This study was funded by UK aid from the UK government. However, the views expressed do not necessarily reflect the UK government’s official policies. Suggested citation: Institute of Development Studies (IDS), Ghana Statistical Services (GSS) and Associates (2016), Domestic Violence in Ghana: Incidence, Attitudes, Determinants and Consequences, Brighton: IDS. © Institute of Development Studies 2016 ISBN: 978-1-78118-314-4 This is an Open Access report distributed under the terms of the Creative Commons Attribution 4.0 International licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are credited. http://creativecommons.org/licenses/by/4.0/legalcode Contents List of tables and figures 4 Foreword 7 Acknowledgements 9 Research team 12 Acronyms and abbreviations 14 Executive summary 15 1. Background to the study 20 1.1. Domestic violence in Ghana 21 1.2. The study 25 2. Conceptual framework 27 2.1. Types and perpetration of domestic violence 31 2.2. Attitudes towards domestic violence 34 2.3. Determinants of domestic violence 35 2.4. Consequences of domestic violence 39 2.5. Use of and access to victims’ support services and institutions 41 3. Methodology 44 3.1. Quantitative instruments 44 3.2. Qualitative instruments 46 3.3. Sampling design: quantitative survey 48 3.4. Sampling design: qualitative fieldwork 49 3.5. Training and pre-testing 50 3.6. Fieldwork implementation 51 3.7. Response rates 52 3.8. Data analysis 57 3.9. Ethical considerations 58 4. Empirical results 61 4.1. Incidence of violence (domestic and non-domestic) against women and men in Ghana 61 4.1.1. Social violence 62 4.1.2. Physical violence 73 4.1.3. Sexual violence 83 4.1.4. Psychological violence 93 4.1.5. Economic violence 103 4.2. Incidence of domestic violence against women and men in Ghana 113 4.2.1. Domestic social violence 114 4.2.2. Domestic physical violence 121 4.2.3. Domestic sexual violence 128 4.2.4. Domestic psychological violence 135 4.2.5. Domestic economic violence 142 4.2.6. Summary 149 1 4.3. Perpetration of domestic violence in Ghana 150 4.3.1. Violence perpetration within domestic relations 150 4.3.2. Profile of domestic violence perpetrators 153 4.3.3. Domestic violence against children 158 4.4. Attitudes towards domestic violence in Ghana 159 4.4.1. Attitudes towards domestic social violence 160 4.4.2. Attitudes towards domestic physical violence 166 4.4.3. Attitudes towards domestic sexual violence 179 4.4.4. Attitudes towards domestic psychological violence 182 4.4.5. Attitudes towards domestic economic violence 184 4.5. Determinants of domestic violence in Ghana 185 4.5.1. Age and intergenerational effects 187 4.5.2. Socio-economic effects 189 4.5.3. Marital status 193 4.5.4. Geographical effects 196 4.5.5. Patriarchal and gender norms 197 4.5.6. Exposure to other forms of violence 200 4.6. Consequences of domestic violence in Ghana 201 4.6.1. Physical health effects 202 4.6.2. Mental health and emotional effects 206 4.6.3. Effects on work and daily activities 208 4.6.4. Consequences for children 209 4.7. Use of and access to victims’ support services and institutions 211 4.7.1. Seeking external help 211 4.7.2. Public services and institutions 215 4.7.3. Knowledge about support services and institutions 219 4.7.4. Levels of satisfaction about support services and institutions 223 5. Summary and discussion 225 5.1. Incidence of domestic violence in Ghana 225 5.2. Attitudes towards domestic violence in Ghana 226 5.3. Determinants of domestic violence in Ghana 227 5.4. Consequences of domestic violence in Ghana 230 5.5. Use of and access to victims’ support services and institutions 231 6. Way forward and policy recommendations 233 6.1. Recommendations with respect to the incidence of domestic violence in Ghana 234 6.2. Recommendations with respect to attitudes towards domestic violence in Ghana 236 6.3. Recommendations with respect to the determinants of domestic violence in Ghana 238 6.4. Recommendations with respect to the consequences of domestic violence in Ghana 240 6.5. Recommendations with respect to the use of and access to victims’ support services and institutions in Ghana 241 6.6. Recommendations for future research on domestic violence in Ghana 244 References 247 2 Appendices 259 Appendix A: Incidence of domestic violence worldwide 260 Appendix B: Timeline of major policies related to domestic and gender-based violence 263 Appendix C: Summary of qualitative sample, methods and instruments 265 Appendix D: Distribution of enumeration areas across regions and survey weights 266 Appendix E: Regression tables 269 Questionnaire annex 275 3 List of tables and figures Table 1: Response rates to household survey (in percentages) .............................................. 53 Table 2: Response rates to the individual survey: women (in percentages) ........................ 54 Table 3: Response rates to the individual survey: men (in percentages) .............................. 54 Table 4: Distribution of survey respondents by key socio-economic characteristics (weighted) ...................................................................................................................................... 55 Table 5: Incidence of violence in Ghana 2015 (in percentages) .............................................. 62 Table 6: Percentage of women who have experienced domestic or non-domestic social violence over their lifetime .......................................................................................................... 65 Table 7: Percentage of women who experienced domestic or non-domestic social violence in the last 12 months ..................................................................................................................... 67 Table 8: Percentage of men who have experienced domestic or non-domestic social violence over their lifetime .......................................................................................................... 69 Table 9: Percentage of men who experienced domestic or non-domestic social violence in the last 12 months .......................................................................................................................... 71 Table 10: Percentage of women who have experienced domestic or non-domestic physical violence over their lifetime .......................................................................................................... 75 Table 11: Percentage of women who experienced domestic or non-domestic physical violence in the last 12 months ..................................................................................................... 77 Table 12: Percentage of men who have experienced domestic or non-domestic physical violence over their lifetime .......................................................................................................... 79 Table 13: Percentage of men who experienced domestic or non-domestic physical violence in the last 12 months ..................................................................................................... 81 Table 14: Percentage of women who have experienced domestic or non-domestic sexual violence over their lifetime .......................................................................................................... 85 Table 15: Percentage of women who experienced domestic or non-domestic sexual violence in the last 12 months ..................................................................................................... 87 Table 16: Percentage of men who have experienced domestic or non-domestic sexual violence over their lifetime .......................................................................................................... 89 Table 17: Percentage of men who experienced domestic or non-domestic sexual violence in the last 12 months ..................................................................................................................... 91 Table 18: Percentage of women who have experienced domestic or non-domestic psychological violence over their lifetime ................................................................................. 95 Table 19: Percentage of women who experienced domestic or non-domestic psychological violence in the last 12 months ..................................................................................................... 97 Table 20: Percentage of men who have experienced domestic or non-domestic psychological violence over their lifetime ................................................................................. 99 Table 21: Percentage of men who experienced domestic or non-domestic psychological violence in the last 12 months ................................................................................................... 101 Table 22: Percentage of women who have experienced domestic or non-domestic economic violence over their lifetime ...................................................................................... 105 Table 23: Percentage of women who experienced domestic or non-domestic economic violence in the last 12 months ................................................................................................... 107 4 Table 24: Percentage of men who have experienced domestic or non-domestic economic violence over their lifetime ........................................................................................................ 109 Table 25: Percentage of men who experienced domestic or non-domestic economic violence in the last 12 months ................................................................................................... 111 Table 26: Incidence of domestic violence in Ghana 2015 in the 12 months prior to the survey (in percentages) .............................................................................................................. 113 Table 27: Percentage of domestic social violence over the last 12 months ......................... 114 Table 28: Percentage of women who experienced domestic social violence over the last 12 months ..................................................................................................................................... 117 Table 29: Percentage of men who experienced domestic social violence over the last 12 months .......................................................................................................................................... 119 Table 30: Percentage of domestic physical violence over the last 12 months..................... 121 Table 31: Percentage of women who experienced domestic physical violence over the last 12 months .............................................................................................................................. 124 Table 32: Percentage of men who experienced domestic physical violence over the last 12 months ..................................................................................................................................... 126 Table 33: Percentage of domestic sexual violence over the last 12 months ........................ 128 Table 34: Percentage of women who experienced domestic sexual violence over the last 12 months ..................................................................................................................................... 131 Table 35: Percentage of men who experienced domestic sexual violence over the last 12 months .......................................................................................................................................... 133 Table 36: Percentage of domestic psychological violence over the last 12 months ........... 135 Table 37: Percentage of women who experienced domestic psychological violence over the last 12 months ........................................................................................................................ 138 Table 38: Percentage of men who experienced domestic psychological violence over the last 12 months .............................................................................................................................. 140 Table 39: Percentage of domestic economic violence over the last 12 months .................. 142 Table 40: Percentage of women who experienced domestic economic violence over the last 12 months .............................................................................................................................. 145 Table 41: Percentage of men who experienced domestic economic violence over the last 12 months ..................................................................................................................................... 147 Table 42: Socio-economic groups with highest incidence of domestic violence ................ 149 Table 43: Perpetrators of domestic violence as reported by women (in percentages) ...... 151 Table 44: Perpetrators of domestic violence as reported by men (in percentages) ........... 152 Table 45: Percentage and number of domestic violence perpetrators ................................. 153 Table 46: Number and percentage of perpetrators reporting violence against same or another sex ................................................................................................................................... 154 Table 47: Percentage of women reporting perpetration of domestic violence ................... 156 Table 48: Percentage of men reporting perpetration of domestic violence ........................ 157 Table 49: Perpetration of domestic violence against individuals who reported domestic violence and lived with their parents (in percentages) ......................................................... 158 Table 50: Responses to first social violence scenario (in percentages) ................................ 161 Table 51: Responses to second social violence scenario (in percentages) ........................... 163 Table 52: Responses to first physical violence scenario (in percentages) ........................... 167 5 Table 53: Responses to second physical violence scenario (in percentages) ...................... 169 Table 54: Percentage of respondents agreeing with wife-beating if the wife… ................. 172 Table 55: Percentage of respondents who think that a man has a good reason to hit his wife if: ........................................................................................................................................... 173 Table 56: Percentage of respondents stating that wife-beating is acceptable if the woman… ...................................................................................................................................... 175 Table 57: Views about statement ‘To bring up, raise or educate a child properly, the child sometimes needs to be physically punished’ (in percentages) ............................................. 178 Table 58: Views about statement ‘If a woman wears revealing clothes, it is okay for men to make some sexual comments as she walks in her neighbourhood’ (in percentages) .. 180 Table 59: Views about statement ‘If a woman wears revealing clothes, it is her fault if she is raped’ (in percentages) ........................................................................................................... 181 Table 60: Regions significantly associated with increased experience of domestic violence (by type of domestic violence and sex) .................................................................................... 197 Table 61: Relationship between experience of domestic violence and physical health (women) ........................................................................................................................................ 204 Table 62: Relationship between experience of domestic violence and physical health (men) ............................................................................................................................................. 205 Table 63: Percentage of respondents who have experienced domestic violence reporting serious mental illness (by type of domestic violence and sex) ............................................. 206 Table 64: Effect of domestic violence on men and women’s mental health ....................... 207 Table 65: Percentage of respondents who experienced domestic violence who reported seeking help (by sex) ................................................................................................................... 212 Table 66: Reasons for approaching external help after exposure to any form of domestic violence by sex (in percentages) ................................................................................................ 212 Table 67: Reasons for not seeking external help after exposure to any form of domestic violence by sex (in percentages) ................................................................................................ 213 Table 68: First choice of help (in percentages) ........................................................................ 214 Table 69: Respondents’ knowledge about public services (in percentages) ....................... 221 Table 70: Satisfaction levels with public services (number of respondents) ...................... 223 6 Foreword Domestic violence, in the form of physical or sexual violence, is experienced by 35 per cent of women globally during their lifetime. Emotional abuse and controlling behaviour is experienced by 75 per cent of women across the world. It is also acknowledged that men experience domestic violence. Nationally representative statistics on the incidence of domestic violence are, however, hard to come by. This study represents a comprehensive attempt to estimate the incidence of domestic violence in Ghana, and analyse the attitudes that sustain domestic violence, its determinants and its consequences. The study estimates that 28 per cent of women and 20 per cent of men experienced domestic violence in 2015. Domestic violence causes immediate devastating consequences to those affected: physical injuries, mental health problems and poor well-being, to name but a few. But it also has long-term, far-reaching effects, including persistent inequalities between men and women, which limit women and girls’ abilities to fulfil their potential. Those who experience domestic violence have longer-term poor physical and mental health, higher risk of contracting HIV and sexually transmitted diseases, and restricted choices in terms of accessing education and jobs. It is also likely that they will have lower levels of education, income and productivity. These consequences are not only life-changing for the women and men concerned, but also for their families, communities and wider societies, as gender inequalities become entrenched. This innovative mixed-methods study, commissioned by the Ministry of Gender, Children and Social Protection in Ghana and the UK Department for International Development, and funded by UK aid, provides an in-depth understanding of the incidence, attitudes, causes and consequences of domestic violence in Ghana, as well as investigating the effectiveness of existing institutional support offered to women and men. Ghana has, in many ways, led the way with its pioneering legislation on domestic violence, which includes various forms of economic abuse, acknowledges that perpetrators and survivors do not have to be married or related by blood ties, and takes a broader perspective on access to justice than other countries, allowing for mediation by alternative dispute resolution methods. The Domestic Violence Act 732, adopted by Parliament in 2007, outlines a comprehensive legal framework for the prevention of and protection against domestic violence and criminalises various acts of physical and sexual violence, economic and psychological abuse, and intimidation in domestic relations. Despite this innovative legislative work, the results of this study show that there is more work to be done to ensure that the laws are enacted and understood by the broader population. The findings of this work will be used to formulate ongoing policies and 7 programmes to ensure that, moving forward, fewer and fewer people will be affected by domestic violence in Ghana. Hon. Nana Oye Lithur Minister for Gender, Children and Social Protection 8 Acknowledgements This report discusses the main findings of a comprehensive study of the incidence, determinants and consequences of domestic violence, attitudes towards domestic violence and the effectiveness of services that support women and men who experience domestic violence in Ghana. The study was commissioned by the Domestic Violence Secretariat of the Ministry of Gender, Children and Social Protection (MoGCSP) of the Government of Ghana and the UK Department for International Development (DFID), and funded by UK aid. The study was led by the Institute of Development Studies (IDS) in the United Kingdom (UK) and carried out in cooperation with the Ministry of Gender, Children and Social Protection, the Ghana Statistical Service (GSS) and researchers at the Institute of Statistical, Social and Economic Research (ISSER) in Accra. The research team is very grateful to Professor Akosua Adomako Ampofo (Institute of African Studies at the University of Ghana), who reviewed the quantitative and qualitative instruments and participated in the quantitative pilot training exercises; to Mrs Dorcas Coker-Appiah for providing invaluable guidance during training to the enumerating team about domestic violence and ways of relating to potential victims of domestic violence during the survey; and to Mr Adolf Awuku Bekoe (Coordinator, DV Coalition) for his help in coordinating referral support provided by different organisations to the study participants who requested it across the regions covered by the study. Mr Adolf Awuku Bekoe also provided invaluable psychological support to the fieldwork team when needed via a telephone helpline set up for the study. At the MoGCSP, the study was supported by the Honourable Minister Nana Oye Lithur and her team. The Honourable Minister chaired the study’s Steering Committee, which provided feedback on the study’s outputs at key stages, facilitated contacts with key stakeholders beyond the Steering Committee representatives, and supported the outreach and uptake of the study. The Steering Committee was comprised of Mr Kwesi Armo-Himbson (Chief Director, MoGCSP), Mrs Efua A. Anyanful (MoGCSP), Ms Fauzia Issaka (DFID Ghana), Ms Lynne Henderson (DFID Ghana), Mrs Patience Opoku (MoGCSP), Mrs Abena Annobea Asare (MoGCSP), Mr Kwabena Twumasi (Human Trafficking Secretariat), Mrs Lilian Ayete- Nyampong (Commission on Human Rights and Administrative Justice – CHRAJ), ACP Habiba Twumasi-Sarpong (Domestic Violence Victim Support Unit – DOVVSU), Mr Adolf Awuku Bekoe (DV Coalition), Mrs Ruth Aba Grant (Network for Women’s Rights – NETRIGHT), Professor Felix A. Asante (ISSER), Mr Daniel Degbotse (Ministry of Health) and Mr Clarke Noyoru (MoGCSP). The research team is grateful to Mrs Victoria Natsu for the logistical support she provided to meetings and interactions with the Steering Committee throughout the study. 9 The study benefited from excellent input, discussions and support from DFID Ghana. Ms Lynne Henderson provided support to the design of the study, participated in several discussion and reviewed drafts of the report. Mr Archie Laing provided invaluable logistical support to the team throughout the study. Ms Fauzia Issaka engaged with the research team at all stages of the study, provided comments on several drafts of the fieldwork instruments, participated in numerous research discussions, helped the team to identify and liaise with key stakeholders and reviewed in detail multiple drafts of the report. The depth and quality of her feedback has helped substantially to improve the report. The research team is extremely grateful for her unfaltering support and encouragement during the entire study. At IDS, the research team benefited from fantastic research assistance by Marco Carreras and Eric Casper. Deborah West and Veronica Moore managed the contractual and financial aspects of the study. Deborah West also provided invaluable assistance with communication activities and editorial support. The study also benefited from excellent comments and support from Dr Jerker Edstrom, Professor John Gaventa, Professor Melissa Leach and, especially, Dr Mariz Tadros, who read and commented on the entire initial draft of the report. The study’s qualitative and quantitative instruments and various drafts of the report were reviewed by four members of the study’s Advisory Board: Mrs Dorcas Coker-Appiah (Executive Director of the Gender Centre, Accra), Dr Akosua K. Darkwah (Director, Centre for Gender Studies and Advocacy, University of Ghana), Dr Nata Duvvury (Co-Director of the Global Women’s Studies Programme, National University of Ireland, Galway) and Dr Katherine Brickell (Royal Holloway University of London). The research team is very grateful for their contribution, support and invaluable and constructive comments at all stages of the study. The research team would also like to thank key stakeholders from the Gender Centre, DV Coalition, DV Secretariat, Ark Foundation, International Federation of Women Lawyers (FIDA), Leadership and Advocacy for Women in Africa (LAWA) and other organisations, in particular Ms Dinah Adiko (consultant, DV Secretariat, MoGCSP), Ms Juliana Amponsah (former Acting Executive Secretary, DV Secretariat, MoGCSP), Superintendent Laurencia Akorli (Deputy Coordinating Director of DOVVSU), Mrs Barbara Ayesu (Coordinator, LAWA (Ghana) Alumnae Inc.), Dr James Clayman (Medical Superintendent and Obstetrician Gynaecologist at the Ga West Municipal Hospital), Mrs Dorcas Coker- Appiah (Executive Director, Gender Center), Mrs Esther Darko-Mensah (Gender Centre), Mrs Angela Dwamena-Aboagye (Executive Director, Ark Foundation), Dr Elvis Sadongo (Head of Social Services Department at Ghana Police Hospital) and Mrs Jane Quaye (Executive Director, FIDA-Ghana) for their warm welcome, encouragement and invaluable support. The research team is thankful for the outstanding work of the fieldwork teams at the Ghana Statistical Service and the Institute of Statistical, Social and Economic Research. 