issue of domestic violence within Seva Mandir’s work area. Likewise, I provide a brief review of the origins of the WRCs in Kherwara block, my research site, then specifically look at how staff define and describe domestic violence in this area. Domestic Violence and Tribal Women in Udaipur District Tribal women in Udaipur district in the state of Rajasthan, India face several oppressive practices such as: domestic violence, being branded witches, desertion by their husbands, and being scapegoats when conflict arises across homes and communities (Seva Mandir, 2011). India’s last National Family Health Survey (2007) found that 46 percent of married women in Rajasthan responded that they have experienced some form of domestic violence. When women suffer these injustices they often have few options for help. The state institutions, such as the police, and traditional institution, the caste panchayat, responsible for dispensing justice have a poor reputation due to a lack of transparency, high fees and penalties, and frequent acts of corruption. More importantly, the caste panchayats do not allow women to be members nor can women directly put 39 their case forward to the cast panchayat members. As a result, tribal women face numerous barriers that prevent them from accessing justice that will strengthen their right to be free from violence. Research Locations: Kherwara Block, Balicha and Kharadiwara WRCs After conversations with staff in the WCD office in Udaipur, we decided that I should focus my research on Kherwara block, since there are three WRCs in this block: Balicha (since 2006), Kharadiwara (since 2006), and Sagwara (since 2011). As mentioned previously, from my initial site visits I decided not to include Sagwara because it has only recently been established and proved difficult to reach logistically. Kherwara Block is predominantly rural, with one urban settlement, covering 110,211 hectares in the Aravalli Mountain Range. Sharing borders with Gujarat state and Dungarpar district to the south, it has a population of 268,000, three quarters of which are tribal, predominantly the Bhil, Meenas and Garsaias tribes (Seva Mandir 2012). People in Kherwara Block depend primarily upon agriculture and animal husbandry for their livelihoods. However, due to small land holdings and poor irrigation infrastructure, a large part of the population migrates to Gujarat or Kuwait to find work as laborers in the non-agriculture season. Because the majority of my data comes from Balicha WRC, I will overview the origin of this WRC, which is similar to how the Kharadiwara WRC also began. In addition, I will note the block staff’s definition of domestic violence and staff impressions of the causes of violence in the Kherwara block. The WRC in Balicha began after domestic violence issues emerged within a women’s cluster level group. This cluster level group, including leaders from seven or eight village level women’s Self-Help Groups (SHGs), decided to elect three women 40 leaders from each SHG group to meet every two months to discuss these issues. These discussions that began at the cluster level raised questions as to how the women’s groups can address the injustices women are facing, considering the lack of viable options to turn to for help. As a result, Seva Mandir facilitated a group of women leaders to go on a site visit to the nari adalat in Gujarat, around 2003. Following the visit to the nari adalats in Gujarat, the cluster level women’s group in Balicha organized an event where groups of women could meet with the resource persons from the nari adalats and strategize how to build a response in their communities. Here it was decided to elect three women to open a space for women to resolve cases and address these issues. In 2006, the Balicha WRC opened. The village of Balicha was selected to be the site of the WRC due to the highly active women’s group there as well as supportive men in the community. In addition, the previous event with discussions with the representatives from the Gujarat nari adalat was held in Balicha. In talking with staff about the origins of Balicha WRC, staff described the issues that were emerging that demanded a new type of response, “Issues were coming such as home violence, physical torture, beating, husband demanding the wife to leave the home. At that time daakan (witch accusation) was a big issue. At the panchayat level, issues were coming such as widow pension, old age pension, and safety of women, rape cases, eve teasing5”. Staff explained that prior to the WRC opening some of these issues were resolved within the village level women’s groups. 5 A euphemism used in India for public sexual harassment or molestation of women by men. It includes acts varying from catcalls, sexually suggestive remarks, to groping. 41 In 2006, when the Balicha WRC first opened, Seva Mandir had rented a space for the center located in the school campus. This was a room owned by the panchayat6, the local government. At the time, there was also a Youth Resource Center (YRC) that shared the space, and this facilitated cooperation between the WRC and the YRC. An YRC leader explained, “In the beginning we both worked there together, the cases were bought before the WRC women…I [still] attend meetings related to the cases that are brought up before the WRC.” Only one year after opening, the leader of the local panchayat asked the WRC and YRC to vacate the room, suggesting it was a disruption to the school campus. From this point on, going on six years, there has been a search for a new space. Meanwhile, the center has operated from one of the homes of a cluster level worker. While conducting interviews in Balicha I found people used the terms “home violence” and “domestic violence” interchangeably. In general, domestic violence was used more than home violence, especially among staff; hence, I will use domestic violence throughout the thesis. According to Seva Mandir staff in Kherwara block, “domestic violence in this local context refers to physical violence, beating, any scratch on the body by a man on a woman. It also included psychological violence. Home violence is when some family member does violence, which could be physical, mental or sexual abuse.” When asked why violence happens specifically within tribal communities, staff explained issues related to son preference, alcohol, women’s work in the home, and general male control over their wives in terms of mobility and interaction in the public 6 Here panchayat, as opposed to the caste panchayat, refers to the official local level governing system comprising of elected officials. 