10 Most importantly, the research team is extremely grateful to all the women and men who took time from their daily activities to talk about their experiences – some quite painful and distressing. This study would have never been possible without them. 11 Research team Core research team: Ms Alia Aghajanian (IDS), Mr Anthony Amuzu (GSS), Dr Elizabeth Asante (ISSER), Professor Patricia Justino (IDS) (Team Leader), Dr Marinella Leone (IDS), Dr Elizabeth Mills (IDS), Ms Sheila Minkah-Premo (Apex Lawconsult), Ms Rebecca Mitchell (IDS), Dr Catherine Müller (IDS), Dr Philomena Nyarko (Government Statistician, GSS), Dr Pauline Oosterhoff (IDS) and Dr Jean-Pierre Tranchant (IDS). Fieldwork team at the Institute of Statistical, Social and Economic Research: Mohammed Ben Abdallah, Adom K.M. Acheampong (Supervisor), Otuo Acheampong, James Allotey, Aboagye Amankona, Dinah Badmus, Daniel Danso, George Yaw Dugbatey, Daniel Gyasi, Samuel Gyasi-Adonteng (Supervisor), Michael Johnson, Randy Korley, Eric Obeng, Joseph K. Okrah and Emmanuel Nuetey Siakwa (Supervisor). Fieldwork team at the Ghana Statistical Service: Dr Philomena Nyarko (Government Statistician/Project Director), Mr Baah Wadieh (Deputy Government Statistician/Deputy Project Director), Mr Anthony Amuzu (Acting Deputy Government Statistician/Survey Director), Mr Peter Takyi Peprah (Project Coordinator), Mr Emmanuel George Ossei (Trainer/Field Monitor), Mr Emmanuel Boateng (Trainer/Field Monitor), Mr Godwin Odei Gyebi (Trainer/Field Monitor), Mr K.B. Danso-Manu (Field Monitor), Mrs Araba Forson (Field Monitor), Mr Owusu Kagya (Field Monitor), Mrs Betty-Love Hermans Cobbinah (Internal Auditor), Mr Robert Kwami (Director of Finance), Mr Kwadwo Asante-Mensah (Director of Administration), Mrs Abena Asamoabea Osei-Akoto (Head, Data Processing), Mrs Lucy Twumwaa Afriyie (Head, Sampling Design), Ms Hannah Konadu Frimpong (Project Secretary), Mrs Rosalind Quartey (Head, Geographic Information System) and Ms Elizabeth Arhin Donkor (Project Accountant). The team of regional statisticians included Isaac Addae, Salifu Amadu, Chris Amewu, Kobina Abeka Ansah, Chris Assem, Samuel Brefo Darkwa, Sixtus Dery, Festus Manuh, Amatus S. Nobabuma and Ernest Nyarku. The team of field supervisors included Barnas Abugre, Michael Opoku Acheampong, Jane Acolatse, Samuel Owusu Agyemang, Dieudonnee Ankamah, Daniel Ansah, Emmanuel Atanga, Solomon Owusu Bempah, Victor Owusu Boateng, Patrick Kwame Djangba, Felix Geli, William Ofosu, Stephen Tabi, Joseph, Asiedu Tenkorang and Richard Boakye Yiadom. The team of interviewers included Daniel Amos Abanyie, Nana Afua Acheampong, Mavis Adebona, Dennis Addo, Judith Addo, John Agbadza, Michael Agyarkwa, Frank Agyei, Habib Alhassan, Sandra Amartey, Anastasia Ammisah, Esther Adwoa Amoah, Jolene Mantey Amoo, Caroline Anipah, Anthony Ankomah, Umul Klusum Arimiyaw, Konlan Alhasssan Awal, Gloria Bekoe, Vida Akuamoah Boateng, Gladys Ataa Dabison, Mabel Diameh-Bentil, Getrude Elleammoh, Afu Anita Elorm, William Eyimah, Bright Gyamfi, Sulemana Idriss, Felicity Kuwornu, Emmanuel Kyei, Evelyn Kyeremateng, Mark Ato Larbi, Mary Larbi, Zainab Ibrahim Maltiti, Happy Krah Mawuse, Richard Mills, Linda Ntiamoah, Israel Kofi Menash Nyarko, Pricilla Opoku, Seth Opoku, Mercy Naa Quarshie, 12 Phylicia Saforo, Rabiatu Seidu, Edward Tebiesinaa, Linda Ama Tenkorang, Isaac Kwame Yeboah and Benjamin Kusi Yentumi. The field teams were supported by a team of drivers: Ebenezer Abrokwah, Emmanuel Acquah, Isaac Addae, Joseph Afedzie, Elvis Agyei, David Agyemang, William Akamaboro, Joseph Ampomah, George Ansah, Eugene Aniapam, Peter Ayensu, Samuel Ayikwei, Emmanuel Chartey, Samuel Dam, Ebenezer Dinsey, Kwesi Donkor, Tweneboah Gyesi, Jacob Jabotir, George Karikari, Simon K. Kasakwa, Wisdom Kuagbelah, James Nii Laryea, Victor Mensah, Gershon Nornyibey, Emmanuel Oduro, Isaac Opare, Felix Osei, Stephen Otoo and Stephen Sanbonaa. 13 Acronyms and abbreviations ACHPR African Charter on Human and Peoples’ Rights CEDAW UN Convention on the Elimination of All Forms of Discrimination against Women CHRAJ Commission on Human Rights and Administrative Justice CSO Civil Society Organisation DE Design effect DEVAW Declaration on the Elimination of Violence against Women DFID UK Department for International Development DHS Demographic and Health Survey DOVVSU Domestic Violence Victim Support Unit DV Domestic violence DV Secretariat Domestic Violence Secretariat EA Enumeration area FGD Focus Group Discussion FGM Female Genital Mutilation FIDA International Federation of Women Lawyers Gender Centre Gender Studies and Human Rights Documentation Centre GFLHS Ghana Family Life and Health Survey GSS Ghana Statistical Service GVSSN Gender Violence Survivors’ Support Network HIV Human Immunodeficiency Virus IDS Institute of Development Studies IPV Intimate Partner Violence JHS Junior High School JSS Junior Secondary School LAWA Leadership and Advocacy for Women in Africa MICS Multiple Indicator Cluster Survey MoGCSP Ministry of Gender, Children and Social Protection NETRIGHT Network for Women’s Rights NGO Non-governmental Organisation NPPOA National Policy and Plan of Action SDGs Sustainable Development Goals SMI Serious mental illness UK United Kingdom UN United Nations UNICEF United Nations Children’s Fund USA United States of America VAWG Violence against women and girls WAJU Women and Juvenile Unit of the Ghana Police WHO World Health Organization WISE Women’s Initiative for Self-Empowerment 14 Executive summary Domestic violence is a violation of basic human rights and a global policy concern. Following several years of advocacy efforts by civil society and women’s rights organisations, the Government of Ghana enacted the DV Act (act 732) in February 2007, followed by the formulation of the National Policy and Plan of Action (NPPOA) in 2008. The NPPOA laid out a strategy for the implementation of the Domestic Violence Act, under the direction of the Domestic Violence Secretariat, part of the Ministry of Gender, Children and Social Protection (MoGCSP). This study forms part of this strategy. The overarching aim of the study is to provide an in-depth understanding of the attitudes to and incidence, determinants and consequences of domestic violence in Ghana. The study addresses the following questions: 1. What types and levels of violence are experienced by women, men, girls and boys in Ghana, and who perpetrates them? What proportion of this violence is domestic violence? 2. What are society, community and people’s attitudes towards domestic violence, and how do attitudes and norms influence the acceptance of different forms of domestic violence in Ghana? 3. What are the main determinants of domestic violence against women, men, girls and boys at the individual, family, community and wider societal levels? 4. What are the consequences of domestic violence in Ghana for people’s physical and mental health and their daily lives? 5. What institutional structures support victims of domestic violence, and what proportion of victims are able to effectively access available services? 6. How can policies and interventions be (re)designed to better support victims of domestic violence and prevent domestic violence in Ghana? Conceptual approach. The conceptual framework adopted in the study – the critical social ecological model – uses an explicit political economy lens to analyse the complex interplay between the determinants and consequences of domestic violence at multiple levels of analysis: the individual, the family, the community and the wider social and political system. This conceptual framework formed the basis for the collection of new empirical evidence across Ghana on: (i) the incidence and types of domestic violence; (ii) the attitudes and norms that drive domestic violence; (iii) the determinants of domestic violence; (iv) the consequences of domestic violence; and (v) the use of and access to victims’ support services and institutions. Conforming to international definitions and to definitions included in the 2007 Domestic Violence Act (Act 732), the study analysed the incidence of the following forms of violence:  Social violence: defined as acts of controlling behaviour, such as preventing someone from seeing friends or family; stopping someone from leaving the house; requiring to know where someone is at all times; stalking; spreading false information, videos 15 or photos without permission; or forcing women to have an abortion.  Physical violence: defined as slapping, pushing, shoving, hitting, kicking, dragging or throwing objects at someone; choking, strangling or burning someone; using a weapon, hazardous chemicals or substances against someone; or kicking or pulling someone’s external genitalia (for male respondents only).  Sexual violence: defined as acts of unwanted sexual comments or physical contact; rape by physical force, or otherwise forced sex (for instance, by blackmail or threats); denial of using protection during sex; a sexual partner hiding their HIV status; sexual acts and intercourse that were performed on the basis of feeling there was no option; or penetration with an object against someone’s will.  Psychological violence: defined as the use of insults, belittling or humiliation in private or in front of others; threats of abandonment; being ignored or treated indifferently; intimidations and acts aimed at scaring someone; threats of using weapons against someone; or threats of hurting someone or someone one cares about.  Economic violence: defined as the denial of household money for expenses (chop money) even if enough financial means are available; unsolicited taking of money; control of belongings and spending decisions; damage to or destruction of someone’s property; denial of the right to work; forcing someone to work against their will; or denial of food and other basic needs. Methodological approach. The study adopted a mixed-methods approach. This involved the collection of primary quantitative data through a representative household-level survey of 4,995 individuals, and the collection of qualitative in-depth data on key factors that may explain the incidence of domestic violence in Ghana. The main quantitative tool used in the study was a survey entitled the “Ghana Family Life and Health Survey 2015” (GFLHS 2015). The GFLHS 2015 was conducted between April and August 2015. Given the sensitive nature of the topic, the preparation of the survey took six months and involved intense testing, piloting and training. This careful approach ensured high response rates, ranging between 97.6 per cent in the Western region and 99.6 per cent in the Brong Ahafo region. The GFLHS 2015 included one household and one individual questionnaire. The study focused on men, women, boys and girls between the ages of 15 and 60 years. This age group extends the usual age group included in international surveys (typically 15 to 49 years), to capture the incidence of domestic violence among older women and men. Children below the age of 15 years old were outside the scope of the study. The GFLHS 2015 asked detailed questions about individual experiences of any of the five broad categories of domestic violence that occurred in the 12 months prior to the survey. The sensitivity of the research topic raised several ethical and methodological challenges that were carefully addressed during the study. The safety of respondents and the research team was paramount, and guided all decisions in the study. Protecting the confidentiality of the respondents was essential to ensure their safety and data quality, and all researchers 16 were carefully selected and received specialised training and ongoing support. Field researchers were also trained to refer individuals requesting assistance to available local services and sources of support. Incidence of domestic violence. The study provides an in-depth analysis of the incidence, attitudes, determinants and consequences of domestic violence in Ghana. The main results show that 27.7 per cent of women and 20 per cent of men in Ghana have experienced at least one type of domestic violence in the 12 months prior to the survey. The most common form of domestic violence reported by women in the 12 months prior to the survey was economic violence (12.8 per cent), followed by social violence (11.6 per cent), psychological violence (9.3 per cent), physical violence (6.0 per cent) and sexual violence (2.5 per cent). The incidence of domestic violence was generally lower among men than among women. The most common form of domestic violence experienced by men was psychological violence (7.9 per cent), followed by social violence (7.7 per cent), economic violence (7.3 per cent), physical violence (2.1 per cent) and sexual violence (1.4 per cent). The GFLHS 2015 is not comparable to prior surveys of domestic violence in Ghana because it has collected data on broader and more in-depth definitions of domestic violence than other surveys. It is, however, possible to compare the incidence of physical domestic violence experienced by women and men in the 2008 DHS and in the 2015 GFLHS. The results show, when using comparable categories of domestic physical violence, that domestic physical violence among women decreased from 17.2 per cent in 2008 to 10.3 per cent in 2015. Domestic physical violence among men decreased from 12.7 per cent in 2008 to 11.2 per cent in 2015. Attitudes towards domestic violence. Respondents did not consider some of the definitions of domestic violence used in the study (and in Act 732) to be a form of violence. Many of the acts mentioned in the study were viewed as accepted social norms, defence mechanisms or ways of resolving disputes. Only physical violence between partners was consistently perceived as violence. Psychological violence, in particular, was generally perceived as a means of defence and a coping strategy – for women in particular – to ward off physical violence or to claim some autonomy or agency in contexts where they may not have the ability to exert much power. Women were on average more likely than men to condone forms of domestic violence: 23.1 per cent of women and 13.8 per cent of men found wife-beating acceptable, while 65.3 per cent of women and 56.2 per cent of men fully agreed that women were to blame for rape if they wore revealing clothes. These findings suggest that domestic violence in Ghana may persist due to harmful social norms that hold women responsible for the violence they experience. 17 Determinants of domestic violence. Age was found to be one of the key determinants of domestic violence in Ghana. Young women and men (15–19 years old) were substantially more likely to have experienced domestic violence than other age groups. Women aged 15–19 years were between 1.3 and four times more likely to experience any form of domestic violence than, for instance, women aged 30–39 years. Differences in the incidence of domestic violence across age groups were similar for men, with the exception of sexual violence, which was more prevalent among men aged 20–24 years. Exposure to violence in childhood was found to be strongly related to the likelihood of an individual being a victim or perpetrator of violence in adulthood. Women who witnessed any form of domestic violence as children were between 1.3 and three times more likely to report domestic violence than women who did not witness violence during their childhood. Men who witnessed social, psychological or economic violence in their homes as children were between 1.2 and 2.2 times more likely to have experienced these forms of violence in adulthood than other men. Taken together, these findings suggest that Ghana may experience vicious intergenerational cycles of domestic violence that need to be broken if further progress is to be made in reducing domestic violence. In addition to age, socio-economic factors were found to be among the most important determinants of domestic violence in Ghana, albeit in complex ways:  Low asset levels were found to be strong determinants of psychological and economic violence among women, but high asset levels were associated with higher levels of social and sexual violence among men.  Tensions and fights over money or property were found to be leading determinants of all forms of domestic violence across all regions.  Women with no education were generally (between 0.5 and 2.5 times) less likely to experience domestic violence than women with some level of education. More educated men were around three times more likely to experience economic or sexual violence than less educated men. This result may be due to the fact that women and men with more education may also be more likely to report domestic violence, as emphasised in previous studies conducted internationally and in Ghana.  The qualitative analysis suggested that factors such as poverty, unemployment and economic shocks were identified as important determinants of domestic violence. Overall, the determinants of domestic violence varied considerably among types of violence, emphasising the need for domestic violence legislation in Ghana to continue to distinguish between different forms of violence (as in Act 732). Consequences of domestic violence. The GFLHS 2015 and the qualitative data revealed serious adverse consequences of domestic violence on physical health, including injuries and illness. The effects of domestic violence on physical health were observed almost 18 exclusively among women: 43.8 per cent of women who had been physically assaulted declared having been ill in the 30 days prior to the survey – a rate that was one third higher than among women who had not been physically assaulted (31.2 per cent). 42.3 per cent of women who had experienced psychological violence reported ill health, compared to 30.9 per cent of women who had not experienced this form of domestic violence. The results also showed a strong correlation between exposure to domestic violence and mental health, particularly among women: women who reported serious mental illness were on average around 25 per cent more likely to have experienced at least one form of domestic violence than women who did not report serious mental illness. The association between serious mental illness and domestic violence was generally lower among men. Only physical violence was associated with men’s mental health status. Domestic violence was also found to have adverse consequences on people’s daily life in terms of their ability to work, go to school or do domestic work, their ability to concentrate on daily activities, levels of confidence and feelings of living in fear, although few respondents answered this question. The effects were larger for women than for men, and for physical violence than for other types of violence. Use of and access to services. The study investigated in detail how victims of domestic violence make use of and access support services and institutions across Ghana. The results showed that just over one in every three women and men who experienced domestic violence approached a person, formal or informal organisation for help or support. The majority of respondents who sought help first approached family members or a friend, but 14.6 per cent of men and 9 per cent of women first contacted the police for help. Individuals who experience domestic violence appear to only resort to help beyond friends and family in extreme circumstances because domestic violence is still considered largely a private family matter. Distrust of, and poor experience with, state authorities and public services also contributed to not seeking help, a result that calls for further efforts to strengthen the services and institutions aimed at supporting women and men affected by domestic violence in Ghana – as well as coordination between them. Recommendations on moving forward are provided at the end of the report, where detailed roadmaps are outlined on how best to address the complex intersecting factors that underlie the incidence of domestic violence among women and men and across different, socio-economic groups and locations across Ghana. 19 1. Background to the study Domestic violence – defined as abuse by one person against another in an intimate relationship including marriage, cohabitation, dating or relations within the family – is one of the most common forms of gender-based violence in the world (Abramsky et al., 2011; Ellsberg et al., 2008; Garcia-Moreno et al., 2006, 2013; USAID, 2006). International studies estimate that approximately 35 per cent of women across the world have experienced physical and/or sexual violence at some point in their lives, largely in the form of domestic violence (García-Moreno et al., 2013; Appendix A). 1 A recent systematic review has reported that at least one in seven homicides and over one third of all female homicides worldwide are perpetrated by an intimate partner (Stockl et al., 2013). Physical and sexual violence are not the only types of domestic violence perpetrated against women: emotionally abusive acts and controlling behaviour are experienced by up to 75 per cent of women worldwide (García-Moreno et al., 2005). Domestic violence is also associated with persistent forms of gender inequality and adverse health and economic outcomes among victims, including poor physical and mental health, higher risks of the human immunodeficiency virus (HIV) and other sexually transmitted diseases, restricted livelihood options and choices, lower human capital and lower productivity (García- Moreno et al., 2005, 2013; Moosa, 2012). Civil society and governments around the world have acknowledged that violence against women and girls (VAWG), including domestic violence, is a violation of basic human rights and a global policy concern. In recognition of this international attention, “eliminating all forms of violence against all women and girls in the public and private spheres” is one of the targets against which Sustainable Development Goal number 5 – aimed at achieving gender equality and empowering all women and girls – will be measured. To further underscore the importance of tackling VAWG, the “elimination and prevention of all forms of violence against women and girls” was chosen as the review theme at the 60th Session of the United Nation’s (UN) Commission on the Status of Women in March 2016. Violence against women within private relations was drafted into international legislation in the 1970s, as women’s movements in the United Kingdom and the United States of America (USA) drew attention to the extent of violence committed by intimate partners.2 Domestic violence was first recognised at the international policy level in 1993 in the 1 This analysis is based on data from 155 studies in 81 countries. For further information on domestic violence internationally, see also the Sexual Violence Research Initiative (http://www.svri.org/index.htm), and the DFID-funded programme on What Works to Prevent Violence: A Global Programme to Prevent Violence Against Women and Girls (http://www.whatworks.co.za/). Appendix A includes a comprehensive table compiled by UN Women on the incidence of various forms of domestic violence across the world according to the latest available information. 2 The first known use of the term ‘domestic violence’ in a political setting can be traced back to 1973, to an address in the UK Parliament: in this address, domestic violence was explained to mean ‘spouse abuse, violence in the home’: http://hansard.millbanksystems.com/commons/1973/jul/16/battered-wives. See also Dobash and Dobash (1977), Freeman (1977) and Shanley (1993). 20 Declaration on the Elimination of Violence against Women (DEVAW), which framed gender-based violence as a human rights violation.3 In these international declarations, sexual and gender-based violence were defined primarily as violence committed by men against women and girls. DEVAW, in particular, articulated three interlinked spheres in which sexual and gender-based violence took place: in the family (including marital rape, sexual abuse, female genital mutilation and dowry-related violence); in the community (including rape, sexual harassment and sex trafficking); and by the State (all forms of violence that are condoned or perpetrated by state actors).4 This approach was informed by feminist theories that considered violence against women to be fundamentally different from violence against men (Africa, 2010; Haraway, 1988; Yodanis, 2004). Emergent fields of sexuality and masculinities studies have shown that domestic violence takes place against women, against some men and also against children (Hunnicutt, 2009; see also Institute of Development Studies et al., 2015). This study provides a nationally representative, in-depth analysis of the incidence, levels of acceptance, determinants and consequences of domestic violence, and the effectiveness of institutional support offered to women and men who experience domestic violence in Ghana. Section 1.1 below discusses in more detail current trends and research on domestic violence in Ghana, while Section 1.2 describes the study. 1.1. Domestic violence in Ghana The Republic of Ghana is located along the Gulf of Guinea in the sub-region of West Africa. It has a total land area of 238,537 square kilometres and is bordered by Ivory Coast in the west, Burkina Faso in the north, Togo in the east and the Atlantic Ocean in the south. Ghana is divided into ten administrative regions: Western, Central, Greater Accra, Volta, Eastern, Ashanti, Brong Ahafo, Northern, Upper East and Upper West. 3 Prior to this, human rights law did not cover gender-based violence because it excluded the ‘private sphere’ in which many forms of gender-based violence, such as domestic violence, took place. 4 1993 UN Declaration on the Elimination of Violence against Women. 21 Ghana’s population is around 24 million, according to the 2010 Census (Ghana Statistical Service (GSS), 2013). Almost half of the total population live in the Ashanti, Eastern and Greater Accra regions. Upper East is the least populated region, with 2 per cent of the total population of Ghana. Over 50 per cent of the population live in urban areas (GSS, 2013). The Ghanaian population is made up of several ethnic groups. The Akans are the largest group (48 per cent), followed by the Mole-Dagbani (17 per cent), Ewe (14 per cent), Ga- Dangme (7 per cent) and others (GSS, 2013). The sex ratio, according to the 2010 Census, is 95.2 males per 100 females. According to the Demographic and Health Survey (DHS) conducted in Ghana in 2008, 38.7 per cent of ever-married women between the ages of 15 and 49 years reported having experienced physical, psychological or sexual violence by a husband or partner at some point in their lives. Over a quarter (27.6 per cent) of Ghanaian males reported having experienced physical or psychological violence by their wife or partner (GSS et al., 2009). Considerable effort has been made in Ghana over the last three decades to reduce the incidence of domestic violence. Some of the first studies on domestic violence in Africa took place in the 1990s in Ghana, as well as in Tanzania, Uganda and South Africa (Hodgson, 2002; Ofei-Aboagye, 1994; Watts, Osam and Win, 1995). These studies were motivated by the actions of activist groups, which played an important role in the formulation and passing of domestic violence laws in Ghana (Bowman, 2002; Kimuna and Djamba, 2008; Schneider, 2008). 22 Ghana’s first legislation efforts against domestic violence reflected the global momentum in pushing for women’s rights to be recognised as human rights (Cook, 2011),5 prompted by years of advocacy from key civil society organisations (CSOs), women’s rights organisations and international bodies. These actions resulted in the Government of Ghana enacting a number of national laws to protect women’s rights and outlaw violence against women and girls. These included a provision in the 1992 Constitution that prohibited discrimination based on sex; the 1998 Criminal Code Amendment Act; 6 and legal amendments criminalising certain harmful traditional practices, such as widowhood rites (1984), female genital mutilation (FGM) (1995) and child abuse (1998). In February 2007, Ghana’s Parliament voted to enact the Domestic Violence Act (Act 732). Appendix B provides a detailed historical timeline of international and Ghanaian policies related to domestic violence. The enactment of the 2007 Domestic Violence Act was the culmination of a long process of advocacy and civil action, which started with the publication of the Gender Studies and Human Rights Documentation Centre’s (Gender Centre’s) research on violence against women and children, by Coker-Appiah and Cusack (1999), and continued under the leadership of the Gender Centre, the Women’s Initiative for Self-Empowerment (WISE) and the Ark Foundation.7 Responses to domestic violence were largely initiated by non- governmental organisations (NGOs), CSOs and networks and coalitions in Ghana that worked in the areas of human rights and women’s rights (Amoakohene, 2004; NETRIGHT, 2012).8 Together, these institutions have been engaged in activities to change perceptions, attitudes and harmful social norms among local populations and within public institutions. The efforts to put the Domestic Violence Act in place in the late 1990s were accompanied by much publicity in the media. Physical spousal abuse became a regular theme in media reports, which resulted in the establishment of the Women and Juvenile Unit (WAJU) of the Ghana Police in 1998, a specialised unit that handled crimes against women and children (Amoakohene, 2004). During early 2002, an increased number of stories on a series of murders of women by their partners were publicised, fuelling further the debate on violence against women in Ghana and enabling CSOs to mobilise public action. The role 5 These include the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW), the African Charter on Human and Peoples’ Rights (ACHPR) and the Maputo Protocol on the Rights of Women in Africa. 6 This Act amended the Criminal Code of 1960 by increasing the age of criminal and sexual responsibility, including a specific offence of indecent assault and revising provisions regarding sexual offences. It also abolished customary or ritual servitude. 7 WISE provides counselling and other support services to survivors of domestic violence and has played a central role in increasing public awareness about domestic violence. The Ark Foundation was instrumental in forming various gender coalitions, such as the Sister’s Keepers, the Gender Violence Survivors’ Support Network (GVSSN), the Network for Women’s Rights (NETRIGHT) and the Violence against Women and Children’s Practitioners Network, among others. 8 These included, among others, the Ghana branch of the International Federation of Women Lawyers (FIDA- Ghana), which worked together with other NGOs to officially abolish trokosi (a traditional religious practice whereby human beings, usually young virgin girls, are used as payment to the Gods) and to lobby the government to pass the Domestic Violence Bill, among other initiatives. 23 of the media at that time was criticised for their sensational framing of events and ill- considered language that left the impression that domestic violence was something unusual, based on individual malfunctioning rather than being situated within wider cultural and institutional norms and practices. But Gadzepko (2009: 266) also describes the mass media as “a key channel of information and debate on domestic violence and consequently [...] influential in constructing public understanding of violence against women and children”. Efforts by civil society actors and NGOs resulted in successful mobilisation, advocacy, sensitisation and education about domestic violence, and the training of state officials on domestic violence and gender issues. Their efforts achieved momentum with the coming together of stakeholder CSOs into the National Coalition on Domestic Violence Legislation, established in 2003, which “at various times worked closely with, and at other times independently of, or even in conflict with, the State” (Adomako Ampofo, 2008: 1). Finally, in 2007, Ghana’s Parliament voted to enact the Domestic Violence Act (Act 732). The actual Bill had been before Parliament since 2003 and was the “subject of heated debate” (Adomako Ampofo, 2008). According to Takyiwaa Manuh, the “process leading to the passage of the law involved not only the introduction of new legislation, but also confronting a social system that tolerates various forms of violence against women and girls, especially in the context of gender relations and in the domestic sphere” (Manuh, 2007: 1). Ghana’s domestic violence legislation takes a broader and, arguably, culturally sensitive approach to access to justice, when compared with other countries. First, the 2007 Domestic Violence Act allows for mediation by alternative dispute resolution methods.9 Second, the definition of domestic violence used in Act 732 does not include reference to a specific sex. Third, the Act acknowledges that perpetrators and survivors do not have to be married or related by blood ties, and applies to live-in household staff too. Fourth, the definition of domestic violence includes various forms of economic abuse, in addition to more conventional definitions of sexual and physical violence. The Act also provides a working definition of domestic violence, and outlines a comprehensive legal framework for the prevention of and protection against domestic violence. Notably, the law criminalises various acts of physical and sexual violence, economic and psychological abuse, and intimidation and harassment in domestic relations. These broad definitions of domestic violence have implications for research on the incidence, nature, causes and consequences of domestic violence in Ghana, which will be discussed in more detail in the next section. 9 Act 732 allows for the Court, with the consent of the complainant, to refer the case for settlement by an alternative dispute resolution method if the criminal trial of a domestic violence case is not aggravated; if the complainant requests to have the matter settled out of court; or if the court determines that the case “can be amicably settled”. Local mediation, however, entails its own challenges, which will be discussed in Section 4.5.4. 24 1.2. The study The Ghana Domestic Violence Act (Act 732) was followed by the formulation of the National Policy and Plan of Action (NPPOA) developed by the former Ministry of Women and Children’s Affairs in 2008. The NPPOA lays out the specific roles of key stakeholders that will effectively implement the Domestic Violence Act, under the direction of the Domestic Violence Secretariat, part of the Ministry of Gender, Children and Social Protection (MoGCSP). This study forms part of this strategy. The overarching aim of the study is to provide an in-depth understanding of the attitudes to and incidence, determinants and consequences of domestic violence in Ghana. The study addresses the following questions: 1. What types and levels of violence are experienced by women, men, girls and boys in Ghana, and who perpetrates them? What proportion of this violence is domestic violence? 2. What are society, community and people’s attitudes towards domestic violence, and how do attitudes and norms influence the acceptance of different forms of domestic violence in Ghana? 3. What are the main determinants of domestic violence against women, men, girls and boys at the individual, family, community and wider societal levels? 4. What are the consequences of domestic violence in Ghana for people’s physical and mental health and their daily lives? 5. What institutional structures support victims of domestic violence, and what proportion of victims are able to effectively access available services? 6. How can policies and interventions be (re)designed to better support victims of domestic violence and prevent domestic violence in Ghana? The study adopted a mixed-methods approach, which involved the collection of primary quantitative data through a representative household-level survey of 4,995 individuals – the Ghana Family Life and Health Survey (GFLHS) 2015 – and the collection of qualitative in-depth data on key mechanisms that may explain trends and the distribution of domestic violence outcomes in Ghana. Qualitative and quantitative methods were fully integrated to enable complementarity between the methods. The quantitative data enabled the production of statistics that are regionally and nationally representative at the household level, as well as an analysis of determinants and consequences of domestic violence. The qualitative data explored the perceptions, attitudes and beliefs of women and men within their everyday lives, how they are related to domestic violence experiences as survivors and perpetrators, and how individual experiences of domestic violence are integrated within family, community and wider social and political structures and relations. The next sections present the detailed findings of this study. Section 2 discusses the main definitions of domestic violence adopted in the study, and outlines the conceptual 25 framework of the report. Section 3 explains the qualitative and quantitative methodologies used in the study to collect nationally representative data on domestic violence in Ghana. Section 4 discusses in detail the main findings of the study. Section 5 summarises the key findings and discusses them in relation to previous studies conducted internationally and in Ghana. Section 6 concludes with proposals for future policy action to reduce and prevent domestic violence in Ghana. 26 2. Conceptual framework Five main sets of factors have historically been used to describe the determinants of domestic violence: individual psychological factors, sociological factors (such as family dysfunction), male aggression, poverty and the prevalence of cultures of violence (Bowman and Schneider, 1998; Schneider, 2000).10 International studies, largely conducted in the USA, have emphasised the role of individual psychology, interpersonal relationships within families and male aggression as the main explanations for the prevalence of domestic violence (Gosselin, 2000; Heath, 2014; Perilla, Bakeman and Norris, 1994; Pollak, 2004; Schneider, 2000; Walker, 1999). These individual- and family-centric approaches to domestic violence have been criticised for their lack of focus on factors such as poverty and culture – arguably more relevant in developing countries. They also tend to prioritise response mechanisms, such as psycho- and family therapy (Pollak, 2004), crime control and retroactive responses to cater to survivors’ needs (Kelly, 2003), which may not be useful in settings where medical and legal recourse may be prohibitively expensive and difficult to access (Mann, 2000; Murphy, 2002). Approaches that rely on expensive therapy to facilitate transformation may not, in addition, account for the deeper structural inequalities that have been shown to contribute to domestic violence (Maguigan, 2002; Schneider, 2008). As a response, feminist theories have attempted to analyse domestic violence as an outcome of broader contexts characterised by unequal gender relations (Bowman, 2002). This approach has been prolific in studies on domestic violence in Africa (Banda, 1995; also see Adjetey, 1994; Bell and Mattis, 2000; Bograd, 1999; Butegwa, 1994; Koenig et al., 2003; Perilla et al., 1994; Sokoloff and Dupont, 2005; Tamale, 1993; Turshen and Turshen, 2000), including one of the first domestic violence studies in Ghana by Rosemary Ofei-Aboagye (Ofei-Aboagye, 1994a; also see Akpalu et al., 2000). These studies have found that domestic violence against women is largely explained by women’s subordination to, and economic dependence on, men. Feminist theories are closely linked to cultural approaches to domestic violence. Among these, several studies have found that some aspects of domestic violence – for instance, wife-battering – are considered ‘normal’ within particular cultures. As an example, a study in Nigeria revealed that police officers tended to remind wives who reported physical violence that their ‘Yoruba culture’ permits men to beat women (Atinmo, 1997; Atinmo, 2001). Other less direct examples of ‘culture’ as an explanation for domestic violence include the widespread practice of bride price, and gender norms that limit women’s 10 The study was informed by a comprehensive literature review and secondary data analysis. Electronic key word searches were initially carried out in the English language in Science Direct, Scopus and Google Scholar. Subsequently, searches for regional and international policy documents were conducted using the United Nations and the African Commission websites. Thereafter, an extended search was done with academic publications by following ‘related’ or ‘cited by’ articles offered on publishing journals’ websites and Google Scholar. Studies that could not be obtained online or as hard copies in the United Kingdom were collected during visits to Accra. 27 access to work outside their home (Abane, 2003; Armstrong, 1998). For instance, research in Zimbabwe has found that asking for money from male partners led to physical violence against women, as the request was interpreted as a challenge to men’s position of power (Armstrong, 1998). Similarly, in Ghana, Amoakohene (2004: 2378) has analysed how women are expected “to be ‘submissive to [their] husbands’, ‘be respectful’, ‘be patient’, and ‘be dutiful and serviceable’, as tradition demands […]”. A related approach has argued that domestic violence is explained (at least partially) by the shift from traditional to modern, urbanised societies, a process that may result in some men feeling threatened and more likely to find ways of asserting their power through violence (Bowman, 2002). Abane (2003) has found that socio-economic change in Ghana has been accompanied by increasing economic strain within households (see also Armstrong, 1998). As a result, some men faced increased difficulties in supporting multiple wives and households, while women took on waged employment to subsidise the family’s income. However, women were expected to do both household and paid work. If they failed to meet their partner’s expectations, they could be beaten. Amoakohene (2004), too, has found that a change from traditional gender roles, where women worked at home, to new roles, where women took on waged employment, precipitated increases in domestic violence. International studies have revealed similar results (Justino et al., 2012). These approaches have resulted in important insights into the causes and consequences of domestic violence. However, most emphasise a unidimensional perspective, either focusing on the individual, on immediate social relations or on wider social and political processes. Structuralist theories, in contrast, have attempted to identify how drivers and responses to domestic violence may be shaped simultaneously by social, political, economic and cultural structures. Historically, structural approaches were championed by some feminists in the 1970s and 1980s who, in line with CEDAW and DEVAW, argued that domestic violence was not a private matter between individuals in the hidden spaces of their homes. Domestic violence should, rather, be understood as a public issue shaped by state and community actors and institutions that privilege certain gender and sexual norms over others (Bowman and Schneider, 1998). Within this theoretical perspective, the social ecological model (Bronfenbrenner, 1979; Carlson, 1984; Heise, 1998; Heise and Garcia-Moreno, 2002) offered a conceptual approach that integrated many of the above theories into a structural model that explored domestic violence across individual, societal, economic and political spheres. The model outlined four key levels of interaction across public and private spaces: the individual (personal history), the family (ecosystem), the community (microsystem) and the social and political context (macrosystem). Subsequent critiques led to the refinement of the original model. The first critique related to the model presenting the four layers as impermeable and unrelated to one other. However, studies in Brazil (Minayo, 1994), South Africa (Jewkes et al., 2002) and the USA (Perilla et al., 1994) highlighted the extent to which the experiences and personal histories 28 of the individual were strongly entwined with the ‘ecosystem’, their workplaces and neighbourhoods. The second critique originated from post-structuralist and post-colonial theorists who emphasised the importance of seeing individuals and structures in a circular dynamic, where people have constrained agency within structures that they both reproduce and challenge (Bourdieu, 2001). These theories resonated with post-colonial critiques, particularly by African feminists (Jones, 2011; Coughtry, 2011; Ewelukwa, 2002; Tsikata, 2014). These argued that both individualist and social ecological approaches were ahistorical and Western-centric. Furthermore, they failed to account for the role of violently unequal economic and social systems – such as colonialism, apartheid and globalisation – in entrenching intersecting forms of violence, including domestic violence, which were linked to class, race, sex, sexuality, age, location and so forth (Africa, 2010; Cornwall et al., 2007; Jones, 2011; McFadden, 2000). These critiques point to the importance of a conceptual framework that considers through an explicit political economy lens the complex interplay between the individual, the family, the community and the wider political and social system in Ghana. This study has, therefore, adopted a ‘critical social ecological’ approach as an amalgamation of the original social ecological theory adapted by Carlson (1984), Heise (1998) and others to domestic violence, and the extensions proposed by Jewkes (2002). At a general level, this integrated approach requires an understanding of individual experiences of domestic violence within family relations, the immediate political and economic environment, local institutions (such as the police and health care providers) and community attitudes towards violence, and the current and historical national policy context and political and economic environment. These four spheres and their interactions are outlined below. The individual. Personal circumstances help to understand the attitudes, incidence, determinants and consequences of domestic violence. As discussed above, domestic violence often goes unrecognised, unreported and denied due to the reluctance of individuals to report ‘private’ matters. One important example is the extent of sexual abuse within marital relationships (Adomako Ampofo, 2008). Other examples include human trafficking, coerced sex work, child marriage, widow inheritance and female genital mutilation (FGM). The international literature on domestic violence has shown that personal histories contribute to experiencing or enacting violence. Examples include witnessing violence as a child, having experienced abuse as a child or having had an absent father (García-Moreno et al., 2005; WHO, 2000; UNICEF, 2002). Based on these findings, this study was designed to explore people’s experiences with domestic violence over time, through survey questions, focus group discussions and in-depth interviews. These methods have been able to identify people’s personal experiences of domestic violence, their attitudes towards domestic violence, and their family and social relationships that may account for these experiences and attitudes – from the perspectives of both survivors and perpetrators. The family. The family is the immediate environment in which people who experience or exert violence live. In this sphere, factors that contribute to domestic violence include male 29 dominance and male control over the family’s wealth. These factors may also work conversely: if men do not hold wealth and women do, women may be subject to violence by her partner, or even by her partner’s family (Armstrong, 1998; Geisler, 1995; Njovana and Watts, 1996). The 2008 Ghana DHS study found that the incidence of domestic violence was associated with a complex variety of causes within the nuclear and extended family network. The results revealed that 36.6 per cent of women aged 15–49 years experienced physical violence, mostly perpetrated within the family by current or previous husbands or partners, as well as by other male and female family members or acquaintances (GSS et al., 2009). The current study, therefore, incorporated methodologies to collect data on and analyse domestic violence from this family perspective. To this purpose, the survey questionnaire included several questions on the determinants of violence within the domestic sphere, and about social norms and expectations around gender roles within immediate and wider family circles. The qualitative instruments were designed to examine personal experiences within families, by allowing interviewees to tell personal stories about relevant events, and explore different elements in the story to understand what happened and why. The community. This level of analysis takes into account the institutions and social structures that surround the family, such as the neighbourhood, social networks and the workplace. Human perceptions and attitudes do not emerge in a vacuum, and communities perform important functions in the lives of individuals and their families. Notably, the attitudes of community members, traditional and religious leaders, neighbours, work groups and other local social groups and networks are likely to shape the attitudes of individuals and their families, including the acceptance or tolerance of domestic violence. In Ghana, key community figures include community leaders and local representatives of state agencies, such as police officers and health care workers. This current study has paid careful attention to the analysis of how attitudes and norms towards domestic violence are shaped by wider community-level social, economic and political dynamics. The surveys for this study were designed to capture different types of violence experienced within the environment surrounding an individual (notably, the workplace and the community). The survey and qualitative instruments also included questions on attitudes and norms about the acceptability of domestic violence, about appropriate reactions and potential punishments, and about beliefs that may affect people’s ability to take cases of domestic violence outside the private sphere and to involve state actors, such as the police. The wider social and political context. Social and political contexts shape general views and attitudes shared across societies. The original social ecological model applied to domestic violence paid limited attention to these wider political or economic structures. Later work by Jewkes et al. (2002) revealed how political and social institutions engage, reproduce and challenge gender inequalities and domestic violence, and how economic inequalities affect people’s experiences of domestic violence. Key political and social institutions include the police, law courts, legal services, and health and social support organisations. This is because domestic violence may go unchecked if not accompanied by 30 integrated prevention and oversight through the State’s law enforcement bodies. Domestic violence may also be unreported if it is socially accepted across the wider society. One objective of the current study was to uncover how institutions set up to respond to and prevent domestic violence in Ghana may have shaped the incidence, nature, determinants and consequences of domestic violence across the individual, their families, their communities and the wider society. Quantitative and qualitative instruments have measured differences in knowledge, availability, accessibility, use and satisfaction with different services related to domestic violence. Information was also collected about views and attitudes across the population that may explain relations between institutional engagement, structural inequalities, gender inequality and domestic violence. The qualitative instruments, in particular, collected data on whether people were aware of available services, the barriers they might have encountered when accessing them, and suggestions to support individuals and communities to hold local and national government accountable to the provision of services listed in the 2007 Domestic Violence Act. The conceptual framework highlights the importance of understanding domestic violence both within and across the different spheres that shape the incidence, distribution and nature of domestic violence: the individual, their immediate social relations, the community and the wider social and political system. The study used this conceptual framework to collect new empirical evidence across Ghana on: (i) the incidence and types of domestic violence; (ii) individual-, household- and community-level attitudes towards domestic violence; (iii) determinants of domestic violence; (iv) consequences of domestic violence; and (v) the use of and access to victims’ support services and institutions. Subsection 2.1 discusses the main definitions and concepts used in the study. Subsections 2.2, 2.3, 2.4 and 2.5 describe, respectively, the current study’s approach to the collection and analysis of information on attitudes to and determinants and consequences of domestic violence, and how people access and use relevant services and institutions across Ghana. 