42 sphere. “If a woman is infertile, unable to have children, or if she is giving a girl child, with no sons. Also, drinking wine. When a woman is sick and not doing enough work at home, then there is beating. There are many reasons for why violence happens.” Several staff members mentioned, “suspicion”, meaning the husband is suspicious that his wife is having a relationship with another man. Similarly, another staff member explained that, “Women do not have any rights to go out anywhere without taking permission from their men in India. When a woman doesn’t ask permission for leaving the house, even for working, then violence occurs.” Outcome Mapping One technique within the theory of change approach is the outcome map. Outcome maps are a visual diagram that depicts the relationships between the various strategies being used to address a problem and how these strategies work together to create the desired change. Both short term and long term change is captured within outcome maps as well as distinctions between changes at different levels such as, at the individual level, organizational level, system level or community level. Another tool typically used alongside the outcome map is a document that captures the assumptions that underlie the program. This includes organizational philosophy, values, ways of working together as well as the overall community context within which the intervention takes place. It is important to first define some of the terms commonly used within the outcome map, specifically three types of outcomes. First, “impact” is used to refer to actual changes in people’s lives. Individual impacts can be considered the foundation for scaling up strategies to the larger population. Without individual impacts, including changes in knowledge, skills, behavior, community-wide change is unlikely to occur. However, 43 programs must recognize that individual impacts alone are insufficient to maintain social change in the long term. Beyond the individual level, changes within institutions, service systems, community norms, the public will or policies are labeled as “influence”. Third, “leverage” is used to denote a change in resources available either through public or private funders to support the program. A final term useful in this version of theory of change thinking is “core capacities”. An organization’s core capacities are what make program strategies effective. These include key inputs that support implementation and program sustainability such as, how the group establishes a collective vision, makes decisions, uses data, activates a communication plan or supports collaborative learning. Well functioning core capacities are important to achieving impact, influence and leverage outcomes. For example, when looking to achieve and influence an outcome where community norms do not tolerate violence against women, core capacities such as a collective vision, engaged local leadership, and capacity to communicate core messages are a few core capacities that support achieving this impact. However, these core capacities can sometimes be difficult to define, particularly within complex grassroots organization efforts. Overall, the aim it to consider core capacities as the inputs or ingredients used to carry out influence, leverage and impact strategies. To create an outcome map for the WRCs alone, as stated previously, would be to misrepresent this facet of Seva Mandir’s efforts as isolated from other social empowerment initiatives. During my first few weeks at Seva Mandir I spoke with staff in Udaipur about how the program is intended to work as a piece of the WCD unit’s comprehensive social empowerment program. These conversations led me to reviewing 44 current documents and previous volunteer reports related to the WRCs. In looking at Seva Mandir’s Seventh Comprehensive Plan, 2012 – 2015, I found the WCD unit depicted its program in a logical framework analysis format (See appendix A). I used this document along with a 2011funding proposal for the WRCs to refer back to while in the field as the most explicit example of the WRC’s program theory. In addition, after leaving Udaipur I referred to these documents to use Seva Mandir’s language in identifying the strategies they are using to address violence against women. Because I was dealing with language barriers as well as limited time and resources, my fieldwork and data collection was a back and forth process between articulating the implied program theory and observing the program’s current operation through interviews and observation in the field. Since I knew I could create the implied program theory while not on site, I maintained documentation regarding this piece and focused my time and energy on the fieldwork aspect. In the field, I worked to capture the voices of the women leaders implementing the program as well as the women who had received help from the WRC. In creating the outcome map I followed a toolkit by the Annie E. Casey Foundation (2004), “Theory of Change: A Practical Tool for Action, Results and Learning.” There are numerous tool kits designed to walk organizations or individuals through this process, and I selected this one because I found it thoroughly described the elements of the outcome map with sufficient examples. The toolkit breaks down the creation of an outcome map into six steps, which I will briefly describe in the following pages. I believe that describing the process illustrates the type of thinking involved in