2.1. Types and perpetration of domestic violence The most prominent international studies on domestic violence are of a quantitative nature, based on national demographic data and health surveys (e.g. UNICEF, 2002; WHO, 2000). These typically measure physical, sexual and psychological abuse by using a set of questions about particular acts of violence and asking respondents to state whether they have ever experienced them, and if so, whether this happened in the 12 months prior to the survey. These studies also usually investigate attitudes towards physical and sexual abuse by men and women. Studies prior to 2008 in Ghana were of a more exploratory nature but included very useful information about context-specific and normative understandings of what constituted domestic violence. The first study on the nature and extent of domestic violence in Ghana, by Rosemary Ofei-Aboagye (1994), focused on women’s understandings and responses to 31 domestic violence. The study noted, in particular, the importance of terminology: only five out of 50 women said it was accepted in their ethnic group that husbands beat their wives; but many more did so when the term ‘disciplining’ was used. A second study by the Gender Centre was designed to better understand the prevalence, type, contexts, responses, barriers to responses and recommendations about domestic violence at the national level (Coker-Appiah and Cusack, 1999). Similar to Ofei-Aboagye’s 1994 study, the researchers asked women to define what they understood as violence. The study revealed that most women understood that violence was not limited to physical abuse but also included acts of psychological mistreatment such as threats, humiliations, being prevented from seeing family and friends and sexual harassment. A later study by Amoakohene (2004) analysed domestic violence from a more multidimensional perspective, by exploring women’s perceptions about their rights, responsibilities, duties and the abuses they suffer, as well as the coping strategies they developed. Domestic violence was seen to include both physical and psychological abuse. Sexual violence was not recognised as an offence by either the State or the community (see also Adu-Gyamfi, 2014). These findings, alongside the definition of domestic violence proposed in the 2007 Domestic Violence Act, have informed the typology of domestic violence used in this study. Notably, the 2007 Domestic Violence Act defines domestic violence as: “engaging in the following within the context of a previous or existing domestic relationship: (a) an act under the Criminal Code 1960 (Act 29) which constitutes a threat or harm to a person under that Act; (b) specific acts, threats to commit, or acts likely to result in (i) physical abuse, namely physical assault or use of physical force against another person including the forcible confinement or detention of another person and the deprivation of another person of access to adequate food, water, clothing, shelter, rest, or subjecting another person to torture or other cruel, inhuman or degrading treatment or punishment; (ii) sexual abuse, namely the forceful engagement of another person in a sexual contact which includes sexual conduct that abuses, humiliates or degrades the other person or otherwise violates another person's sexual integrity or a sexual contact by a person aware of being infected with human immunodeficiency virus (HIV) or any other sexually transmitted disease with another person without that other person being given prior information of the infection; (iii) economic abuse, namely the deprivation or threatened deprivation of economic or financial resources which a person is entitled to by law, the disposition or threatened disposition of moveable or immovable property in which another person has a material interest and hiding or hindering the use of property or damaging or destroying property in which another person has a material interest; and (iv) emotional, verbal or psychological abuse namely any conduct that makes another person feel constantly unhappy, miserable, humiliated, ridiculed, afraid, jittery or depressed or to feel inadequate or worthless; (c) harassment including sexual harassment and intimidation by inducing fear in another person; and (d) behaviour or conduct that in any way 32 (i) harms or may harm another person, (ii) endangers the safety, health or well-being of another person, (iii) undermines another person's privacy, integrity or security, or (iv) detracts or is likely to detract from another person's dignity and worth as a human being. Following this definition, the study has collected detailed information on all categories of physical, sexual, economic and psychological abuse. In line with prior studies conducted in Ghana and information provided to the research team in initial scoping interviews with key stakeholders, the study also collected information on acts of controlling behaviour (social violence). Detailed definitions of the types of violence are provided in Section 4. Conforming to international definitions (e.g. UNICEF, 2002; WHO, 2000), the Domestic Violence Act, and prior studies conducted in Ghana, domestic violence is defined in this study as violence committed against a person by an intimate partner, including co-habiting partners, and by other family members, irrespective of whether this violence occurs within or beyond the confines of the home. Ghana’s Domestic Violence Act provides a broad definition of what constitutes a ‘domestic relation’, which has been adopted in this study. ‘Domestic relation’ means “a family relationship, a relationship akin to a family relationship or a relationship in a domestic situation that exists or has existed between a complainant and a respondent and includes a relationship where the complainant (a) is or has been married to the respondent; (b) lives with the respondent in a relationship in the nature of a marriage even if they are not or were not married to each other or could not or cannot be married to each other; (c) is engaged to the respondent, courting the respondent or is in an actual or perceived romantic, intimate, or cordial relationship not necessarily including a sexual relationship with the respondent; (d) and respondent are parents of a child, are expecting a child together or are foster parents of a child; (f) and respondent are family members related by blood ties, affinity or adoption, or would be so related if they were married either customarily or under an enactment or were able to be married or if they were living together as spouses although they are not married; (g) and respondent share or shared the same residence or are cotenants; (h) is a parent, an elderly blood relation or is an elderly person who is by law a relation of the respondent; (i) is a house help in the household of the respondent; or (j) is in a relationship determined by the court to be a domestic relationship”. To distinguish whether acts of violence are domestic or non-domestic in nature, respondents who reported having experienced any one of the acts of violence above during the 12 months prior to the survey were asked whether the perpetrator was part, or not, of the household. If the perpetrator and survivor shared the same roof and meals (the definition of a household adopted in this study, following the standard definition adopted by GSS), the respondent was asked to identify the perpetrator. In cases where the perpetrator was identified as living outside the household, respondents were asked about their relationship to the perpetrator. Answer categories included (former) partners, extended family members living within or outside the same community, non-related 33 community members, teachers, colleagues, friends and other relations. This choice of questions allowed the study to identify which violence was domestic in nature, as well as to frame experiences of violence within the wider community and societal contexts. One innovative aspect of the 2007 Domestic Violence Act is that its definition of domestic violence does not refer exclusively to women. Although domestic violence is generally disproportionately directed at women, some studies – almost exclusively based on data from the developed world – have argued that this is not always the case: women can be equally violent, and children also ought to be considered potential victims of domestic violence (Hunnicutt, 2009). The DHS 2008 study conducted in Ghana was one of the few studies to date to include a domestic violence module that elicited men’s experiences. Over a quarter (27.6 per cent) of ever-married Ghanaian men aged 15–49 years in the 2008 DHS reported having experienced physical or psychological violence by their wife or partner at some point during their lives (GSS et al., 2009). Using the same data, Kishor and Bradley (2012) found that 19 per cent of married women and 10 per cent of married men had experienced physical, psychological or sexual spousal violence, confirming that it is not only women who are victims of intimate partner violence.11 In addition, the Domestic Violence Victims Support Unit (DOVVSU) reported to the research team during interviews in Ghana that it had processed several cases reported by men and boys. At the international level, similar evidence is reported in IDS et al. (2015).12 This study has attempted to uncover the experiences of women and girls, as well as those of men and boys. It is also important to note that different types of domestic violence and reporting behaviour of domestic violence vary across different age groups, regions and socio-economic characteristics (Coker-Appiah and Cusack, 1999; García-Moreno et al., 2005; GSS et al., 2009). The study will address these differences in detail in Section 4. 2.2. Attitudes towards domestic violence Similar to other social attitudes, attitudes towards domestic violence depend on a range of factors. These include individuals’ beliefs about gender norms and norms that regulate the use of violence; how such beliefs are internalised; social norms that regulate the distribution of power and resources within the household and immediate family; and social norms that shape gender roles and behaviours at the community and wider societal levels (Justino, Leone and Salardi, 2015). Multi-country studies and international comparisons have shown that women’s attitudes towards violence vary substantially across settings. Furthermore, women living in countries where domestic violence is more prevalent tend to find domestic violence more acceptable than women in countries where prevalence is lower (García-Moreno et al., 2005; Appendix A). The acceptance of wife- 11 However, the authors show that women are consistently and significantly more likely than men to experience all forms (physical, sexual and emotional) of such violence, and that intimate partner violence is more commonly experienced by women. It is also more severe and more likely to result in injuries than intimate partner violence experienced by men. 12 See also the research programme on Men, Boys and Gender Equality at the Institute of Development Studies: http://menandboys.ids.ac.uk/. 34 beating also tends to be higher among women who have experienced abuse than among those who have not (García-Moreno et al., 2005). Taken together, these findings suggest that women seem to learn to accept violence when they are victims themselves – a result that will be explored further in the study. Ofei-Aboagye’s (1994) study asked respondents about the level of beating that was perceived to be acceptable. The findings showed that one to three slaps were considered acceptable. However, the study also found that there was a threshold beyond which beating ‘transcended the norm’. This happened when the victim was permanently hurt, scarred or deformed, or was beaten in public. A later study by Takyi and Mann (2006) found that 40.5 per cent of never-married and 33 per cent of ever-married men perceived wife-beating to be justified. The most justifiable reason given was when they perceived that women neglected their children. Using data from the 2008 DHS, Mann and Takyi (2009) found that women were more likely than men to believe that spousal violence against them was justified: 64.2 per cent of men stated that abuse would never be justified, but only 44.6 per cent of women agreed (see also Kishor and Bradley, 2012; Tenkorang et al., 2013). These findings raise important questions about the prevalence of social norms and beliefs that may sustain domestic violence – an issue that will be explored in detail in Section 4. It appears that wife-beating may be becoming less accepted. Ghana DHS statistics for 2003 and 2008 show that, across almost all regions, the percentage of respondents condoning wife-beating decreased. The reductions were up to 35.7 percentage points among men, and up to 45.9 percentage points among women, with the largest changes taking place in the Upper East and Upper West regions. These were also regions with a traditionally high prevalence of domestic violence, and where awareness-raising campaigns have concentrated their sensitisation efforts. This study will explore how these trends may have continued until 2015 and what social norms and changes in social norms may explain them. 2.3. Determinants of domestic violence The existing literature on domestic violence has found evidence for strong correlations between the incidence of domestic violence and a wide range of individual socio-economic factors, such as age, education and marital status (Aizer, 2011; García-Moreno et al., 2005; Jewkes et al., 2002; La Mattina, 2012). In line with the conceptual framework outlined above, evidence shows that these individual determinants of domestic violence interact with factors that play out at the level of families (notably, differences in intra-household gender dynamics), as well as wider community and societal levels – in particular, the prevalence of patriarchal norms, perceptions of masculinity and exposure to other forms of violence. These determinants of domestic violence, and how the study was designed to capture them, are discussed below. Age and intergenerational effects. Several studies have shown a close association between the incidence of domestic violence and the age of victims and perpetrators, with, in general, 35 incidence being lower among older women over their lifetime and in the recall period of the surveys (Aizer, 2011). For example, in urban Peru, 41 per cent of women aged 15–19 years old reported having experienced intimate partner violence, in comparison to 8 per cent among women aged 45–49 years old (García-Moreno et al., 2005). The research team conducted a background analysis of the Ghana DHS 2008 for this study. In contrast to international evidence, the results did not show any particular age patterns among victims of domestic violence. The data indicated, however, that the incidence of domestic violence in Ghana was higher among women and men whose father beat their mother. Several other studies have found exposure to domestic violence during childhood to be positively associated with the probability of being a victim of domestic violence as an adult (Clark et al., 2010; Gallegos and Gutierrez, 2011; Bobonis et al., 2013). These findings are consistent with theories of intergenerational transmission of domestic violence (Pollak, 2004). This study will investigate further the distribution of the incidence of domestic violence across age and generations in Section 4. Marital status. Qualitative research has shown that, in many parts of the world, married men are assumed to have unconditional sexual access to their wives (Ellsberg et al., 2001). In the context of Ghana, Adu-Gyamfi (2014) has reported how attitudes, beliefs and practices have led to violent actions against women being dismissed due to widespread beliefs that women give up their rights to their husbands at the point of marriage, particularly when bride prices are paid. In the international literature, marital status is also an important factor associated with domestic violence. Women who are separated or divorced tend to report higher incidence of domestic violence than those who are currently married, with violence persisting after the divorce or separation (García-Moreno et al., 2005). The timing of marriage also appears to be an important determinant of domestic violence, with the incidence of domestic violence lower among victims who marry later (La Mattina, 2012). The background analysis of the 2008 DHS data conducted for this study showed that the incidence of domestic violence in Ghana was lower among currently married individuals (women and men) than among those who were formerly married and separated or divorced afterwards. The role of marital status on domestic violence in Ghana in 2015 will be explored further in Section 4. Socio-economic status and education levels. Though domestic violence affects everyone, existing literature indicates that the incidence of domestic violence is usually associated with low income and wealth levels (Bobonis et al., 2013). For instance, Aizer (2011) showed that poor women were at a higher risk of abuse in the USA. In contrast, Amoakohene (2004) reported domestic violence in Ghana to be widespread among all socio-economic classes, with 70 per cent of educated and employed respondents in her sample reporting some form of abuse (predominantly physical). The study by Adinkrah (2014) showed that homicide-suicides occurred across the whole socio-economic spectrum but were more prevalent among individuals with lower socio-economic background. Education levels are also typically associated with incidence of domestic violence. As argued in Jewkes et al. (2002), “education confers on individuals social empowerment, self-confidence, and the ability to use information and resources to one’s advantage”. Aizer (2011) found that less 36 educated women were at a higher risk of abuse in the USA, while Panda and Agarwal (2005) showed that women who owned land in Kerala (India) were less likely to experience forms of domestic violence. García-Moreno et al. (2005) reported that in many of the countries they analysed the prevalence of domestic violence decreased with rising education levels. However, in some of the countries (Brazil, Namibia, Peru, Thailand and Tanzania) the prevalence of domestic violence was lower only among women with education levels above secondary school. This may imply that, at higher levels of education, women acquire a greater capacity to negotiate autonomy and control of resources within their marriage or families. Consistent with these findings, Tenkorang et al. (2013) have shown that women with higher levels of education in Ghana in 2008 were less likely than women with lower levels of education to experience physical violence. Yeboah and Batse (2009) found some level of schooling (primary education) to be correlated with higher risks of psychological violence. Education has also been found to explain attitudes towards domestic violence: Takyi and Mann (2006) reported that younger men and men with little education or social status are more likely than other socio- economic groups to believe that wife-beating is justified. According to the 2008 Ghana DHS, the incidence of domestic violence in Ghana was lower among individuals with no education than among those with primary or higher levels of education. La Mattina (2012) found a similar result in Rwanda. These results could suggest that individuals who are more educated may also be more likely to be aware of existing laws against domestic violence and to report being victims of domestic violence. The 2008 data for Ghana showed that the percentage of individuals who condoned wife-beating was higher among those with no education than among those with some level of education. The role of socio-economic status, and education in particular, will be further explored in Section 4. Intra-household gender dynamics. Several studies have attributed changes in rates of domestic violence to changes in gender dynamics within the household. A corollary of the feminist theories discussed in the previous section is that women who have more autonomy and earn independent incomes may be less likely than other women to experience domestic violence (see Goode, 1971; Dyson and Moore, 1983). For instance, Aizer (2010) showed that increases in women’s relative wages and in women’s labour force participation resulted in a lower incidence of domestic violence in the USA. In contrast, others have noted that newly economically empowered women may threaten men’s identity, resulting in increases in domestic violence. In these cases, husbands may resort to violence to reassert their sense of power (Heath, 2014; Menon and Johnson, 2007; Tanchen, Witte and Long, 1991). Consistent with these predictions, studies have found that increases in women’s labour force opportunities and their access to income and other resources have at times been associated with increases in domestic violence. This is particularly true if women’s initial level of bargaining power before entering the labour market was low (Bobonis et al., 2013; Heath, 2014; Hidrobo and Fernald, 2013; Justino et al. 2012). 37 In the case of Ghana, Mann and Takyi (2009) have shown that the likelihood of an individual tolerating domestic violence was 50 per cent higher among men who contributed less than half of the household budget (when compared to men who contributed to the totality of the household budget). Interestingly, when men and women were equals in terms of decision-making, men were more likely than women to believe that abusing women was not justifiable. The background analysis conducted in this study using the 2008 DHS data also showed that the incidence of domestic violence was lower when men and women made decisions jointly. The incidence of domestic violence was higher among women who made decisions about household expenses alone. These patterns will be explored further for the 2015 data in Section 4. Patriarchal norms. Several theories have proposed patriarchy as an explanation for the prevalence of domestic violence, whereby violence is a way in which men exercise control over women (Dobash and Dobash, 1979; Eswaran and Malthotra, 2011; Martin, 1976; Wilson and Daly, 1993, 1996; Yllo and Strauss, 1990). A number of studies have reinforced the notion of Ghana as a patriarchal country. Ofei-Aboagye (1994), for example, described how in proverbs, oral traditions and folklore, women were celebrated for their ‘life-giving capacities’. The duty of husbands, on the other hand, was to secure the obedience and fidelity of the wife. Cultural traditions, such as wife inheritance and dowry payments, forced marriages, widowhood rites and FGM, have reinforced the belief in men’s superiority and encouraged violence against women (Adu-Gyamfi, 2014; Amoah, 2007; Amoakohene, 2004; Adomako Ampofo, 1993; Tenkorang et al., 2013). This is particularly true when women are dependent on men because of limited education and training opportunities, and restricted access to cash, credit and employment (Amoakohene, 2004; UNICEF, 2000). Mann and Takyi (2009) found further that respondents from matrilineal groups were less likely to condone spousal abuse than those from patrilineal backgrounds (see also Panda and Agarwal, 2005). This study will investigate further the role of patriarchal norms as a determinant of domestic violence through the analysis of specific questions included in the qualitative and quantitative instruments to uncover how normative values at the community and society levels may affect individual and household behaviours and attitudes towards domestic violence. Masculinities. The way men express masculinity in multiple and, sometimes, contradictory ways has been linked to gender-based violence (Edström et al., 2014a, 2014b; Freedman and Jacobson, 2012). The focus on masculinities since the late 1980s has been due to a number of factors with significance for sub-Saharan African countries, including the emergence of sexually transmitted diseases such as HIV (Barker and Ricardo, 2005; Eves, 2010; Harrison et al., 2006; Mills et al., 2009), resource-scarcity and conflict-related migration (Campbell, 1997; Jones, 2006; Cornwall and White, 2000; Little and Jones, 2000), and social and economic crises that trigger crises of masculinities (Richards, 1996, 2005). The ‘crisis of masculinities’ literature, however, has been recently challenged. Studies in Sierra Leone, South Africa and Kenya, among other countries, have emphasised the importance of fostering positive masculinities among men to encourage them to challenge harmful patriarchal norms and collectively address gender-based violence (Mills et al., 38 2015a, 2015b). In Ghana, Mann and Takyi's (2009) study found that expectations of masculinity that focused on men’s financial responsibility to provide for their households was linked to men believing that intimate partner violence was acceptable. The pressures placed on men, through these and other harmful constructions of masculinity, not only reinforce harmful gender dynamics and violence but have also been found to contribute to men’s ill health. For instance, Adinkrah (2012) investigated male suicide behaviour in Ghana using data from the Statistics and Information Technology Unit of the Ghana Police Service. The study found that over half of the 287 (attempted) suicides covered in the data were motivated by an attempt to avoid the dishonour, humiliation, public disgrace and shame associated with economic failure, disease and physical conditions, deviant or criminal conduct and sexual impotence. In a follow-up study, Adinkrah (2014) focused on homicide-suicides. The majority of the 50 cases analysed between 1990 and 2009 showed a particularly severe form of intimate partner violence: they involved husbands suspecting infidelity and men whose wives threatened them with divorce or separation. Studies on masculinities, both in and beyond Ghana, therefore, show the harmful effects of negative masculinities on both men and women. Emerging studies have also revealed that, as a socially constructed phenomenon, masculinities can be positively harnessed to transform detrimental gender dynamics (Mills et al., 2015a, 2015b). Exposure to other forms of violence. A smaller literature has examined how domestic violence may be related to other forms of violence prevalent within the wider society. Recent evidence has shown that exposure to a violent conflict or civil war at some point of an individual’s life could be associated with rises in domestic violence. La Mattina (2012) found that exposure to the 1994 Rwandan genocide led to an increase in domestic violence against women who married after the genocide. Gallegos and Gutierrez (2011) found a positive and large effect of conflict exposure in Peru on the incidence of domestic violence, particularly when conflict exposure took place during late childhood and teenage years. The study reported, in addition, that women who were exposed to civil conflict at a younger age were more likely to justify wife-beating. Noe and Rieckmann (2013) showed that women living in districts in Colombia with high conflict intensity had a higher probability of being a victim of domestic violence than those living in districts with low conflict intensity. These results may be due to post-traumatic stress disorder or desensitisation caused by exposure to wider violence. This literature is only in its infancy. This study has attempted to analyse further this interesting hypothesis through questions in the qualitative and quantitative instruments on experiences of violence, crime and safety in and outside the communities sampled. Results are discussed in Section 4. 2.4. Consequences of domestic violence The persistence of domestic violence is not only a violation of basic human rights but also restricts the achievement of gender equality across a variety of social, economic and political outcomes. Research on the consequences of domestic violence is, however, 39 limited. The few studies that have analysed the consequences of domestic violence show that domestic violence is typically associated with a long list of adverse consequences – notably on physical and mental health, and on the capacity of women and men exposed to violence to carry on with their work and other daily activities. The literature has also reported large direct and indirect effects of domestic violence within the household on children’s physical and emotional well-being and future life prospects. Physical and mental health. The most widely reported consequences of domestic violence are poor physical and mental health. Adverse effects include physical injuries, pregnancy complications, depression and higher risk of contracting HIV and other sexually transmitted diseases (García-Moreno et al., 2005; Moosa, 2012). García-Moreno et al. (2013) reported that women who had experienced physical or sexual abuse by their partners were 16 per cent more likely to give birth to low-weight babies than women who did not experience intimate partner violence. They were also almost twice as likely to experience depression, up to 1.5 times more likely to contract HIV, and more than twice as likely to have an abortion. García-Moreno et al. (2005) showed that domestic violence has particularly severe consequences on mental health: women who had ever experienced physical or sexual violence by an intimate partner reported significantly higher levels of emotional distress,13 and were more likely to have thought about suicide or have attempted to commit suicide. In the case of Ghana, Amoakohene (2004) reported widespread adverse psychological and emotional effects as a result of domestic violence: women in the study mentioned being in constant fear when their husbands came home, depressed over their lack of freedom, and suffering from stress, tension and low self-esteem. Three other recent studies from Ghana showed harmful effects of domestic violence on taking pregnancy to full term and the health of babies born (Pool et al., 2014), between domestic violence and health status (Addai and Adjei, 2014), and between domestic violence and mental health (Sipsma et al., 2013). Adu-Gyamfi (2014) reported similar non-physical side effects of domestic violence, such as an erosion of a person’s sense of dignity, self-esteem, confidence and respect, particularly when injuries make it difficult for victims to go out in public for fear of being ‘found out’. Work and daily activities. Some studies have noted a close association between experiences of domestic violence and restricted access to livelihood options, fewer skills and lower productivity among victims (García-Moreno et al., 2005; Moosa, 2012). Lloyd (1997) found that female victims of domestic violence in the USA were more likely to have been unemployed, had multiple jobs, had lower personal incomes and were significantly more likely to receive public assistance than women who did not experience domestic violence. Similar results were found in Friedman and Couper (1987) and Shepard and Pence (1988), also for the USA. In Colombia, Ribero and Sanchez (2005) showed that domestic violence against women caused increased unemployment (by 6.4 percentage points), a 40 per cent reduction in earnings, and adverse health outcomes. In Ghana, Britwum and Cusack (2009) found that individuals who experienced domestic violence 13 Emotional distress was identified through symptoms such as crying easily, inability to enjoy life, fatigue, and thoughts of suicide in the four weeks prior to the interview (García-Moreno et al., 2005). 40 took time off work as a direct effect of the violence.14 Duvvury, Minh and Carney (2012) showed that domestic violence also affects men, with men losing about 6.5 days of work per incident of violence. Eswaran and Malthotra (2011) found a negative effect of domestic violence on women’s autonomy, while Bloch and Rao (2002) showed that domestic violence was a means used by perpetrators (males in their study) in India to ensure the maintenance of their bargaining power within the household. Child development and behaviour. Domestic violence within families is associated with direct and indirect effects on children. The most prominent consequences include lower birth weight, lower IQ scores and increased emotional and behavioural problems (Aizer, 2011; Koenen et al., 2003; Sternberg et al., 1993; Wolfe et al., 2003). Carrell and Hoekstra (2010) have also shown that children who suffered from social and emotional problems related to exposure to domestic violence also experienced lower academic achievement. To capture the consequences of domestic violence for individuals, their families and their communities, the quantitative survey conducted in this study included questions about the immediate consequences of domestic violence, particularly physical and sexual violence. Questions elicited information about whether and how violence influenced men and women’s ability to work, attend school or provide childcare, as well as about health and psychological consequences, including fear and confidence. This information was complemented by a number of interviews and focus group discussions designed to understand the consequences of domestic violence at the level of the individual, the household and the community. Specific questions addressed social relations, the well- being of children, and levels of cohesion within households and communities, including tensions that may result from broken marriages and the escalation of conflicts beyond the immediate relationship between victim and perpetrator. 2.5. Use of and access to victims’ support services and institutions The final research question addressed by this study asks how men and women affected by domestic violence use and access support services and institutions. Levels of domestic violence are typically under-reported across the world. García-Moreno et al. (2005) asked about the use of formal services across several countries (health services, legal advice and shelters), and whether victims of domestic violence asked for support from the police, women’s NGOs, local leaders or religious leaders. The majority of physically abused women reported that they had never approached any formal organisation or service. When asked about the reasons for (not) seeking help, most respondents mentioned a general acceptance of these behaviours as normal. Others feared consequences such as further violence, losing their children or bringing shame to their family. Seeking help – a limited occurrence in the sample – was usually associated with the severity of the violence, the 14 This study interviewed 1,588 women aged over 19 years and 481 girls aged 13–18 years. Nearly a third reported having experienced physical violence, a quarter had experienced psychological violence, and one fifth had experienced sexual violence at least once in their lives. 41 potential negative impact on children and encouragement from friends and family. Only victims of very severe acts of violence left their husbands. Similarly, Amoakohene (2004) found that victims of domestic violence in Ghana generally did not report abuse. Support was not always forthcoming when the victims approached relatives, the village chief or the police. The main explanation for these findings was the widespread view that domestic violence was normal. Other barriers that have prevented victims of domestic violence in Ghana from seeking help from outside the family included concern for the children’s economic and emotional well-being, lack of family support, lack of information or economic means to pursue legal rights and the belief that women have to obey their husbands (Coker-Appiah and Cusack, 1999; Ofei-Aboagye, 1994). Several institutions have been set up to respond to domestic violence in Ghana but are constrained by capacity and other logistical challenges (Mitchell, 2011). Challenges include the lack of initial and continuing training of staff, the slow pace of the courts and the lack of specific support services, such as counselling, emergency shelters, legal advice and medical services. Adu-Gyamfi (2014) analysed the process of prosecuting cases of domestic violence after the implementation of the 2007 Domestic Violence Act in the Mampong municipality of Ghana. The study found that the number of abuse reports increased significantly after the implementation of the Act but was not accompanied by increases in the number of prosecutions or convictions. According to the DOVVSU 2011 annual report, 12,706 cases of domestic violence were filed in 2010. Among these, only 954 cases were sent to court, resulting in only 118 convictions. One of the objectives of this study was to analyse how resources and services were used in Ghana to support victims of domestic violence, and how and when victims accessed them. There appears to be a consensus among service users about a lack of resources and poor implementation of relevant policies. The perceptions and experiences of authorities who work at these specialised services are less clear. This study designed a series of structured observations of these services (discussed in more detail in Section 3), to better understand the opportunities and barriers for domestic violence survivors to seek and receive support. To operationalise this framework, the study interviewed several individuals at different levels, including opinion leaders, activists, legal experts, community leaders (religious and political leaders, including traditional authorities), legal practitioners and journalists. The evidence above also suggests that under-reporting of domestic violence may be due to social normative pressures. This evidence connects to a growing body of feminist criminology literature that challenges the focus on the criminal justice system as the main avenue for punishing perpetrators of domestic violence (Buzawa and Buzawa, 2003; Rowe, 2015). These studies show that the apprehension and punishment of perpetrators through penal and criminal systems might offer short-term protection for women but does not effectively challenge patterns of violence at a societal or systemic level (Belknap, 2014; Hoyle, 1998; Sherman et al., 1992). Recent studies show, too, that victims of domestic violence, especially intimate partner violence, might forgo accessing the criminal justice 42 system altogether because their adjudication wishes (such as anonymity) are not met (Artz, 2014; Cerulli et al., 2015). Further, stigmatisation of those who report violence has also been found to discourage victims of domestic violence from accessing resources available through the criminal justice system (Sulak et al., 2014). The study will explore these factors, alongside other potential explanations such as people preferring other forms of intervention and mediation due to differing perceptions and trust about the justice system and other government-appointed support structures (lawyers, nurses, social workers and teachers, for instance), community leaders and the police. This approach connects to a current debate in the development literature on whether or not there is value in using ‘organic’ or ‘existing’ informal mechanisms to mediate individual-, household- and community-level conflict (Lugo, 2015; Rowe, 2015). Research in Rwanda and Uganda, for instance, has suggested that ‘organic’ responses foster greater levels of reconciliation among perpetrators and victims than imposed criminal justice routes (Doughty, 2014; Burgess and Campbell, 2015). This study will explore how men and women perceive government institutions and actors (such as police officers and health care providers). It will also explore how the police and other service providers view the help- seeking behaviour of citizens. This analysis will make use of a number of questions included in the formal survey about access to formal and informal services and support structures, use of these structures and perceptions about their usefulness and effectiveness. 43 3. Methodology Ghana is a culturally diverse country, and the incidence, causes and consequences of domestic violence are likely to vary across regions, broader kinship and socio-economic relations and individual characteristics. The study, therefore, adopted a mixed-methods approach to collect information about the incidence, nature, determinants and consequences of domestic violence. This approach involved the collection of primary quantitative data through a representative household-level survey and qualitative in- depth data on key mechanisms that may explain trends, patterns and the distribution of domestic violence in Ghana. The qualitative and quantitative instruments were designed to capture information related to the dimensions of the social ecological model – personal histories; individual and household circumstances and relations within; individual and community perceptions and realities of violence; access to, use of and quality of local resources and services at local and regional levels; and experiences of violence perpetrated by a range of possible actors. From the outset, the aim of the research was to ensure complementarity between the qualitative and quantitative research methods. The quantitative data enabled the production of statistics that are regionally and nationally representative at the household level. The qualitative data explored the perceptions, attitudes and beliefs of women and men within their everyday lives, to understand how they are related to experiences of domestic violence as survivors and victims and to observe how individual experiences of domestic violence are integrated within family, community and wider social and political structures and relations. The following sections describe the design of the quantitative survey (subsection 3.1) and the qualitative instruments (subsection 3.2); sampling procedures for the quantitative survey (subsection 3.3) and qualitative fieldwork (subsection 3.4); the training of enumerators and pre-testing processes (subsection 3.5); the implementation of fieldwork across all sites in Ghana (subsection 3.6); response rates (subsection 3.7); the methods used in the analysis of the qualitative and quantitative data (subsection 3.8); and the strict ethical procedures followed in the study (subsection 3.9). 3.1. Quantitative instruments The main quantitative instrument used in the study was a survey entitled the ‘Ghana Family Life and Health Survey 2015’ (GFLHS 2015). The survey deliberately did not include any mention of ‘domestic violence’ in the title, and was described to communities in generic terms as a survey of health and family life. Specific information about the survey was provided only when the research team had selected the participant within each household approached, and privacy was ensured. At that point, the survey was explained in detail to the participant, who was offered the option of refusing to continue with the interview. Few participants refused to continue (see subsection 3.3 below). This approach ensured high rates of response because participants were not prevented from answering the questionnaire by other household members worried about a survey on domestic violence. The introduction of the study as a health and family life survey also provided a 44 level of safety to the respondents and interviewers, by avoiding any immediate conflicts or tensions that may have arisen if the topic of domestic violence had been introduced publicly at the outset, as well as preventing the possibility of adverse follow-up consequences for the respondent. The GFLHS 2015 included one household- and one individual-level questionnaire. The design of the questionnaires was based on the World Health Organization (WHO) survey for the Study on Women’s Health and Violence against Women (Garcia-Moreno et al., 2005), DHS and Multiple Indicator Cluster Survey (MICS) domestic violence modules applied in several countries (see Appendix A), the 1998 study on Violence against Women and Children in Ghana (Coker-Appiah and Cusack, 1999) and modules used in surveys implemented by research team members elsewhere. The questionnaire design also built on the definition of domestic violence used in the 2007 Domestic Violence Act but included an additional category of domestic violence to capture instances of controlling behaviour (social violence). As discussed in the previous section, the definition of domestic violence adopted in this study is broader than definitions used in standard domestic violence surveys implemented to date. The questionnaires used in the GFLHS 2015 included questions that address this wider definition of domestic violence and, therefore, may not be directly comparable to existing surveys such as the DHS 2008 and the 1999 Gender Centre study (Coker-Appiah and Cusack, 1999). For instance, the DHS 2008 focused only on physical, sexual and psychological violence. Therefore, the total incidence of domestic violence captured in this study will not be directly comparable to the DHS findings. However, by adopting this broader definition of domestic violence and capturing several acts of violence committed within domestic relations but not captured before, this study provides for the first time a comprehensive analysis of how women and men in Ghana experience all forms of domestic violence, which can be used to fine-tune policy, programmes and monitor future progress. The household-level questionnaire was designed to capture information about all individuals in each sampled household. For the purpose of the study, the standard definition of a household used by the GSS was followed, which includes a group of individuals who normally live under one dwelling, and prepare and eat their meals together. The household survey included standard socio-economic questions about age, sex, education levels, children’s school attendance, occupation, work activities and related payments, religion and ethnicity. The household questionnaire also collected information on food consumption, household responses to economic shocks, asset ownership, decision- making processes within the household, and conflicts in the community. The individual-level questionnaire was administered to only one man or one woman in each sampled household, in line with WHO (2001) guidelines. This questionnaire included five modules on individual experiences of domestic violence from the perspective of both victims and perpetrators. The modules followed the definitions and typologies of domestic violence discussed in the previous section, and elicited information about experiences of 45 social, sexual, physical, psychological and economic violence. These modules were applied to all respondents and woven into the questionnaire between other modules to avoid repetition and respondent fatigue. Other additional modules included questions designed to capture the various dynamics, determinants and consequences of domestic violence across the four spheres of the social ecological model: safety in the community, social capital, relationships within and outside the household and local community, reproductive, physical and mental health, individual agency, access and use of services, and approaches to parental and child discipline. The individual questionnaire also contained a detailed module on attitudes towards domestic violence that made use of attitudinal questions, scenario questions and vignettes, as used, for instance, in the DHS domestic violence modules (García-Moreno et al., 2005; see also Thaler, 2012). The complete individual and household questionnaires are included in the Questionnaire Annex at the end of this report. 