creating an outcome map, which I argue would be helpful for Seva Mandir staff to do as 45 the implementing agency. While some practitioners contest that the theory of change thinking still focuses too heavily on causal thinking and over simplifies change processes, there are several benefits to using this approach. First of all, when done collaboratively, this process establishes a clear vision of what the program is ideally trying to achieve. Second, the process facilitates discussion and ideally consensus regarding fundamental assumptions underlying the program. Likewise, the process fosters accountability and transparency among staff and other stakeholders. Discussion should emerge regarding unintended outcomes that can also be used to capture learning. Finally, and most relevant to this thesis, the creation of an outcome map aids in designing monitoring and evaluation systems that are sensitive to the program’s unique and complex context. Step 1: Identify the Ultimate impact Looking at the WCD unit’s logical framework analysis, the ultimate impact is stated as, “improved position of women in household, community and society, politically, socially and economically” (Seva Mandir 2012). While this ultimate impact clearly represents the WCD unit’s vision I wanted to make the ultimate impact more specific to the issue of violence against women. My aim was to situate the WRCs within the unit’s comprehensive empowerment programming; therefore, I made the ultimate impact, “women have access to justice that supports their right to be free of violence.” Step 2: Identify strategies to address the ultimate impact Because the aim was for my empirical observations to converse with this theoretical model, I did not include strategies within the logical framework analysis that I did not observe in the field. This includes activities such as: strengthening savings and credit 46 activities; capacity building of group accountants; linking women members with external financial institutions; providing support for various Income Generation Activities (IGA). In addition to having no direct observation related to these activities, I also found through my conversations that these activities did not have a strong social empowerment element to them. As such, I wanted to focus on those activities that more directly correspond to social empowerment and therefore, the ultimate impact. With this in mind, I identified seven strategies that Seva Mandir is using to address this ultimate impact: (1) create and strengthen women’s groups at the village and cluster level; (2) create and strengthen WRCs; (3) strengthen the social empowerment aspect of Self-Help Groups; (4) coordinate awareness campaigns, public hearings and solidarity events; (5) build capacity of WRC leaders and Cluster Level Association (CLA) leaders; (6) provide specific gender sensitization trainings for caste panchayat, youth and women leaders; (7) support and strengthen the counseling cell and short stay home in Udaipur. Step 3: Create “So That” chains for each strategy So That chains are an important step in the creation of an outcome map, where each strategy is broken down into a sequence of parts that link and together form a pathway to the ultimate impact. In looking at each strategy individually and linking it to the ultimate impact, this generates reflection on how the program is intended to work. Recalling the three types of outcomes, impact, influence, and leverage, this step in the outcome mapping process will illuminate key assumptions implied within program components. The seven So That chains I created can be found in Appendix B. Step 4: Link strategies with outcomes and ultimate goal During this step all of the outcomes from each of the So That chains are linked to 47 the ultimate goal, which reveals how several strategies are intended to achieve similar intermediate outcomes in impact, influence or leverage. These overlaps are important to recognize the interdependence of efforts to address complex social issues. Likewise, these linkages reveal how this interdependence can create areas where progress might require relatively more time or increased resources to achieve an intermediate outcome that is key to the path towards the ultimate impact. Also, at this point in the process questions begin to arise as assumptions become more apparent. Step 5: Review completed map and make adjustments Typically this outcome mapping process would be done collaboratively, showing the completed map to stakeholders for feedback. It is recommended to test the logical linkages between strategies, outcomes and impacts. Additional input from a diversity of perspectives is likely to contribute to the accuracy of the outcome map. However, because my priority was in fieldwork and observation, and also because of language barriers, I did not have time to make this step possible. Since I completed the outcome map after leaving Udaipur, one way I was able to review my map was by revisiting the WCD proposal document for the WRCs. The final section of this proposal asked, “where do you see the innovation five years from now?” The list of outcomes here was particularly useful for me to double check that these outcomes were within my own outcome map. Step 6: Articulate the Assumptions The process of articulating the assumptions within the program theory, as depicted by the outcome map, begins with the So That chains. In looking at each of the seven strategies that lead to women having access to justice that supports their right to be 48 free from violence, various assumptions are implied in order to move through the chain towards the ultimate impact. These assumptions will be discussed throughout the following chapter as I compare my observations with the outcome map, which represents how Seva Mandir conceptualizes their programming is addressing domestic violence. 