3.2. Qualitative instruments The qualitative instruments included community-level focus group discussions, interviews with community leaders and key informants, field observations and individual in-depth interviews. The complete list of instruments used in the qualitative stage of fieldwork is provided in the Questionnaire Annex at the end of this report. The qualitative instruments were designed to generate data that could be analysed across sex, age and location (rural/urban), and to triangulate key findings across the different social ecological spheres. Focus group discussions generated data at the individual, family and community levels and explored interactions across each of these levels. Interviews engaged closely with community leaders and service providers to generate data at the level of the community and the broader social and political contexts. The focus group discussions took place among four distinct groups: (i) men between 18 and 50 years old who were married or living with a partner; (ii) women between 18 and 50 years old who were married or living with a partner; (iii) men and women between 18 and 25 years old who were unmarried; and (iv) men and women over 50 years old who were married or unmarried. The groups were disaggregated by sex and age to allow a better understanding of how these factors may affect different responses to the research questions. Group sizes varied between 8 and 12 people, and were guided by one facilitator and supported by one note- taker and one translator. Further details about the composition of the groups are provided in Appendix C. The discussions were conducted in the main local language of each region and explored norms, beliefs and practices regarding domestic violence, as well as how and when individuals affected by domestic violence seek medical and legal advice. Exact questions are provided in the Questionnaire Annex at the end of this report. 46 The focus group discussions were able to draw out a great deal of detail on perceptions, determinants and consequences of and responses to domestic violence at the level of individuals, households, communities and the wider society. The questions not only probed interactions across the model’s spheres and the different contexts in which domestic violence takes place (such as asking about violence in public arenas such as market places, compared to violence in private spaces such as homes), but they also explored different types of domestic violence (controlling behaviour, economic violence, sexual violence and so forth). Because the respondents were clustered in specific groups, it was also possible to understand some of the factors (for instance, sex, age, location) that explain the incidence of domestic violence, compared to others. The qualitative instruments also included in-depth interviews with community leaders and key informants, designed to understand experiences of violence and perceptions about the causes and consequences of domestic violence, and ways and barriers to address it. These interviews provided important insights into the practicalities of everyday work related to domestic violence, particularly with respect to law enforcement and engagement with community organisations dealing with domestic violence. By conducting interviews with opinion leaders alongside the focus group discussions, it was possible to generate a detailed picture of how people access state resources (such as health care services), alongside the factors that affect the capacity of service providers to meet the needs of those affected by domestic violence. These interviews enabled the team to also explore the interplay of dynamics occurring at a political and economic level (for example, the failure to provide resources to remote clinics or police stations), and the implications they have for community workers, families and individuals who are affected by domestic violence. Details about the nature and number of interviews are included in Appendix C. The qualitative fieldwork also made use of observations made in the main health facilities and police services of each selected district in each of the ten regions. The aim of these observations was to examine first-hand the availability and use of public services, and understand the barriers that women, men and service providers face. These observations made it easier to understand the context and situations that people described in the in- depth interviews and focus group discussions. Details of the observations conducted in the study are provided in the Questionnaire Annex at the end of this report. The observations were used in particular to complement individual interviews with key staff at the observed institutions. These explored the barriers officials face in implementing national-level policies, and their views about the relevance and effectiveness of policies to address domestic violence. In addition to the methods described above, 15 in-depth individual interviews were conducted with women and men affected by domestic violence identified through specific trigger questions in the quantitative survey. These interviews took place privately immediately after the individual surveys by enumerators specifically trained on the use of qualitative data collection methods (see subsection 3.3 below). These interviews allowed more time with the respondents to discuss in detail the occurrence, perceptions, 47 consequences and ways of coping with domestic violence. Although at first the survey team was concerned about potential respondent fatigue, it found instead that respondents seemed “excited and appreciative to get someone to share their story with”.15 Detailed questions are included in the Questionnaire Annex at the end of this report. 3.3. Sampling design: quantitative survey The study focused on men, women, boys and girls between the ages of 15 and 60 years. This age group extends the usual age group included in international surveys (typically 15 to 49 years), to capture the incidence of domestic violence among older women and men, an issue brought to the attention of the research team during initial scoping interviews with key stakeholders in Ghana. Children below the age of 15 years old were outside the scope of the study due to ethical considerations and the fact that a focus on young children would require a different research design from the one used to elicit information from adults. The quantitative survey aimed to conduct interviews among representative samples of individuals aged 15 to 60 years old at national, regional and urban/rural levels, and by sex. Either one man or one woman was interviewed in each household. The first stage of the sampling design involved compiling basic information about the population of Ghana. According to the 2010 census, the population is spread unevenly across the ten regions of the country, with each region having between 2.5 and 19.4 per cent of the overall population (see Appendix D). The level of urbanisation also varies significantly between regions, ranging between 16.3 and 90.5 per cent.16 Sex ratios are on average balanced, varying only slightly across regions. This information was used to stratify the population of Ghana into 20 strata at the regional level, and by rural and urban areas within regions. The sample design then followed a two-stage procedure. The first stage consisted of randomly selecting primary sampling units within the 20 strata using the sampling frame of the 2010 census enumeration areas (EAs, also referred to as ‘clusters’). In the second stage, 15 households were randomly selected from a listing done by the GSS team in an average of 30 EAs in each region. When designing a representative sample, it is important to take into account the prevalence of the outcome being measured (domestic violence, in the case of this study). According to GSS et al. (2009), the prevalence of domestic violence in 2008 varied considerably between regions (between 26.8 and 49.4 per cent for women, and between 14.3 and 42.3 per cent for men). 17 To be confident that the GFLHS 2015 would capture the ‘true’ prevalence of domestic violence, the study used a more conservative estimate of 50 per cent incidence across both sexes. 15 Correspondence with Mr Peter Takyi Peprah (Project Coordinator), 3rd February 2016. 16 Urban localities were defined as localities with 5,000 inhabitants or more (Ghana Census, 2010). 17 ‘Domestic violence’ in the DHS 2008 is defined as physical, sexual or psychological violence by husband/wife/partner against the respondent. 48 The sample size per region required to accurately estimate a prevalence of 50 per cent across the population of Ghana is given by the following formula (Daniel, 1999): 𝑍2(𝑝(1−𝑝)) n = 2 × 𝐷𝐸, 𝑑 where n = sample size, z is the z statistic for a level of confidence, p is the prevalence, d is the level of precision chosen, and DE represents the design effect. Originally, a statistical confidence of 95 per cent was chosen, which is associated with a Z-score of 1.96. The desired precision of the estimated prevalence was set at 5.5 per cent.18 The design effect (DE) was calculated as follows: 𝐷𝐸 = 1 + 𝜌(𝑚 − 1), where 𝜌 is the intra-regional correlation coefficient, and m the number of regions (10). Similar to García-Moreno et al. (2005), the intra-regional correlation for the different types of domestic violence in the Ghana DHS 2008 data is very low, usually around 1–2 per cent, and never above 4 per cent. As a result, a conservative intra-regional coefficient of 7 per cent was chosen. These calculations resulted in a sample size of 4,995 individuals in 331 EAs. Appendix D shows the allocation of the EAs for each region and describes the survey weights. 3.4. Sampling design: qualitative fieldwork The design of the qualitative fieldwork was based on information on the length of the various instruments, travel times, time needed in each community, transcript preparation times and time spent on the analysis of the data. To minimise travel costs and times, the team selected one district in each of the ten regions of Ghana. Since in all ten regions the regional capital covers the full district and there is no rural site, the district with the second largest urban town within the region was selected. This ensured that both urban and rural sites were covered within the same district. The second most populated urban town was selected as the urban field site, and the least populated in that same district was selected as the rural study site. Overall, the team conducted 80 focus group discussions with women and men across Ghana, 210 in-depth interviews with opinion leaders, activists and legal experts, and observations and in-depth interviews with institutions of justice (20) and health (30). The selection of participants for the qualitative fieldwork differed depending on the qualitative tool used and type of respondent. For focus group discussions with community members within urban areas, respondents from each peri-urban, lower-, middle- and upper-class areas were selected using the help of local assembly staff to map out residential areas by 18 The precision parameter is related to the confidence interval of the estimate, with the width of the confidence interval being 2d. We increased slightly this precision from the conventional 5 per cent to 5.5 per cent due to budgetary constraints. 49 class. In rural areas, participants for the focus group discussions were randomly selected using a list of households in each of the areas, obtained from GSS. For interviews in and observations of police facilities, a representative at the district police office was chosen, as well as staff of the DOVVSU unit at the district police office – if one existed. If the district police office did not have a DOVVSU unit, the DOVVSU unit that covered the area of the district police office was selected instead. There, the research team interviewed the head of the unit and one assistant, and conducted on-site observations of its services. With respect to health facilities, the research team selected in each district two or three members of staff from the district hospital’s outpatient department and emergency unit and, if available, counsellors or psychologists working with individuals affected by domestic violence. Many sampled districts in rural areas (which covered the least populated rural area) did not have health facilities. Instead, staff of available community- based health planning services were interviewed as part of the interviews with opinion leaders. Exact details are provided in Appendix C. In both urban and rural areas, traditional chiefs, the queen mother,19 a religious leader, a headmaster or a teacher of a public school, a traditional healer and/or a private health practitioner were selected to participate in the community/opinion leader interviews. In urban areas, the person responsible for gender-related services at the district assembly was also approached. Depending on availability, one or two representatives from each category of those listed were chosen in each site. Certain categories of respondents, such as legal practitioners and other professionals, were not resident in rural areas and were, therefore, selected in urban sites only. All sub-groups of opinion leaders in each community were found and interviewed (see Appendix C). Exceptions were health care workers in the Upper East and Volta regions, where two (and not the planned three) respondents were available to interview. These were replaced by available representatives of other eligible groups of opinion leaders. Additional key informant interviews in each district targeted legal practitioners identified by the Commission on Human Rights and Administrative Justice (CHRAJ), activists from organisations working on domestic violence at national and regional levels identified by the DV Coalition, media professionals (radio or TV) in each district, and journalists at the national level. 3.5. Training and pre-testing The implementation of all quantitative and qualitative instruments was preceded by extensive training, and by in-depth pre-testing of all instruments. 19 Queen mothers are women who rule alongside the chiefs of a traditional area and who play a central role in the traditional governance of that area. They typically come from a royal line of descendants within a particular clan/family, usually the same as that of the traditional chief. In Ghana, queen mothers are more prominent in the Akan culture, which practises matrilineal descent. Queen mothers’ traditional intermediary roles have been used by international networks and local groups campaigning against domestic violence. 50 The training of the quantitative fieldwork team was done in two stages. The first training event took place in early May 2015 using draft questionnaires. The main aim of this training was to prepare the pre-testing of the survey instruments. The training consisted of six main components: (i) introduction to the study and sensitisation about domestic violence in Ghana; (ii) outline of the quantitative survey and how to introduce oneself and the study to households; (iii) familiarisation with the questionnaires; (iv) introduction to the use of tablet PCs and related software; (v) working through examples of the questionnaires using tablets; (vi) discussion about research ethics; and (v) practice in the field, followed by a detailed debrief session. The survey questionnaires were pre-tested after this training exercise to assess the feasibility and structure of the questions, improve the flow of the questionnaire and refine answer codes. The re-drafted final questionnaires were then used as the basis for the second training exercise in late May 2015. This session offered the team the opportunity to refine the questions and answer codes further, and to adjust how questions were phrased. Enumerators worked in different language groups to agree on the language and terminology they would use when translating certain terms in the questionnaire. This was important to ensure that all research methods used the right language and measured real incidence levels and perceptions of domestic violence. The second training event also included special exercises on qualitative methods involving members of the quantitative survey team assigned to conduct the in-depth individual interviews. The training of the qualitative fieldwork team took place in early April 2015. Eighteen qualitative researchers affiliated with ISSER and 18 enumerators from GSS took part in this event. Supervisors of the three teams (of five members each) were selected on the basis of their leadership expertise in past research projects. The topics covered by this training exercise included: (i) violence, gender and health, including background information on domestic violence in Ghana; (ii) overview of the methods used for the data collection; (iii) discussion about research ethics; (iv) training on the use of in-depth interviews and focus group discussions, including selection criteria, planning and implementation; (v) theory, practice and quality control with respect to notes and transcripts and to observational methods; (vi) methods for data analysis; and (vii) planning fieldwork. The training event was based on role-playing and mock interviews, in which scenarios were played out to help identify the ‘do’s and don’ts’ of sensitive behaviour towards participants, and to test the logic, length and practicality of the drafted instruments. The training of the qualitative field team was also used as an opportunity to refine the instruments, which were tested in the field during a piloting exercise immediately after the training event. 3.6. Fieldwork implementation Fieldwork took place between April and August 2015. The GFLHS 2015 data collection started on 5th June and lasted until 26th August. Fifteen teams of three enumerators and one supervisor visited 331 EAs overall. The teams spent 2.5 days in each EA, devoting one day 51 to the listing of all households and 1.5 days to conducting 15 interviews. The quantitative data collection was done using tablet PCs, which allowed for greater efficiency in terms of time spent in each EA, and better precision of the answers entered into the questionnaire.20 The use of tablets also allowed the teams to send data regularly to a main server. Incoming data were then observed through graphs and exported to a spreadsheet, which allowed the team to look out for errors and inconsistencies in almost real time. The qualitative fieldwork took place between 27th April and 26th July. The qualitative teams travelled and collected data for two weeks and then returned for one week to Accra to finalise their transcripts. They repeated this process four times. This structure was designed to reduce pressure on the field team, and increase the quality of the transcripts. The time spent in Accra also allowed the qualitative research lead to perform a second round of quality checks. Control systems were put in place to ensure quantitative and qualitative data quality. The survey data were checked twice – first by the supervisors before uploading the data to the server, and then by a data monitoring team, who listed and plotted key variables to look for outliers and inconsistencies. Any abnormal findings were fed back to and addressed by the field supervisors. The supervisors also visited the field teams regularly to carry out quality control checks and provide technical and emotional support. With respect to the qualitative fieldwork, interviewers peer-reviewed each other’s notes every day before filing them. Transcripts were reviewed several times during the weeks spent in Accra. Spending several days in any one area also allowed the qualitative team to identify gaps and inconsistencies in time and to go back to the respondents if required. 3.7. Response rates Table 1 shows the survey response rates. These rates ranged from 97.6 per cent in the Western region to 99.6 per cent in Brong Ahafo.21 These high rates of response were due to the precautions taken when administrating the survey. First, as discussed above, information about the survey being about domestic violence was only disclosed after the individual participant had been chosen and privacy had been ensured. Second, the household questionnaire was designed so that it could be administered to heads of households without disclosing the content of the individual questionnaire. This was done because past experience of the GSS with household surveys in Ghana had shown that an engagement with heads of households – usually male – would subsequently lead to higher participation and completion rates among other household members. Third, the individual questionnaire – containing the modules on experience and perpetration of domestic violence – was administered to a randomly chosen individual aged 15–60 years using tablet 20 For example, data are only entered once, rather than recorded on paper and then entered electronically, which minimises data entry error. 21 The percentage of completed interviews is the same as the household response rate, as there were no vacant or destroyed dwellings or absent households, thanks to the listing exercise that preceded the survey collection. 52 PCs. The tablets in turn enabled the research team to make use of pictorial responses, as in García-Moreno et al. (2005), where respondents were able to choose a ‘X’, depicting no experience of physical violence or sexual abuse (used for both separately), or a ‘tick’ depicting having been subject to physical or sexual abuse in the past (see the questionnaires in the Questionnaire Annex at the end of this report). This choice was made at the end of the physical and sexual violence module, without the enumerator observing it. These tools were used to counteract potential levels of under-reporting of these types of violence, as indicated in other studies. Table 1: Response rates to household survey (in percentages) W C GA V E A BA N UE UW Total Completed 97.6 98.1 98.2 98.4 98.3 98.9 99.6 98.8 98.1 97.8 98.5 No one at 2.4 1.9 1.4 1.4 1.5 0.9 0.4 0.8 1.9 1.5 1.3 home Refused 0.0 0.0 0.1 0.0 0.2 0.0 0.0 0.4 0.0 0.0 0.1 Dwelling 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 destroyed Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Notes: W: Western region; C: Central region; GA: Greater Accra region; V: Volta region; E: Eastern region; A: Ashanti region; BA: Brong Ahafo region; N: Northern region; UE: Upper East region; UW: Upper West region. The response rates to the individual survey were also consistently high (Tables 2 and 3). Response rates for women ranged between 96.3 per cent in the Volta region and 100 per cent in Upper West (Table 2). Response rates for men ranged between 98 per cent in Eastern region and 100 per cent in the Upper East and Upper West regions (Table 3). 53 Table 2: Response rates to the individual survey: women (in percentages) W C GA V E A BA N UE UW Total Completed 98.9 99.6 98.2 96.3 99.1 99.3 99.0 98.9 99.0 100.0 98.8 Not available 1.1 0.4 1.0 1.6 0.6 0.0 0.7 0.4 1.0 0.0 0.6 First 0.0 0.0 0.0 0.4 0.0 0.0 0.3 0.0 0.0 0.0 0.1 respondent did not allow second respondent to answer Second 0.0 0.0 0.2 1.2 0.3 0.3 0.0 0.4 0.0 0.0 0.3 respondent refused Lack of privacy 0.0 0.0 0.6 0.0 0.0 0.0 0.0 0.4 0.0 0.0 0.1 Other 0.0 0.0 0.0 0.4 0.0 0.3 0.0 0.0 0.0 0.0 0.1 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Notes: W: Western region; C: Central region; GA: Greater Accra region; V: Volta region; E: Eastern region; A: Ashanti region; BA: Brong Ahafo region; N: Northern region; UE: Upper East region; UW: Upper West region. Table 3: Response rates to the individual survey: men (in percentages) W C GA V E A BA N UE UW Total Completed 99.0 99.3 98.4 98.4 98.0 98.6 98.1 99.2 100.0 100.0 98.7 Not available 1.0 0.0 1.3 0.5 0.5 0.3 0.6 0.4 0.0 0.0 0.6 First 0.0 0.0 0.0 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.1 respondent did not allow second respondent to answer Second 0.0 0.0 0.0 0.5 0.0 0.3 1.2 0.0 0.0 0.0 0.2 respondent refused Lack of privacy 0.0 0.0 0.3 0.0 1.5 0.3 0.0 0.4 0.0 0.0 0.3 Other 0.0 0.7 0.0 0.0 0.0 0.6 0.0 0.0 0.0 0.0 0.2 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Notes: W: Western region; C: Central region; GA: Greater Accra region; V: Volta region; E: Eastern region; A: Ashanti region; BA: Brong Ahafo region; N: Northern region; UE: Upper East region; UW: Upper West region. Table 4 presents the breakdown of the sample along key socio-demographic characteristics. Just over a quarter (27 per cent) of the individuals sampled were aged 30–39 years. Only 8 per cent were aged 15–19 years. The majority of respondents (60 per cent) were self- employed, and 17 per cent were employed. Over two thirds of the sample were married or living with someone. One fifth of the individuals sampled were divorced, separated or 54 widowed, and one tenth had never been married. The most common education level was middle school (37.7 per cent), followed by no education (19 per cent), secondary and primary education (around 15 per cent each), higher education (9.5 per cent) and technical school (3 per cent). In terms of asset quintiles,22 20 per cent of all individuals in the sample are in the bottom quintile, 26.5 per cent in the second quintile, 16.6 per cent in the third quintile, 21.2 per cent in the fourth quintile, and 15.8 per cent in the top quintile. Table 4: Distribution of survey respondents by key socio-economic characteristics (weighted) Sample All All Women Women Men Men number percentage Number percentage number percentage Age group 15–19 394 7.9 191 6.4 203 10.3 20–24 682 13.7 380 12.7 301 15.2 25–29 936 18.8 556 18.6 378 19.1 30–39 1,362 27.4 886 29.6 475 24.0 40–49 856 17.2 512 17.1 343 17.3 50–60 744 15.0 463 15.5 281 14.2 Total 4,974 100.0 2,989 100.0 1,982 100.0 Employment Self-employed 2,984 59.7 1,951 65.3 1,032 52.0 Employed 845 16.9 322 10.8 522 26.4 Not working 1,166 23.3 716 24.0 428 21.6 Total 4,995 100.0 2,989 100.0 1,982 100.0 Marital status Never married 496 10.0 156 5.2 339 17.1 Married or living together 3,472 69.9 2,189 73.4 1,281 64.7 Divorced, separated or 998 20.1 638 21.4 359 18.1 widowed Total 4,966 100.0 2,984 100.0 1,979 100.0 Residence Urban 2,559 51.2 1,580 52.9 964 48.7 Rural 2,436 48.8 1,409 47.1 1,018 51.3 Total 4,995 100.0 2,989 100.0 1,982 100.0 22 Individuals in the lowest asset quintile have four assets from the list provided in the individual questionnaire (see Questionnaire Annex). Individuals in the second quintile have between five and seven assets. Individuals in the third quintile have between eight and nine assets. Individuals in the fourth quintile have between 10 and 12 assets. Individuals in the highest quintile have between 13 and 21 (maximum) assets. 55 All All Women Women Men Men number percentage Number percentage number percentage Region Western 506 10.1 280 9.4 226 11.4 Central 440 8.8 290 9.7 145 7.3 Greater Accra 949 19.0 590 19.7 351 17.7 Volta 454 9.1 251 8.4 200 10.1 Eastern 488 9.8 303 10.1 185 9.3 Ashanti 954 19.1 575 19.2 378 19.1 Brong Ahafo 507 10.1 331 11.1 175 8.8 Northern 411 8.2 216 7.2 192 9.7 Upper East 173 3.5 86 2.9 86 4.4 Upper West 114 2.3 67 2.2 45 2.3 Total 4,995 100.0 2,989 100.0 1,982 100.0 Education level None 958 19.3 684 22.9 274 13.8 Primary 751 15.1 504 16.9 246 12.4 Middle/JSS/JHS* 1,871 37.7 1,167 39.1 703 35.6 Secondary 769 15.5 361 12.1 407 20.6 Technical 144 2.9 78 2.6 66 3.4 Higher 472 9.5 190 6.4 281 14.2 Total 4,964 100.0 2,984 100.0 1,977 100.0 Asset quintile Lowest 999 20.0 583 19.5 410 20.7 Second 1,324 26.5 802 26.8 517 26.1 Third 828 16.6 487 16.3 338 17.1 Fourth 1,057 21.2 667 22.3 384 19.4 Highest 788 15.8 450 15.1 332 16.8 Total 4,995 100.0 2,989 100.0 1,982 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Note: Differences in totals are due to missing values in the respective questions. * JSS = junior secondary school; JHS = junior high school. The survey captured a good balance of male and female respondents in the household interview. There are, however, more female respondents (60 per cent) in the individual interviews because men were much more often away from the household than women – a common feature of most household surveys. This did not affect the analysis or main results, which control for sex composition. The geographical representativeness of the sample was also close to the census distributions. Table 4 shows the distribution of the sample across rural and urban areas: 48.8 per cent in rural areas (49.1 per cent in the 2010 census) and 51.2 per cent in urban areas (50.9 per cent in the 2010 census). It also shows the distribution of the sample across regions: 10.1 per cent in the Western region (9.6 per cent in the 2010 census), 8.8 per cent in the Central region (8.9 per cent in the 2010 census), 19 per cent in Greater Accra (16.3 per 56 cent in the 2010 census), 9.1 per cent in the Volta region (8.6 per cent in the 2010 census), 9.8 per cent in the Eastern region (10.7 per cent in the 2010 census), 19.1 per cent in the Ashanti region (19.4 per cent in the 2010 census), 10.1 per cent in the Brong Ahafo region (9.4 per cent in the 2010 census), 8.2 per cent in the Northern region (10.1 per cent in the 2010 census), 3.5 per cent in the Upper East region (4.2 per cent in the 2010 census) and 2.3 per cent in the Upper West region (2.3 per cent in the 2010 census). 3.8. Data analysis The quantitative survey data were weighted to account for differences in the selection probability and in the number of completed interviews among primary sampling units (see details in Appendix D). The quantitative data were analysed using cross-tabulations between domestic violence outcomes and socio-economic variables such as age, sex, region, educational level, marital status, employment status and wealth, among others. The study also used multivariate regressions to understand in more detail the determinants and consequences of the various forms of domestic violence. Regression analysis account for the correlations between potential determinants or consequences of domestic violence – something that is not possible when using only cross- tabulations. For instance, individuals with higher levels of education also tend to be wealthier. It is then possible that both higher levels of education and higher levels of wealth are associated with higher incidence of domestic violence in isolation but not together. It may be that, in fact, one of those factors is more important than the other. Alternatively, both variables may be independent drivers of domestic violence. Multivariate regressions help to identify the most significant determinants of domestic violence by taking into account correlations between those determinants. Second, multivariate regressions enable the analysis of regional differences. Domestic violence in Ghana varies considerably across regions. This may be because regions are different in the ways in which domestic violence occurs. However, the same result may also be explained not by genuine cultural differences between regions but, rather, because regions with higher levels of domestic violence are also poorer or have higher levels of less educated populations. Multivariate regressions distinguish between these two explanations. This information is, in turn, important for designing the most effective policies. Third, multivariate regressions allow the inclusion within a single framework of determinants of violence at the individual, household and wider community or national level – in keeping with the social ecological framework of domestic violence discussed in Section 2. It is important to note that while multivariate regressions enable the identification of the most significant determinants and consequences of domestic violence in Ghana, the results suggest an association between the variables but do not indicate causality. 57 The analysis of the qualitative data was based on over 300 transcripts. The data were analysed using Grounded Theory (Glaser and Strauss, 1998), 23 which encourages an iterative approach. The qualitative data were analysed using a three-fold analytical system. First, the data were clustered by region. Two researchers each reviewed data from three regions, and one researcher reviewed data from four regions. Data were then coded against the main research questions. This entailed systematically combing through the transcripts from each region and looking for evidence that related to each of the main research questions. This first stage of analysis was collated, and circulated to other research team members to identify any initial queries or inconsistencies. In the second stage, the research team reviewed one another’s coding, and suggested areas that required further substantiation or clarification. In the third stage of the process, the researchers generated a narrative analysis for each research question, based on the empirical material. This analysis was circulated again among the research team for validation. Through these three stages of coding, the research team sought to ensure that the claims made in one set of transcripts, in one region, were validated by triangulating the findings from these transcripts with data from other regions. During the coding, the following rules were applied to ascertain validity across a scale of representativeness. ‘Nationally representative’ findings are those findings that are corroborated by data in all regions, in both rural and urban sites, and in data from a range of stakeholders (in the key informant interviews) and group compositions of the focus group discussions. ‘Regionally representative’ findings are those findings that are supported by data in both rural and urban sites in one region, and in data from a range of stakeholders (in the in-depth interviews) and group compositions of the focus group discussions. 3.9. Ethical considerations The sensitivity of the research topic raised serious ethical and methodological challenges. Notably, the study involved the collection of data that could have potentially affected the privacy of subjects, and may have induced some level of stress and anxiety. The study also involved potential safety risks for interviewers and respondents. Accordingly, the study followed closely the Ethical and Safety Guidelines developed by the WHO (2001) for conducting research on domestic violence. Research was conducted professionally and ethically, with strict respect for principles of integrity, honesty, confidentiality, voluntary participation, impartiality and the avoidance of personal risk. Adherence to these principles was overseen and monitored by the team supervisor and field monitors, in collaboration with the IDS Research Ethics Committee. The detailed mechanisms that were put in place to ensure compliance with the principles included: 23 Grounded Theory is a systematic methodology that groups concepts and categories through induction based on data collected. 58 1. The safety of respondents and the research team was paramount, and guided all decisions in the study. To ensure the safety of both respondents and the research team, the survey was introduced to communities and households as a study on women’s health and family life, as discussed earlier. Interviews were conducted in private settings at convenient and safe times for the participant, and interviewers were trained to change the subject to a less sensitive topic if an interview was interrupted. The survey participants were informed about this beforehand. The safety of interviewers was also ensured by planning visits to research sites in teams, particularly during the evening and to areas known to be unsafe. 2. Protecting confidentiality was essential to ensure both safety and data quality. The questionnaires were implemented and analysed anonymously, with the names of individuals in each household replaced by a numeric code (identifier). Respondents were informed about protocols for confidentiality and anonymity. No names or addresses appeared on the questionnaires or transcripts. The survey questionnaires were linked to the in-depth interview transcripts via the personal identifiers only. The address of individuals willing to participate in in-depth interviews was noted when they consented to be interviewed, and destroyed after completion of the interview. Interviewers were trained and instructed to obtain the permission of respondents to record interviews beforehand (when this was necessary, for the in-depth interviews and focus group discussions). Tapes and digital recorders were locked in safe places at all times, and destroyed after transcription. 3. All research team members were carefully selected and received specialised training and ongoing support. The training entailed an introduction to gender, gender discrimination and domestic violence, as discussed earlier. Training also addressed the issue of stereotypes, biases and fears regarding domestic violence, and provided mechanisms to overcome these. Field researchers learned and practised how to ask questions in a supportive and non-judgemental manner, and were given opportunities to come to terms with emotions evoked during the interview process or by the topic itself (for instance, if the field researcher had experienced abuse her/himself). Selected field researchers were encouraged to think about and discuss openly or in private any matters of concern with training staff and field managers. Field researchers who faced emotional distress during fieldwork were free to take breaks, carry out less emotionally draining tasks or withdraw from the fieldwork altogether without judgement. Field researchers were also trained to regard themselves as researchers and not counsellors. However, they received sufficient and adequate information to assist those respondents who asked for help (in the form of pamphlets that directed study participants to appropriate services and counselling). 4. The study design included actions aimed at reducing any possible distress caused to the participants by the research. As discussed above, the language of the questionnaire and the behaviour of field researchers was non-judgemental and supportive at all times, and interviewers were trained on how to respond to distress caused by the interviews. They were also encouraged and learned how to terminate interviews if necessary. All interviews 59 ended on a positive note, reminding the participants of their self-worth and of the importance of having shared their experiences. 5. Field researchers were trained to refer individuals requesting assistance to available local services and sources of support. In cooperation with government agencies and services, local partners and other stakeholders, such as NGOs and women’s groups, the research team identified and obtained consent from local formal and informal providers to direct to them individuals in need. Pamphlets containing discreet information about these services were given to respondents, regardless of whether they reported being affected by domestic violence or not. 6. Findings will be properly interpreted and used to advance policy and intervention development. The research study involved from the outset formal and informal groups involved in addressing domestic violence, mitigating its consequences and increasing the well-being of those exposed to domestic violence (see the composition of the Steering Committee). The research team sought their advice and input at key stages of the study. The study also paid close attention to the language used when interpreting results, to avoid stigmatisation or reinforcement of stereotypes of sub-groups. Similar procedures will be employed when disseminating the results of the study. 60 4. Empirical results This section presents the main findings of the study for each of the research questions outlined in Section 1. This analysis brings together the qualitative data, and the descriptive statistics and regression analysis based on the GFLHS. This joint analysis of the qualitative and quantitative data has allowed the research team to identify points of similarity and contrast between the data collected using different methods. Each of the subsections below presents and discusses the main results of the study against each research question in turn. Subsection 4.1 analyses the incidence of social, physical, sexual, psychological and economic violence across different population groups in Ghana – independently of whether this violence was perpetrated or not within domestic relations. Subsection 4.2 focuses on the main remit of the study – the analysis of violence against women and men perpetrated within domestic relations (i.e. domestic violence). Subsection 4.3 examines the profiles of perpetrators of domestic violence in Ghana. Subsection 4.4 analyses attitudes and social norms that drive domestic violence in Ghana across individuals, families, communities and the wider society. Subsection 4.5 discusses the main determinants of domestic violence uncovered in the data at the individual, family, community and wider societal levels. The consequences of domestic violence are analysed in subsection 4.6, with a focus on physical and mental health and the impact of domestic violence on daily lives and children. The final subsection (4.7) analyses how women and men affected by domestic violence in Ghana access and use institutions and available public services. 4.1. Incidence of violence (domestic and non-domestic) against women and men in Ghana Conforming to international definitions (e.g. UNICEF, 2002; WHO, 2000), the national 2007 Domestic Violence Act (Act 732), and prior studies conducted in Ghana, the study has analysed the incidence of the following forms of violence (both within and outside domestic relations):  social violence – acts of controlling behaviour, such as preventing someone from seeing friends or family of birth; stopping someone from leaving the house; requiring to know where someone is at all times; stalking; spreading false information, videos or photos without permission; or forcing women and girls to have an abortion;  physical violence – slapping, pushing, shoving, hitting, kicking, dragging or throwing objects at someone; choking, strangling or burning someone; using a weapon, hazardous chemicals or substances against someone; or kicking or pulling someone’s external genitalia (for male respondents only);  sexual violence – acts of unwanted sexual comments or physical contact; rape by physical force, or otherwise forced sex (for instance, by blackmail or threats); denial of using protection during sex; a sexual partner hiding their HIV status; sexual acts or intercourse that were performed on the basis of feeling there was no option; or penetration with an object against someone’s will; 61  psychological violence – insults, belittling or humiliation in private or in front of others; threats of abandonment; being ignored or treated indifferently; intimidations and acts aimed at scaring someone; threats of using weapons against someone; or threats of hurting someone or someone one cares about; and  economic violence – denial of household money for expenses (chop money) even if enough financial means are available; unsolicited taking of money; control of belongings and spending decisions; damage to or destruction of someone’s property; denial of the right to work; forcing someone to work against their will; or denial of food and other basic needs. The GFLHS 2015 asked detailed questions about individual experiences of any of these five broad categories of violence over the individual’s lifetime, and in the 12 months prior to the survey. This allowed the study to capture both current incidence levels of overall levels of (domestic and non-domestic) violence and the lifetime accumulation of experiences of overall violence among men and women. Table 5 summarises this information. Table 5: Incidence of violence in Ghana 2015 (in percentages) Social Physical Sexual Psychological Economic Total violence violence violence violence violence Percentage of respondents who experienced violence over their lifetime Women 41.0 42.4 30.0 48.0 28.2 71.5 Men 34.4 50.8 23.1 51.9 25.8 71.4 Percentage of respondents who experienced violence in the 12 months prior to the survey Women 20.8 8.9 10.6 22.5 14.4 42.9 Men 18.7 10.2 9.5 28.0 12.2 43.6 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Table 5 shows that 71.5 per cent of women and 71.4 per cent of men reported having experienced at least one form of violence (domestic and non-domestic) over their lifetime, and 42.9 per cent of women and 43.6 per cent of men experienced at least one form of violence in the 12 months prior to the survey. The following subsections discuss in detail how each of these forms of violence was experienced by the respondents of the GFLHS 2015 across their characteristics and geographical location. 4.1.1. Social violence Table 5 shows that 41.0 per cent of female respondents and 34.4 per cent of male respondents experienced at least one type of social violence over their lifetime, committed by perpetrators both within and outside domestic relations. Just over 20 per cent of women and 18.7 per cent of men experienced at least one form of social violence during the 12 months prior to the survey. The survey explored further the incidence of social violence along its different dimensions (as defined above), and across regional and socio-economic 62 characteristics of the sample. This disaggregated information is included in Tables 6–9 and discussed in the following paragraphs. Dimensions of social violence. Social violence experienced by both men and women over their lifetime was mostly in the form of rumours, photos or videos. Almost a quarter (24.0 per cent) of women (Table 6) and 22.8 per cent of men (Table 8) experienced this form of violence. The second and third most common forms of social violence over the lifetime were, respectively, not being allowed outside their homes (experienced by 14.4 per cent of women and 10.6 per cent of men), and being kept away from family and friends (experienced by 14.1 per cent of women and 8.9 per cent of men). These patterns were similar when considering the incidence of social violence in the 12 months prior to the survey (Tables 7 and 9). Age patterns. The incidence of social violence among men and women was strongly associated with age, with older people being less likely to have experienced social violence. For example, 53 per cent and 49.4 per cent of women aged 15–19 and 20–24 years, respectively, reported having experienced social violence over their lifetime. This was in contrast with 33.1 per cent of women aged 40–49 years and 33.7 per cent of women aged 50–60 years (Table 6). There were also large differences by age among men: almost 40 per cent of young men aged 15–19 years and 36.6 per cent of men aged 20–24 years reported having experienced social violence over their lifetime (Table 8). The incidence of social violence over the lifetime of men aged 40–49 years was 29.7 per cent. It was 32.2 per cent among men aged 50–60 years. These age patterns were similar when considering the incidence of social violence in the 12 months prior to the survey (Tables 7 and 9). Employment status. The incidence levels of social violence over the lifetime of women and men were slightly higher for women not working (i.e. not in waged employment): 42.8 per cent of currently non-working women reported having experienced social violence, compared to 40.5 per cent of self-employed women and 40.3 per cent of employed women (Table 6). Social violence was most prominent among employed men (37.6 per cent), followed by men who were not working at the time of the survey (34.6 per cent) and self- employed men (32.7 per cent) (Table 8). Tables 7 and 9 show similar patterns across employment status of women and men for the incidence of social violence in the 12 months prior to the survey. Marital status. The results indicate that married or divorced, separated or widowed men and women were more likely than those never married to have experienced social violence. Almost 41 per cent of women who were married or living with a partner, and 41.5 per cent of women divorced, separated or widowed, experienced social violence, in comparison to 39.9 per cent of women who were never married (Table 6). Thirty per cent of never-married men reported having experienced at least one form of social violence over their lifetime, in comparison to, respectively, 35.1 and 35.8 per cent of men who lived with a partner (or were married) or men who were divorced, separated or widowed (Table 8). Women who were never married (32.8 per cent) reported higher levels of social violence in the 12 63 months prior to the survey than women with other marital status (this was the reverse of the pattern observed for lifetime social violence) (Table 7). Geographical patterns. The incidence of lifetime social violence was higher among women living in rural areas (42.3 per cent) than women living in urban areas (39.9 per cent). It was also higher among women living in the Eastern (67.7 per cent) and Western regions (48.3 per cent) (Table 6) than for those living in other regions. The incidence of lifetime social violence among men showed a similar pattern (Table 8). Differences across urban and rural areas were less pronounced among women for the 12 months prior to the survey than among women reporting lifetime experiences of social violence. Among the former, women living in urban areas were slightly more likely than women in rural areas to have experienced this form of violence (21.5 per cent, in comparison to 20.1 per cent in rural areas) (Table 7). Men living in urban areas were also less likely to report having experienced social violence over the 12 months prior to the survey than men in rural areas (21.5 per cent, in comparison to 16.1 per cent of men living in rural areas – a reverse of the pattern observed among men (and women) reporting social violence over their lifetime) (Table 9). These differences may be due to recent changes in media technology (since the main form of social violence was the spread of rumours, videos and photos), which are more likely to be more widespread in urban areas. Education levels and asset ownership. The incidence of lifetime social violence was lower among women with no formal education (31.8 per cent) and among women with higher (university) education (35.1 per cent) than among women with intermediate levels of education (primary, middle, secondary and technical) (Table 6). Men with no education or with technical education were less likely to report social violence (25.6 per cent and 26.2 per cent, respectively), than men with other levels of education (Table 8). The results show no clear association between the lifetime incidence of social violence and asset ownership for men or women. These education and wealth patterns were similar when considering the incidence of social violence in the 12 months prior to the survey (Tables 7 and 9). 