49 CHAPTER IV COMPARING THEORY WITH PRACTICE: STRENGTHS AND CHALLENGES Based on my seven weeks of observation and interviewing, mainly the women implementing the Balicha WRC, several assumptions were revealed indicating that the current operation of the WRCs is slightly different than how the program was originally conceptualized. This chapter will hit upon these assumptions by describing several characteristics or concerns related to the operation of the Balicha WRC, integrating quotes from interviews as well as observations from the field. The WRC Team Without a doubt, the most critical piece of Seva Mandir’s intervention to address domestic violence within its work area is the women leaders who implement the centers using conflict mediation. Seva Mandir has invested a tremendous amount of resources supporting the capacity building of these women leaders. Some women leaders have been working with Seva Mandir’s women’s groups for over fifteen years. Not surprisingly, the women leaders I met at Balicha and Kharadiwada are incredibly strong, confident women who are passionate about their work. In addition to the three women leaders, I observed three other key people are equally as involved in operating the WRC in Balicha, two cluster level workers and one man working with Seva Mandir. These six individuals form a solid WRC team committed to the cause of building the capacity of the WRC and establishing the center 50 as a permanent feature of the community. As one woman leader was describing a case, I asked her who went to the home to do the mediation. “We all go together,” she said (Interview). This “we” referred to is what I now call the “WRC team”, consisting of the three women leaders, two cluster level workers, and one male staff member. In Kharadiawada, I did not find the cluster level workers as involved as in Balicha. Furthermore, the WRC leaders themselves were not consistently dedicated to the center. Of the three women leaders in Kharadiwara, one had recently replaced a leader that was sick. She had only been with the WRC for roughly three months and had yet to resolve a case. Another woman leader was currently suffering from abuse within her own family, with Seva Mandir staff intervening to help her resolve with conflict and be free from abuse. While I did not know this at the time of my interview with her, staff later explained the situation to me. This case involved verbal abuse, which was coming from both the woman leader’s daughter in law and son, “torturing her” and telling her to leave the home. This staff member and the other two women leaders responded by reporting her case to the WRC themselves, and visited her home to intervene. When they visited the home, they found, “Her son left her home, fleeing somewhere, so no step can be taken right now against her son, but we are waiting for him and whenever he will return he will be punished. She [the woman leader] was crying very much as her son did this bad thing to her, abused her badly.” While interviewing this leader, that I later learned was experiencing abuse, I asked her if violence happens to women in her community, and she said, “Violence happens sometimes. Verbal abuse. But this is culture.” During this interview she gave 51 very brief responses and indicated that she only does this work because Seva Mandir chose her to do it. This situation, along with the new women leader with very little experience made for a WRC team with relatively low capacity compared to the solid six- member team in Balicha. While the presence of a strong team that works together to resolve cases is certainly an asset to the program in Balicha, the WRC’s activity is highly dependent upon the support of the cluster level workers. One cluster level worker explained, “The three leaders are illiterate, so they never go without the cluster level workers.” Indeed, throughout my time interviewing, the women leaders, separately, would tell me numerous stories about how they have resolved cases. In each case, they confirmed that both cluster level workers came to resolve the case with them, as well as the male staff member. As a strategy for conflict resolution, working as a team means that six people show up at a home in the village to talk to each person involved in the case. This is not something that will go unnoticed by neighbors. Indeed, the women leaders described how shocking it could be for people, inciting anger and humiliation. This strategy very much helps create pressure for the conflict to be resolved more quickly and prevents future conflict out of fear that this team will return to the home. Furthermore, working as a team serves to protect the women from aggressive or even potentially violent reactions to their arrival at a home. The commitment and involvement of the cluster level workers is not surprising; however, the committed male staff member was unexpected. This man’s position is a smaller role within the zone level staff, mainly assisting with documentation at the SHG (self-help group) meetings. But, he has personally become an advocate for the WRCs and 52 is now a fully functioning member of the team. He not only joins the women when mediating a registered case, but also reported that he has helped resolve cases with one of the cluster level workers that are not registered, voluntarily. Based on conversations with him, his strategy for resolving cases and perceptions of gender inequality and violence against women are similar to the women working for the WRC. Most importantly, he is very much well received by the women working for the WRC. They turn to him for support and I sense a strong relationship of trust and respect between them. In addition to his personal investment in the project, I found he was able to discuss the WRC as a program at a different level than the other WRC women comprising the WRC team. Perhaps due to different life experience as well as an obviously higher education level, he could more clearly articulate critical reflections about the WRC’s role in the community compared to the women leaders. During our conversations, he made several suggestions about the challenges the WRC faces in resolving cases, which I will incorporate throughout this discussion as part of my feedback from Seva Mandir staff. I integrate his feedback regarding the program as a staff member in order to provide greater confidentiality, even though I consider him as a “member” of the WRC team due to his consistent involvement in resolving cases. The discovery of a more diverse team supporting the WRC, as opposed to the implied theory where it is primarily the three women leaders resolving cases, highlights an assumption within the implied program theory that contrasts with the current operation. The added diversity and clear commitment of the cluster level workers builds in greater sustainability for the program and enhances the case resolution strategy in Balicha. However, my observations also showed that the team might be too dependent on 53 the cluster level workers. This became clear much later during fieldwork when a cluster level worker was sick and the follow up with a new case did not happen until she became healthy and initiated a response. The most obvious issue that is raised with regard to the dedication of these cluster level workers is salary. The cluster level workers are definitely expected to be key supporters for the WRCs, if there is one located in their village in particular. However, this level of involvement goes beyond what is expected, and certainly adds more work, more time away from the home, without direct compensation for this higher commitment to the program. Further discussion on salaries will be discussed later on in this chapter. Location The most obvious assumption that was immediately in contrast to the program theory was regarding the physical location of the WRCs. When I asked women what challenges the WRC was facing, the WRC team always spoke about location, first and foremost. “We need a new center badly. This is the main challenge the WRC is facing,” a woman leader stated. In both Balicha and Kharadiwara, the WRCs are located in the homes of women working for the WRC. During meeting days the women bring a carpet out and then gather on this carpet. When I asked one of the women whose home is used for the center how she feels about this, she stated, “Having the center here is ok, but it does sometimes creates a problem if I have guests on the days of the meetings.” From the very first day I started talking to the women, this is the issue that came up again and again with every single person I interviewed. There was a real sense of hope that there might be a way to have their own space to talk to women, put up flyers, posters, etc. Some spoke excitedly that the community could provide land, if Seva Mandir can 54 help with a building. “Land will be provided by us, but materials should be funded by the agency, according to me this is the solution which if the WRC has its own center, it could run more efficiently. If there is a separate sitting area or center for the WRC, the women working will feel happy and satisfied about their working.” When I talked to block staff about this option, it is quickly considered not possible. One zone staff member suggested the WRC should be run like an office, open everyday with information about government programs. Another suggested the center could be used as a location for case resolution and also as a place for Seva Mandir volunteers to stay when doing their volunteer projects. Undoubtedly, the WRC team in Balicha is eager to have its own center, as well as certain Seva Mandir staff. That the program was designed at the outset to include renting a physical space for the centers, and that this is not occurring, illustrates a divergence from the program theory that can be explored to identify if and how the current situation can move back towards the intention of creating an actual physical space in the community for the WRC. Fees “…It’s helpful for both parties because it is free.” This quote is from a male community member, who has a family member who works with Seva Mandir. It indicates another assumption, that the WRC has incorporated a fee structure into its case resolution services, to build in sustainability of the program and lessen reliance on Seva Mandir in the long term. However, I found the concept of fees was not clearly understood within the WRC team itself, and several interviewees indicated that they believed the WRC provides free services. A cluster level worker from 55 Kharadiwara said, “[the WRC] has made an impact. Before women had nowhere to go. Also, it is free, not expensive.” A cluster level worker in Balicha explained, “In the beginning we did everything for free, no receipt or payment even for registering. But now Seva Mandir is saying to charge after resolution. People don't have the habit of paying.” This quote illustrates the main source of confusion, in that the WRC was free to begin with, but then Seva Mandir introduced a fee structure. One Balicha WRC team member said, “Many times parties are offering us money, but we refuse because we say that Seva Mandir pays us. But Seva Mandir only pays us for 3 days work.” This quote shows that sometimes the WRC team feels uncomfortable collecting fees from the women themselves, instead believing that Seva Mandir should be providing the financial support for the center. The fee structure imposed by Seva Mandir raises questions regarding the appropriateness of fees in this context, where the barriers that women face in coming forward to speak up about conflict, especially violence, are already significant. Does the fee actually inhibit women from approaching the WRC? In my conversations about fees, some community members felt the fees could be raised, and people would pay. However, they were speaking from the perspective of people using the WRC as an alternative to the caste panchayat, which involves expensive fees. It was also clear that the WRC is perceived as a service for poorer people in the community, which could be used to address land disputes, for example at the fraction of the cost it would be to use the caste panchayat. However, the women do not have the training to resolve these matters anyway, and are trained to be addressing family conflict causing violence against women. 56 Remembering its purpose, to serve women experiencing violence, the fees do seem problematic. During my first meeting with all the women leaders at Balicha I asked them what fees they charge, and was told 11 for registration and 25 for resolution. At this time they did not mention issues related to people not paying the fee. Looking at the case receipts, the first receipt for over 11 rupees is from 2009, for 51 rupees. As of the end of October 2012, there are 10 receipts (out of 46 total) for case resolution: 21 rupees (4 receipts), 25 (3 receipts), 50, 51 and 100 rupees. The women leaders themselves paid four of these 10 fees over 11 rupees, when they used the WRC to resolve their own individual conflicts. This breakdown of the fees paid shows that the 11-rupee fee for registration is more or less consistently paid, but the fee for resolution is highly erratic. Since all but one case is considered resolved, this indicates that only 22 percent of resolved cases paid a resolution fee. Granted, three out of those ten resolution fees were well over the recommended 25 rupees, at 50, 51 and 100. This is highly encouraging to see that the WRC was so appreciated in these cases. Another issue that arises in thinking about the matter of fees is that of documentation. If the documentation only occurs when a case pays the registration fee, all other matters that the WRC responds to remain undocumented. As a result, the number of cases reported really represents the number of registered cases. It appears that the women do respond in some way to other “cases” that are not registered. Indeed, registering a case certainly requires a much higher level of certainty in calling upon the WRC for help, as well as a fee that some women would not be able to pay. 57 Stipend and Bank Account At my final meeting with the WRC team in Balicha, the women appeared confused and frustrated when I was asking them about how the bank account works. I had heard from staff that the women leaders have access to their own bank account, where they manage an annual budget that covers their salaries. One WRC team member stated, “X is supposed to help us [with the bank account] but they are busy. Y told us to talk to Z but they are busy.” At this meeting the conversation became quite heated with several people talking at once, describing what was happening and not happening with regard to the bank account. One woman even said, “I don’t want to work anymore, no one is supporting us.” The women repeatedly mentioned they were unsatisfied with the stipend considering the hours of work they do. One woman compared the rate to what a government scheme pays for day labor, saying that they can get more for doing manual labor. In addition, the several people mentioned that the stipend is a barrier to recruiting new women leaders. A staff member suggested, “These five ladies who are working right now at the WRC were requested to be here. I don’t think new ladies will prefer to join WRC if their needs are not met.” Overall, there was a clear sense that the women leaders felt they were not adequately being compensated for their work at the WRC. The designated workdays, on the first, tenth and twentieth of each month, are the only days the women leaders receive a stipend of 100 rupees per day. Additional time they spend outside of these three days, in theory, can be paid through an annual budget that allows for travel or phone expenses. However, in Balicha the women were struggling to access this bank account and manage it to cover these additional costs. This gap 58 between the theoretical program model and current operation indicates discussion about routine operation of the WRCs, including communication between the zone staff and the WRC leaders in terms of the bank account. New WRC Leaders “New women are not interested in this work.” Numerous times, when I asked how the WRC could be improved, people mentioned that there should be more women involved. However, the women leaders and staff mentioned several barriers to recruiting and training new women to do the WRC work. We often discussed the issue of age and literacy, due to the current leaders being older and illiterate, and the barriers this created in term of documentation. “Young men don't want to send their women here to resolve cases or be involved in the process.” “Younger women don’t have time to get involved. But older women don’t have capacity to walk long distances.” This issue is further complicated by some conflicting perceptions between the WRC team and Seva Mandir staff. While the WRC team has a strong position that it is difficult to find women to do this work, some staff have a different view, “Women don’t want to give up their leadership, they don’t want to give up this role.” This discrepancy indicates a need for greater understanding between the WRC team and staff in order to best address the critical issue of bringing in fresh leadership to the WRCs. With some exceptions, where leaders were replaced due to illness or another reason they were unable to do the work, the same women leaders have been in place since 2006, across all of Seva Mandir’s nine WRCs in operation as of November 2012. Furthermore, some of these transitions I encountered appear to have occurred without an election. “I came to work 59 here after another woman left. I replaced her…the cluster level worker brought me” (Interview). I found another woman leader who reported the same case, where she replaced a leader by appointment you could say, without an election. “I’m part of an SHG group, and the block staff talked to me about it [becoming a new leader to replace a woman who became to ill to continue].” Clearly, staff is aware of the challenges of bringing in new leaders at the WRCs, as well as the importance of this transition to ensure the sustainability of the project. However, my interviews show that the implied theory of WRC leaders being elected is not consistently applied. This alone shows a need to re-clarify how women come to be leaders for the WRCs, which is an important participatory feature that illustrates the connection between the women’s groups and the WRC. If transitions in leadership are happening by appointment rather than by election, this shows missed opportunity to garner more clear connections between the women’s groups and the WRCs. Cases – What Prevents Women From Coming to the WRC? According to staff impressions, the number of cases coming to the WRCs over the past couple years has decreased. This is something that the WCD unit continues to question, and is one of my primary research questions that I returned to whenever I saw an opportunity in my interviewing. A few responses are as follows, “They are worried about their image, reputation in the community.” “Yes, we have said [told women] but they don't come. Every one knows but people don't use the center.” “Not many people are coming unless they are connected with Seva Mandir.” One of the assumptions in Seva Mandir’s implementation of the WRCs is that women in SHG groups can spread awareness, and this will bring more cases. Seva 60 Mandir is also doing gender sensitization trainings with women and youth leaders, as well as the male leaders of the caste panchayat to build an environment where women are more able to speak up. However, responses indicate that the constraints to speaking up are still too strong, preventing many women from speaking up and using the WRC to resolve conflict. “They feel that if violence is happening in home, they want to resolve it in home. They don't want to break the family. They think that if they talk about it they will break the family, and then she will suffer from a bad reputation.” Considering the threat of reputation, the current implied program model requires the woman to physically come to the WRC to register a case. In addition to the availability and cost of transportation, the woman must also come on one of the three days that the center is open. Otherwise, the only other theoretical way for a case to be brought to the attention of the WRC is through the women’s groups. My experience observing a group meeting where a woman spoke up about violence will be discussed later in this chapter. One way of gaining more insight into this topic is by asking, who is coming to the center? “The WRC is serving poor people, because they don't have money. Rich people don't use the WRC because they have money so they go to the court and the police station.” According to the WRC team, some cases have come from quite far away from the WRC (10 kilometers), and from families that are not associated with Seva Mandir. However, according to the existing documentation and discussions with the women, the majority of cases do come from people who are involved with Seva Mandir or live near one of the WRC team members. Staff and two men in the community mentioned that the 61 WRC is only serving poor people. “People with money would not come to the WRC,” they said. In theory, with proper documentation, there should be more data regarding who comes to the WRC. Likewise, the women leaders themselves or other WRC team members have accurate memories shortly after cases to recall the details in an interview. Further dialogue with women regarding the barriers in accessing the center will probably not reveal new information. However, further understanding with regard to who comes to the WRC and when they come would be informative. This could be captured through effective monitoring and documentation as well as through interviews with the women resolving cases. Tolerance of Violence “In this village, 90% of families have conflict between husband and wife at home. In this village a lot of conflict happens.” “If it's 1-2 times it's ok, but more then that it's not ok. If they feel afraid, then they will inform the WRC.” Speaking with the WRC team or women members of the SHG group in Balicha, as well as staff, everyone indicated that violence happens on a regular basis in homes. As expected, people said that women prefer to resolve this conflict in the home, often tolerating verbal or physical abuse for some amount of time. “Women are avoiding it. They think this is happening, they have a small family, so they avoid it. They want to live peacefully, so they avoid.” “When there is continuous conflict or physical violence, then they will come [to the WRC].” “If it is verbal abuse, then people may not use the WRC. This is small conflict. If one or two slaps happens…if any more happens then we use. If very big conflict then I would go to the police station.” 62 The data from my interviews confirms the literature regarding tolerance of verbal or physical violence (one, two slaps). That the WRC is seen as an option for more extreme cases of violence, particularly physical violence, is encouraging. Again, with effective documentation describing the types of violence women are experiencing in the cases that are registered at the WRC, Seva Mandir would have greater understanding about help seeking behavior and tolerance of violence to share within the field of practitioners addressing domestic violence. Documentation I have already mentioned challenges and opportunities regarding documentation, but here I will specifically analyze the quantitative data that was available through the receipts at both Balicha WRC and Kharadiwara WRC. Balicha “When he has time he comes and documents, but not regularly.” I never met the man who is doing the documentation in Balicha, but did a thorough review of the receipt book and photocopied samples from the case register notebook and a written agreement. The photocopies were at the request of Seva Mandir staff so that they could better understand the system as it is currently operating. The register notebook contained a mixture of meeting minutes and case details, which made it difficult to extrapolate information about cases. See Table 2 on p. 64 for a breakdown of cases according to the receipt book. When asked if the number of cases coming to the center is less compared to before, both staff and women working for the WRC in Balicha replied no, the number of 63 cases has not decreased. Based on receipts, the number of cases has remained more or less the same from 2007 – 2010, but has in fact decreased slightly since 2011 (see Figure 2 below). The receipt book includes receipts for fees paid when a case is registered as well at resolution. There are at least five receipts from resolution, plus an additional four receipts that are over 11 rupees (the registration fee), but not marked as registration or resolution. Therefore, a more accurate number of cases would be between 37 and 41 cases total. That is an average of between 6.1 to 6.8 cases per year. Based on receipts, 22 cases or nearly 50 percent (22/46), were conflict between husband and wife or other family members. From conversations with the women leaders, most cases involved verbal abuse, with a few cases of physical violence. The women also frequently spoke about daakan cases (witch accusation), although there are only five receipts, 11 percent, identified as daakan cases. See Table 1 on the following page for a breakdown by year of receipts, including amount paid and type of case. Fig. 2: Balicha: Number of Receipts 2007 - Oct 2012 "!#! $!%! &!'"! '#! #""(! #""&! #"")! #"'"! #"''! #"'#! ! "# $% &' () '* +, %, ' -%+&' .+/012+'34*'' !"#$%&'()'*+,%,'5667'8'91:'56;5' 64 Table 1. Balicha Receipts 2007 – Oct 2012 Year Number of Receipts Number of Receipts from Resolution or Issues Related to Documentation Type of Case 2012 *as of Oct 5 1 – 50 Rs., resolved 1 – no amount recorded 4 –Not recorded as resolution but 25, 25, 25, 21 Rs. 4 – recorded on May 20th 2 – conflict husband and wife 1 – conflict husband and daughter 4 – cases from women leaders 2011 6 1 – 100 Rs., resolved 2 – 21 Rs., resolved 1 – no year recorded 1 – no amount recorded 1 – daakan 2 – conflict husband and wife 1 – conflict two women 2010 9 1 – 21 Rs., resolved 7 – no amount recorded 1 – no year recorded 2 – conflict (no detail) 4 – recorded August 1st 1 – accident with husband 1 – conflict over employment 1 – son’s wife left with 3 children (leader’s case) 1 – daakan 1 – conflict over a girl 1 – conflict husband and wife 2009 8 1 – 51 Rs., resolved 1 – no year recorded 2 – daakan 4 – conflict husband and wife 1 – daughter ran away with man/daakan 2008 8 3 – conflict husband and wife 1 – conflict two women 1 – conflict son and wife (his mother registered) 1- conflict neighbors 1- domestic violence 1- daakan 2007 10 3- no case description 1- conflict between each other (who?) 2- conflict husband and wife 1 – domestic violence, husband’s younger brother beating elder brother’s wife 1 – conflict mother and son 1- husband married another woman 1 – mother not sending daughter to husband’s house Total 46 Resolution Fees = 5 to 9 2 – domestic violence 5 – daakan 8 – conflict other family members 14 – conflict husband and wife 3 – no case description 65 Kharadiwara I also reviewed documentation in Kharadiwara. I looked at the receipt book, case register and took photocopies for staff based in the main office, in Udaipur. For Kharadiwara, as like Balicha, the case register was difficult to read, with a mixture of meeting minutes and details about cases. A more consistent format would make it easier to follow. When asked if the number of cases coming to the center is less compared to before, both staff and women working for the WRC in Kharadiwara agreed that the number of cases has decreased. Based on the receipt book, the Kharadiwara WRC had 14 cases in six years, an average of 2.3 cases per year (see Table 2 on the following page). The number of cases per year appears to have been consistent; however, the women also told me they have resolved many more cases but do not have receipts. Unlike Balicha, the majority of case receipts (57 percent) noted physical violence or home violence. 66 Table 2. Kharadiwara Receipts 2007 - Oct 2012 Year Number of Receipts Number of Receipts from Resolution or Issues Related to Documentation Type of Case 2012 *as of Oct 2 1 – mother with physical violence 1 – mother with verbal abuse (leader’s case) 2011 3 1 – daakan, verbal abuse and physical violence 1 – conflict neighbor, verbal abuse 1 – conflict over property 2010 4 1 – no case description 1 – physical violence; woman beaten by mother in law 1 – physical violence; sister in law 1 – conflict between husband and wife, verbal abuse 2009 4 1 – no case description 1 – wedding? (unclear) 1 – home violence 1 – physical violence; husband beating wife with 2 daughters 2008 2 1 – underage girl (unclear) 1 – verbal abuse, physical violence between husband and wife 2007 3 2 – no case description 1 – home violence Total 17 Resolution Fees = 0 8 – physical violence or home violence 4 – no case description ! 67 Gaps in Information: Who Does the WRC Serve? When Are Cases Registered? These basic questions are difficult to answer due to the current documentation system in place at both Balicha and Kharadiwara. Consistent documentation would be helpful to show the types of cases coming to the center, including the date that cases are being registered. I took the data from the receipt book and charted cases based on month the receipt was written (see Figure 3 below). I know that for many cases the receipt is not written when the case was actually registered, so this chart is not accurate. May and July had the most total cases over the six-year period (eight cases), while zero receipts were made during both September and December. More accurate information about who comes to the center to register a case when, both in terms of the conflict and the time of year, would be valuable information for Seva Mandir and other violence prevention initiatives. If there is a seasonal trend, related to holidays or periods of drought, for example, the center could possibly be open more during these times or consider outreach activities to encourage women to register cases. Fig. 3: Balicha WRC Receipts by Month, 2007-2012 "!'! #!*! $!+! %!(! ,-.! /01!2-3!453!2-6! ,7.! ,78! 479! :05! ;<=! >?@!A0