64 Table 6: Percentage of women who have experienced domestic or non-domestic social violence over their lifetime Kept from Stopped from Asked to Stalked Rumour, Forced to Any social seeing leaving home report photos or have an violence friends or activities videos spread abortion family Age group (years) 15–19 23.0 27.8 15.4 4.2 24.6 6.7 53.0 20–24 20.1 20.7 18.1 6.0 24.9 9.4 49.4 25–29 16.7 13.2 12.7 4.2 27.6 12.7 47.7 30–39 11.1 12.4 10.0 2.9 24.1 8.5 39.0 40–49 10.8 10.3 5.2 1.2 20.7 7.2 33.1 50–60 11.6 13.3 5.2 1.8 22.1 5.1 33.7 Employment Self-employed 13.7 14.0 8.6 2.7 25.2 8.2 40.5 Employed 11.4 11.3 13.8 3.5 21.5 9.5 40.3 Not working 16.2 16.8 13.4 4.2 21.8 9.1 42.8 Marital status Never married 18.8 19.9 10.1 1.6 17.6 1.5 39.9 Married or living together 13.4 14.0 10.7 2.9 23.9 8.8 40.9 Divorced/separated/widowed 15.2 14.2 9.1 4.2 26.1 9.3 41.5 Residence Urban 13.2 12.4 11.6 3.8 22.9 8.9 39.9 Rural 15.0 16.6 8.8 2.4 25.2 8.2 42.3 65 Kept from Stopped from Asked to Stalked Rumour, Forced to Any social seeing leaving home report photos or have an violence friends or activities videos spread abortion family Region Western 14.9 16.0 6.0 1.1 32.3 12.7 48.3 Central 12.4 15.4 7.7 2.4 18.6 11.5 38.1 Greater Accra 8.9 6.7 9.6 3.0 17.0 6.1 29.4 Volta 10.7 10.9 9.3 3.6 37.3 4.9 47.3 Eastern 35.2 41.0 17.1 6.3 42.0 11.9 67.7 Ashanti 7.0 9.0 12.1 2.4 25.9 11.7 42.2 Brong Ahafo 17.2 10.1 13.8 4.4 16.2 9.7 40.3 Northern 19.5 19.7 4.9 2.9 13.2 0.9 31.0 Upper East 15.1 19.2 3.3 2.7 14.7 0.0 27.3 Upper West 8.0 7.5 12.6 1.1 11.1 2.3 24.6 Education level None 13.0 12.9 5.1 1.5 18.7 5.3 31.8 Primary 17.5 18.5 10.1 3.5 29.3 9.8 46.2 Middle/JSS/JHS 14.1 14.9 10.6 3.3 26.7 10.7 44.5 Secondary 13.0 12.8 16.8 4.3 19.9 7.6 42.6 Technical 12.7 11.3 11.9 3.3 28.8 13.5 43.8 Higher 11.6 10.3 14.1 5.0 18.2 3.6 35.1 Asset quintile Lowest 15.4 16.2 6.9 2.5 24.8 6.9 41.3 Second 15.0 15.7 9.7 3.5 24.4 8.8 41.2 Middle 12.3 13.5 10.6 3.3 26.1 11.6 43.8 Fourth 14.0 13.5 12.2 3.5 23.0 9.0 39.4 Highest 12.9 11.9 12.7 2.8 21.4 6.3 39.6 Total 14.1 14.4 10.3 3.1 24.0 8.5 41.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 66 Table 7: Percentage of women who experienced domestic or non-domestic social violence in the last 12 months Kept from Stopped from Asked to Stalked Rumour, Forced to Any social seeing leaving home report photos or have an violence friends or activities videos spread abortion family Age group (years) 15–19 16.9 21.8 12.2 3.9 17.9 3.9 44.2 20–24 7.8 7.1 12.7 3.7 16.7 2.1 31.5 25–29 5.0 4.3 8.4 2.4 17.7 0.9 25.7 30–39 2.4 2.3 4.2 1.7 15.2 0.4 18.7 40–49 1.3 1.5 2.3 0.8 9.6 0.0 12.7 50–60 0.9 0.7 0.7 0.4 8.2 0.0 9.8 Employment Self-employed 2.4 2.5 3.7 1.5 14.0 0.6 18.1 Employed 2.2 2.7 7.0 1.2 14.0 1.5 22.7 Not working 9.3 9.3 10.5 3.2 13.8 1.1 27.5 Marital status Never married 13.1 14.2 8.9 1.6 14.9 0.5 32.8 Married or living together 3.7 3.9 6.1 1.8 13.9 0.8 20.7 Divorced/separated/widowed 3.3 2.7 3.5 2.1 14.2 0.9 18.5 Residence Urban 4.1 3.6 6.6 1.9 14.4 0.7 21.5 Rural 4.0 4.7 4.7 1.8 13.5 1.0 20.1 67 Kept from Stopped from Asked to Stalked Rumour, Forced to Any social seeing leaving home report photos or have an violence friends or activities videos spread abortion family Region Western 4.7 6.6 3.9 1.1 15.1 2.2 23.3 Central 5.0 4.6 3.9 0.3 8.5 2.1 16.8 Greater Accra 3.1 3.2 6.0 2.5 14.0 0.1 19.5 Volta 5.4 6.2 6.8 2.3 19.2 0.4 24.9 Eastern 5.2 5.5 7.6 2.9 25.9 1.1 32.3 Ashanti 3.2 2.8 6.8 1.5 16.3 0.4 23.5 Brong Ahafo 1.8 1.4 6.1 2.6 7.2 1.0 13.3 Northern 7.2 4.9 2.9 2.0 4.7 0.2 13.9 Upper East 1.8 6.7 0.6 0.6 9.6 0.0 14.5 Upper West 8.0 5.6 10.2 1.1 8.9 1.1 18.1 Education level None 2.4 2.2 1.7 1.3 8.5 0.3 12.5 Primary 4.1 4.9 4.9 1.5 17.9 0.7 23.6 Middle/JSS/JHS 5.0 5.3 5.7 2.0 16.3 1.3 23.7 Secondary 4.5 4.6 13.1 3.4 13.9 0.5 26.9 Technical 0.0 1.3 7.5 1.7 11.3 0.6 15.8 Higher 4.8 2.4 7.4 1.2 10.4 0.8 17.0 Asset quintile Lowest 4.4 4.5 3.9 1.6 14.0 0.9 20.2 Second 4.3 4.5 5.1 2.4 13.3 1.0 20.2 Middle 3.7 4.6 5.8 1.9 14.7 1.6 22.6 Fourth 4.0 4.1 7.7 1.9 14.7 0.4 21.2 Highest 3.6 2.5 6.0 1.2 13.3 0.2 20.4 Total 4.1 4.1 5.7 1.9 14.0 0.8 20.8 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 68 Table 8: Percentage of men who have experienced domestic or non-domestic social violence over their lifetime Kept from Stopped from Asked to Stalked Rumour, Forced to Any social seeing leaving home report photos or have an violence friends or activities videos spread abortion family Age group (years) 15–19 11.8 19.0 5.5 2.5 18.5 N/A 39.8 20–24 11.4 10.0 15.0 4.5 24.1 N/A 36.6 25–29 8.6 10.1 14.0 4.4 22.4 N/A 36.3 30–39 6.8 9.2 10.5 2.5 24.6 N/A 33.9 40–49 7.9 8.0 4.3 2.9 22.5 N/A 29.7 50–60 9.6 11.1 4.3 2.0 22.4 N/A 32.2 Employment Self-employed 9.9 11.6 7.8 2.1 22.1 N/A 32.7 Employed 7.1 8.2 13.1 5.2 26.2 N/A 37.6 Not working 8.8 11.2 8.5 3.1 20.4 N/A 34.6 Marital status Never married 10.6 13.6 5.4 2.1 15.5 N/A 30.0 Married or living together 9.0 10.5 9.7 2.8 24.0 N/A 35.1 Divorced/separated/widowed 7.0 8.1 12.2 5.6 25.1 N/A 35.8 Residence Urban 7.7 8.4 10.0 3.6 22.5 N/A 33.8 Rural 10.1 12.6 8.8 2.8 23.1 N/A 35.0 69 Kept from Stopped from Asked to Stalked Rumour, Forced to Any social seeing leaving home report photos or have an violence friends or activities videos spread abortion family Region Western 5.9 9.3 10.7 6.3 34.8 N/A 49.3 Central 10.1 9.6 9.0 4.1 25.1 N/A 35.0 Greater Accra 5.3 3.6 9.6 2.6 12.0 N/A 22.9 Volta 2.2 5.7 7.6 1.7 33.5 N/A 39.6 Eastern 28.4 33.1 10.4 4.4 34.7 N/A 59.2 Ashanti 4.1 7.9 11.8 3.9 24.3 N/A 33.9 Brong Ahafo 11.1 11.0 12.8 3.1 21.4 N/A 33.0 Northern 13.2 14.3 4.9 0.6 11.9 N/A 21.5 Upper East 12.2 11.8 4.0 1.0 6.1 N/A 17.4 Upper West 6.1 6.1 1.8 0.0 14.9 N/A 19.2 Education level None 10.6 14.5 3.0 0.5 14.1 N/A 25.6 Primary 12.7 14.4 5.8 1.3 25.4 N/A 37.0 Middle/JSS/JHS 8.1 11.1 7.7 2.5 23.2 N/A 34.6 Secondary 8.4 9.3 13.4 6.6 29.6 N/A 41.0 Technical 7.8 1.2 9.0 4.0 17.9 N/A 26.2 Higher 7.1 6.3 17.4 4.1 19.6 N/A 32.8 Asset quintile Lowest 12.0 13.9 5.7 1.9 20.2 N/A 31.5 Second 9.8 11.9 7.5 2.5 22.3 N/A 34.5 Middle 7.3 8.9 8.6 2.2 26.9 N/A 37.1 Fourth 7.6 10.0 12.0 4.0 24.3 N/A 35.6 Highest 7.0 6.9 14.6 5.7 21.0 N/A 33.7 Total 8.9 10.6 9.4 3.2 22.8 N/A 34.4 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 70 Table 9: Percentage of men who experienced domestic or non-domestic social violence in the last 12 months Kept from Stopped from Asked to Stalked Rumour, Forced to Any social seeing leaving home report photos or have an violence friends or activities videos spread abortion family Age group (years) 15–19 5.3 7.1 4.1 1.7 13.0 N/A 23.1 20–24 2.3 1.0 10.6 3.5 16.4 N/A 26.1 25–29 1.0 1.4 8.1 3.5 12.1 N/A 20.5 30–39 0.6 1.5 5.7 1.5 13.3 N/A 18.2 40–49 0.8 0.7 2.1 1.8 11.2 N/A 13.9 50–60 0.8 1.8 2.2 2.0 8.5 N/A 12.1 Employment Self-employed 1.2 1.4 4.3 1.8 11.9 N/A 16.2 Employed 1.2 1.6 7.7 3.3 15.0 N/A 22.7 Not working 2.5 3.4 6.2 2.4 10.8 N/A 19.9 Marital status Never married 2.5 4.0 3.1 1.5 8.8 N/A 15.1 Married or living together 1.4 1.7 5.8 1.8 12.9 N/A 18.7 Divorced/separated/widowed 0.8 0.5 7.6 4.9 14.5 N/A 22.3 Residence Urban 1.7 2.3 6.4 2.8 13.6 N/A 21.5 Rural 1.3 1.5 4.9 1.9 11.4 N/A 16.1 71 Kept from Stopped from Asked to Stalked Rumour, Forced to Any social seeing leaving home report photos or have an violence friends or activities videos spread abortion family Region Western 1.8 4.3 7.7 5.6 17.8 N/A 30.4 Central 2.3 2.0 6.9 3.3 15.8 N/A 19.7 Greater Accra 1.9 1.8 7.1 2.6 6.9 N/A 15.8 Volta 0.5 1.3 5.6 1.1 18.3 N/A 24.1 Eastern 1.8 2.1 5.1 2.5 20.6 N/A 24.5 Ashanti 1.1 1.0 8.2 2.9 13.4 N/A 20.6 Brong Ahafo 1.4 1.8 2.0 0.9 12.9 N/A 16.0 Northern 0.8 1.1 1.4 0.3 2.3 N/A 4.6 Upper East 0.7 0.7 1.0 0.0 3.1 N/A 4.8 Upper West 4.4 4.4 0.0 0.0 10.2 N/A 14.6 Education level None 1.0 3.3 0.4 0.0 6.5 N/A 10.2 Primary 2.8 2.2 2.6 1.3 14.4 N/A 17.9 Middle/JSS/JHS 2.1 2.5 5.0 1.8 11.4 N/A 17.7 Secondary 0.5 0.9 7.6 4.6 18.1 N/A 24.4 Technical 3.1 0.0 9.0 4.0 10.5 N/A 21.6 Higher 0.3 0.5 11.3 3.3 11.3 N/A 21.6 Asset quintile Lowest 1.2 2.6 2.5 1.7 9.8 N/A 13.3 Second 1.6 1.2 4.0 1.9 12.2 N/A 17.3 Middle 1.5 1.3 5.2 1.2 13.5 N/A 18.3 Fourth 1.9 2.1 7.5 2.0 13.9 N/A 22.5 Highest 1.1 2.5 10.1 5.3 13.5 N/A 23.6 Total 1.5 1.9 5.6 2.3 12.5 N/A 18.7 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 72 4.1.2. Physical violence Men reported higher incidence levels of physical violence than women (Table 5): 42.4 per cent of female respondents and 50.8 per cent of male respondents experienced at least one of the above acts of physical violence over their lifetime. Almost 9 per cent of women and 10.2 per cent of men experienced at least one act of physical violence during the 12 months prior to the survey. Tables 10–13 show further disaggregated detail on the incidence of physical violence across different population groups and characteristics, which are discussed in the paragraphs below. Dimensions of physical violence. The most common forms of physical violence experienced by both men and women over their lifetime were being slapped or being hit by thrown objects. This form of violence affected 32.7 per cent of women (Table 10) and 38.3 per cent of men (Table 12). The next most prominent form of physical violence was being hit by another person, experienced by 15.1 per cent of women (Table 10) and 20.8 per cent of men (Table 12). These patterns of physical violence were very similar across the lifetime of individuals and in the 12 months prior to the survey. Age patterns. The incidence of physical violence over people’s lifetimes was strongly associated with age for both women and men, with older individuals less likely to have experienced physical violence. Over 47 per cent of women aged 24 years and below reported having experienced physical violence, compared to 40.6 per cent of women aged 40–49 years and 35.7 per cent of women aged 50–60 years (Table 10). Younger men aged 15–19 years were particularly at risk of experiencing physical violence: 64.3 per cent of men in this age group reported having experienced at least one form of physical violence over their lifetime, compared to, for instance, 42.9 per cent of men aged 50–60 years (Table 12). These age patterns were very similar across the lifetime of individuals and in the 12 months prior to the survey (Tables 11 and 13). Employment status. Employed and non-working women were more likely to have experienced physical violence over their lifetime (43.7 and 43.6 per cent, respectively), than self-employed women (41.8 per cent) (Table 10). This pattern was similar among the male sample, but incidence levels were higher: 56.9 per cent of men who were not working reported having experienced at least one form of physical violence over their lifetime, while 55 per cent of employed and 46 per cent of self-employed men reported experiencing this (Table 12). The employment patterns were similar to those observed when considering experiences of physical violence in the 12 months prior to the survey (Table 11 and 13). Marital status. The lifetime incidence of physical violence was higher among individuals, particularly women, who were divorced, separated or widowed, or married. Experiences with at least one form of physical violence were reported by 46.3 per cent of divorced, separated or widowed women and by 41.8 per cent of women who were married (or living with a partner), in contrast to 35.1 per cent of women who were never married (Table 10). The differences in the results for the male sample were less pronounced (Table 12): 52.6 per 73 cent of divorced, separated or widowed men reported having experienced physical violence, compared to 51.4 per cent of men who were never married and 50.1 per cent who were married. The incidence of physical violence in the 12 months prior to the survey was highest among never-married women (Table 11). Geographical patterns. The incidence of lifetime physical violence was higher among women living in rural areas (44.4 per cent) than women living in urban areas (40.7 per cent) (Table 10). A similar proportion of men in rural and urban areas experienced physical violence (50.7 per cent and 50.9 per cent, respectively) (Table 12). Women and men in the Eastern region reported having experienced higher levels of physical violence than those in any other region. These patterns were common across the lifetime of individuals and in the 12 months prior to the survey (Table 11 and 13). Education levels and asset ownership. Women and men with no formal education (34.6 and 31.7 per cent, respectively) and women and men with higher education (38.0 and 51.2 per cent, respectively) were less likely to experience physical violence than women and men with primary education (45.3 and 52.1 per cent, respectively) or other levels of education (Tables 10 and 12). This suggests that intermediate levels of education may be associated with increased risk of physical violence among women and men. Women in the highest and fourth asset quintiles were less likely to experience physical violence (38.9 and 41.4 per cent, respectively) than women in the three lowest asset quintiles. For instance, 44.7 per cent of women in the lowest asset quintile reported experiencing physical violence (Table 10). The opposite was true for men: men in the lowest asset quintile were less likely to experience physical violence (46.1 per cent) than men in the highest asset quintile (50.1 per cent) (Table 12). These education and wealth patterns were very similar across the lifetime of individuals and in the 12 months prior to the survey (Tables 11 and 13). 74 Table 10: Percentage of women who have experienced domestic or non-domestic physical violence over their lifetime Slapped or Pushed Hit Kicked, Choked or Burnt Attacked Any physical thrown or shoved dragged or strangled with violence things at beaten up weapon Age group (years) 15–19 35.0 18.5 13.5 16.8 2.4 1.7 1.3 47.6 20–24 36.2 15.0 14.7 15.2 1.7 1.2 2.2 47.0 25–29 34.7 18.9 14.5 14.2 2.7 1.7 2.6 45.5 30–39 32.0 13.3 15.3 13.7 2.1 0.7 4.1 42.0 40–49 32.1 14.6 17.0 12.7 2.3 1.0 2.8 40.6 50–60 28.5 12.3 14.2 11.5 2.1 0.5 1.7 35.7 Employment Self-employed 32.8 14.5 16.1 13.6 2.6 1.1 2.8 41.8 Employed 29.6 17.5 14.4 13.6 1.8 0.3 3.9 43.7 Not working 33.9 15.1 12.6 14.0 1.4 1.0 2.3 43.6 Marital status Never married 22.3 17.9 10.4 9.4 0.0 1.3 1.3 35.1 Married or living together 32.2 14.1 15.2 13.7 2.1 0.9 2.8 41.8 Divorced/separated/widowed 36.9 17.2 15.5 14.5 3.3 1.2 3.3 46.3 Residence Urban 31.6 14.0 15.4 11.6 1.8 0.8 2.9 40.7 Rural 34.0 16.0 14.7 16.0 2.7 1.2 2.7 44.4 75 Slapped or Pushed Hit Kicked, Choked or Burnt Attacked Any physical thrown or shoved dragged or strangled with violence things at beaten up weapon Region Western 35.5 8.8 5.6 9.0 0.4 0.0 2.3 40.5 Central 36.0 11.4 16.0 13.0 5.2 0.5 3.7 44.2 Greater Accra 21.1 12.1 13.1 5.1 1.1 0.9 2.9 29.8 Volta 44.4 31.0 29.0 22.7 2.5 1.0 0.8 56.3 Eastern 42.8 25.3 29.4 28.2 4.9 3.9 4.6 61.9 Ashanti 37.6 11.9 11.0 16.5 1.3 0.4 3.3 46.3 Brong Ahafo 35.5 16.1 15.9 10.2 3.0 1.5 3.1 44.2 Northern 25.9 15.8 12.4 17.0 0.5 0.4 1.3 39.0 Upper East 11.2 5.5 3.2 4.4 2.4 1.1 0.0 13.5 Upper West 13.9 4.3 6.1 7.4 3.1 0.0 2.4 19.9 Education level None 27.0 14.5 14.6 12.7 2.5 0.5 1.8 34.6 Primary 35.5 15.9 14.6 17.0 3.3 1.8 4.0 45.3 Middle/JSS/JHS 37.3 15.5 16.9 13.8 2.3 1.2 2.9 47.0 Secondary 28.2 12.5 13.2 13.0 1.1 1.2 1.6 40.3 Technical 34.8 19.5 17.5 10.3 0.0 0.0 5.9 45.2 Higher 25.5 13.5 9.9 10.8 0.9 0.0 3.6 38.0 Asset quintile Lowest 36.1 20.2 18.7 15.6 3.9 0.9 2.3 44.7 Second 32.7 14.6 11.9 15.4 2.7 1.1 3.1 44.0 Middle 33.6 13.8 14.1 13.1 1.5 0.9 2.3 42.9 Fourth 29.0 12.5 17.4 12.5 1.3 1.0 2.5 38.9 Highest 33.0 13.9 13.7 10.6 1.6 1.2 4.0 41.4 Total 32.7 15.0 15.1 13.7 2.2 1.0 2.8 42.4 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 76 Table 11: Percentage of women who experienced domestic or non-domestic physical violence in the last 12 months Slapped or Pushed Hit Kicked, Choked or Burnt Attacked Any thrown or shoved dragged or strangled with physical things at beaten up weapon violence Age group (years) 15–19 17.7 7.8 6.1 7.8 1.7 1.1 1.1 28.0 20–24 10.1 3.1 4.1 2.5 0.2 0.3 0.0 14.1 25–29 6.5 3.9 2.2 2.6 1.4 0.3 1.0 9.5 30–39 6.2 1.2 1.7 2.5 0.6 0.0 1.6 9.2 40–49 3.2 0.8 1.1 0.5 0.2 0.0 0.2 3.8 50–60 1.0 0.0 0.2 0.0 0.0 0.0 0.0 1.2 Employment Self-employed 5.0 1.5 1.7 1.8 0.6 0.0 0.7 7.2 Employed 4.8 2.4 2.2 1.4 0.4 0.0 0.7 7.0 Not working 10.0 3.7 3.1 3.3 0.7 0.7 1.1 14.5 Marital status Never married 11.7 6.4 3.5 4.5 0.0 1.3 1.3 20.0 Married or living together 5.9 1.6 2.0 1.6 0.6 0.1 0.7 8.2 Divorced/separated/widowed 5.6 2.9 2.0 3.2 0.9 0.0 0.8 8.5 Residence Urban 6.0 1.7 2.3 2.1 0.5 0.1 0.9 8.6 Rural 6.4 2.6 1.7 2.2 0.7 0.2 0.6 9.3 77 Slapped or Pushed or Hit Kicked, Choked or Burnt Attacked Any thrown shoved dragged or strangled with physical things at beaten up weapon violence Region Western 6.1 1.6 1.3 1.3 0.0 0.0 0.6 8.0 Central 6.3 1.4 2.9 2.8 0.8 0.5 1.6 10.9 Greater Accra 4.9 1.9 1.8 1.5 0.2 0.0 1.2 6.5 Volta 8.1 2.7 2.8 2.7 0.3 1.0 0.0 10.3 Eastern 8.1 2.3 3.1 3.1 1.4 0.0 1.2 11.1 Ashanti 7.3 2.0 1.8 2.2 0.3 0.0 0.4 10.4 Brong Ahafo 6.1 3.3 2.7 2.2 1.2 0.2 0.7 9.5 Northern 4.5 2.3 0.6 2.0 0.5 0.0 0.0 7.4 Upper East 1.8 1.8 0.8 0.8 1.8 0.0 0.0 1.8 Upper West 2.4 1.6 1.7 2.3 1.6 0.0 1.2 7.6 Education level None 3.5 1.3 1.1 0.6 0.6 0.0 0.3 4.3 Primary 7.7 3.6 2.6 4.0 1.2 0.2 0.8 12.1 Middle/JSS/JHS 7.0 2.0 2.3 2.4 0.5 0.3 0.9 10.2 Secondary 6.9 2.6 3.2 2.2 0.5 0.0 0.5 10.9 Technical 7.6 3.5 1.9 4.0 0.0 0.0 3.5 9.5 Higher 3.4 0.2 0.2 0.3 0.0 0.0 0.5 4.1 Asset quintile Lowest 6.5 3.1 3.1 2.4 1.6 0.4 0.3 10.0 Second 6.5 2.2 1.7 2.3 0.5 0.1 0.8 8.9 Middle 7.8 1.7 2.1 2.8 0.2 0.0 1.0 12.1 Fourth 4.7 1.6 2.0 1.1 0.5 0.2 0.3 6.4 Highest 5.5 1.7 1.4 2.2 0.0 0.0 1.6 7.8 Total 6.2 2.1 2.1 2.1 0.6 0.2 0.7 8.9 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 78 Table 12: Percentage of men who have experienced domestic or non-domestic physical violence over their lifetime Slapped or Pushed or Hit Kicked, Choked or Burnt Attacked Any thrown shoved dragged or strangled with physical things at beaten up weapon violence Age group (years) 15–19 47.2 30.9 22.8 15.9 3.7 2.1 3.7 64.3 20–24 38.6 17.9 20.1 16.9 1.4 3.1 5.1 54.2 25–29 37.4 16.5 16.9 12.0 3.6 2.1 6.9 47.5 30–39 37.9 18.3 21.6 14.5 4.6 1.6 10.2 50.9 40–49 39.0 20.8 22.6 16.7 3.3 1.3 8.3 49.5 50–60 32.6 22.6 21.9 17.5 3.3 1.9 3.8 42.9 Employment Self-employed 34.3 18.5 18.5 14.3 3.4 1.6 6.8 46.0 Employed 42.1 21.9 22.9 16.8 3.8 2.6 9.6 55.1 Not working 43.2 22.4 23.8 16.1 2.8 2.1 3.7 56.9 Marital status Never married 36.6 20.9 17.7 15.4 2.0 1.9 3.0 51.4 Married or living together 37.4 20.4 21.6 16.2 3.9 1.7 7.9 50.1 Divorced/separated/widowed 42.8 18.5 20.7 12.3 3.0 3.1 7.1 52.6 Residence Urban 38.3 19.0 19.8 13.7 3.9 2.0 8.0 50.9 Rural 38.2 21.4 21.8 16.9 2.9 1.9 5.9 50.7 79 Slapped or Pushed or Hit Kicked, Choked or Burnt Attacked Any thrown shoved dragged or strangled with physical things at beaten up weapon violence Region Western 46.1 16.7 17.3 13.9 3.8 1.4 12.2 60.0 Central 41.4 19.9 21.7 6.7 4.5 3.6 7.0 52.6 Greater Accra 26.7 15.2 17.9 5.8 2.8 2.1 5.0 35.5 Volta 44.6 33.4 35.5 32.1 1.8 0.0 2.5 59.5 Eastern 44.2 28.6 27.5 20.3 3.9 3.2 7.3 66.1 Ashanti 49.8 20.4 23.1 23.1 3.9 2.9 9.4 63.0 Brong Ahafo 35.0 20.4 19.1 6.7 6.2 3.6 10.2 48.7 Northern 31.3 16.2 12.9 18.7 0.0 0.0 3.3 41.6 Upper East 16.3 13.0 10.1 6.0 5.2 0.0 0.4 19.1 Upper West 14.3 14.3 7.1 1.5 3.2 0.0 7.0 21.7 Education level None 23.8 15.4 15.6 12.3 2.8 1.4 3.2 31.7 Primary 39.7 23.3 20.7 13.2 2.2 1.6 6.3 52.1 Middle/JSS/JHS 39.4 20.0 20.8 15.6 4.1 2.3 7.5 52.1 Secondary 44.4 21.3 25.0 16.0 3.7 2.3 6.7 59.2 Technical 44.4 19.7 19.9 14.8 0.0 1.4 6.8 57.8 Higher 38.3 21.2 20.1 18.8 3.7 1.5 9.8 51.2 Asset quintile Lowest 37.2 22.0 20.8 15.4 3.2 1.2 3.3 46.1 Second 36.3 20.3 22.5 16.7 2.9 2.7 5.9 51.4 Middle 40.9 18.3 22.6 15.4 4.6 1.6 7.8 53.1 Fourth 41.4 19.1 19.0 15.9 4.1 3.1 8.0 53.4 Highest 36.5 21.1 18.6 12.3 2.3 1.0 10.7 50.1 Total 38.3 20.2 20.8 15.3 3.4 2.0 6.9 50.8 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 80 Table 13: Percentage of men who experienced domestic or non-domestic physical violence in the last 12 months Slapped or Pushed or Hit Kicked, Choked or Burnt Attacked Any thrown shoved dragged or strangled with physical things at beaten up weapon violence Age group (years) 15–19 18.6 14.9 7.8 3.3 3.1 0.8 2.0 30.6 20–24 7.2 3.0 3.0 2.1 0.8 0.5 1.0 14.2 25–29 4.9 2.0 1.1 0.4 1.5 0.0 0.4 8.7 30–39 3.9 2.4 1.6 0.8 0.3 0.0 1.7 7.3 40–49 2.1 1.8 2.2 0.5 0.5 0.0 0.5 5.1 50–60 1.0 1.4 1.2 0.3 0.3 0.8 0.7 4.0 Employment Self-employed 3.0 1.7 1.9 0.7 0.7 0.2 1.0 6.5 Employed 5.2 3.2 1.7 0.7 0.6 0.0 1.1 9.9 Not working 11.4 8.1 4.5 2.3 1.9 0.7 1.1 19.4 Marital status Never married 10.9 8.4 5.1 1.8 1.1 0.9 0.3 18.6 Married or living together 3.6 2.3 1.8 1.0 0.8 0.2 1.1 7.7 Divorced/separated/widowed 6.6 2.8 2.2 0.5 1.1 0.0 1.6 11.0 Residence Urban 5.0 3.8 2.2 0.6 0.7 0.6 0.8 9.7 Rural 5.7 3.1 2.6 1.5 1.1 0.0 1.2 10.6 81 Slapped Pushed or Hit Kicked, Choked or Burnt Attacked Any or thrown shoved dragged or strangled with physical things at beaten up weapon violence Region Western 9.0 1.8 2.2 0.7 0.6 0.0 1.6 12.5 Central 9.0 3.4 1.8 1.8 1.4 1.2 1.4 14.9 Greater Accra 4.3 4.3 1.7 0.0 1.0 0.0 0.9 8.4 Volta 3.4 1.5 3.0 2.7 0.0 0.0 0.4 9.1 Eastern 6.1 4.8 3.6 0.0 0.0 0.0 0.6 12.3 Ashanti 6.2 4.4 3.3 2.6 1.9 1.0 1.5 14.0 Brong Ahafo 5.0 4.3 2.2 0.0 1.8 0.0 1.5 8.4 Northern 2.0 1.4 0.7 0.0 0.0 0.0 0.4 3.1 Upper East 3.5 4.9 3.1 1.1 1.1 0.0 0.0 5.6 Upper West 2.2 3.8 1.7 0.0 0.0 0.0 1.7 6.1 Education level None 3.1 1.2 1.2 0.3 0.6 0.0 0.0 3.8 Primary 5.3 7.2 3.6 1.8 1.2 0.7 1.0 11.1 Middle/JSS/JHS 5.6 3.3 2.8 1.1 0.9 0.3 1.7 11.0 Secondary 8.7 4.4 3.4 1.9 1.6 0.4 1.3 16.2 Technical 6.3 3.4 0.0 0.0 0.0 0.0 0.6 9.1 Higher 2.3 1.6 0.7 0.0 0.3 0.0 0.3 5.0 Asset quintile Lowest 7.3 2.9 2.0 0.6 1.1 0.0 1.3 10.8 Second 6.9 4.3 2.5 1.4 1.6 0.3 1.1 13.1 Middle 4.1 2.5 2.0 1.2 0.4 0.0 0.7 6.8 Fourth 3.3 2.5 2.8 1.8 0.5 1.0 0.5 9.1 Highest 4.4 4.9 2.6 0.0 0.6 0.0 1.6 9.6 Total 5.4 3.5 2.4 1.0 0.9 0.3 1.0 10.2 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 82 4.1.3. Sexual violence Table 5 shows that 30 per cent of women and 23.1 per cent of men experienced sexual violence at least once over their lifetime, and 10.6 per cent of women and 9.5 per cent of men reported having experienced sexual violence at least once over the 12 months that preceded the survey. Tables 14 to 17 provide further disaggregation of the data on the incidence of sexual violence, as discussed below. Dimensions of sexual violence. The most prevalent form of sexual violence among both women and men over their lifetimes was being subject to sexual comments (18 per cent among women and 12.1 per cent among men) (Tables 14 and 16). This was followed by unwanted touches (15.3 per cent among women and 10.3 per cent among men) and being physically forced to have sex (9 per cent among women and 6.9 per cent among men). These patterns of sexual violence were very similar across the lifetime of individuals and in the 12 months prior to the survey (Tables 15 and 17). Age patterns. Similar to social and physical violence, the incidence of lifetime sexual violence was higher among younger women: 38.2 per cent of women aged 15–19 years, 40.4 per cent of women aged 20–24 years and 38.3 per cent of women aged 30–39 years reported having experienced at least one act of sexual violence. In contrast, 18.8 per cent of women aged 50–60 years experienced these forms of violence (Table 14). The incidence of sexual violence was highest among men aged 30–39 years (29.7 per cent) and men aged 20– 24 years (29.2 per cent). The lowest incidence was again among the oldest male age group (13.8 per cent) (Table 16). These age patterns were also observed when considering the 12 months prior to the survey (Tables 15 and 17). Employment status. Employed women and men (41.9 and 31.5 per cent, respectively) were more likely to have experienced sexual violence than men and women not in employment (Tables 14 and 16). This is an issue that will be explored further when focusing the analysis on domestic violence. These patterns were very similar across the lifetime of individuals and in the 12 months prior to the survey (Tables 15 and 17). Marital status. As observed with social and physical violence, the incidence of sexual violence was more prominent among women and men who were divorced, separated or widowed at the time of the survey (31.7 per cent of women and 29.4 per cent of men) (Table 14). The differences in relation to other marital status were considerable, with 22.6 per cent of never-married women and 16.9 per cent of never-married men having reported sexual violence (Table 16). As with other forms of violence, the incidence of sexual violence in the 12 months prior to the survey was highest among never-married women (Table 17). This may indicate a generational effect, which will be explored further in the next sections. Geographical patterns. Contrary to other forms of violence analysed above, the incidence of sexual violence was consistently higher among women and men living in urban areas (32.1 per cent of women and 28.5 per cent of men, compared with 27.8 and 18.1 per cent in 83 rural areas, respectively) (Tables 14 and 16). As with other forms of violence, the incidence of sexual violence among women was higher in the Eastern region (45.5 per cent) than in any other region. The highest levels of incidence of sexual violence among men were observed in the Ashanti region (32.4 per cent). The lowest incidence of sexual violence against women and men was in the Upper East region (8.8 per cent and 6.7 per cent, respectively). These geographical patterns were very similar when considering the incidence of sexual violence in the 12 months prior to the survey (Tables 15 and 17). Education levels and asset ownership. Women with no formal education were about half as likely as women with some level of education to report sexual violence. Women with secondary (39.1 per cent) and higher education (37.1 per cent) were more likely to have experienced sexual violence than women with primary (32.8 per cent) or middle levels of education (32.0 per cent) (Table 14). Men with higher levels of education were more likely to have experienced sexual violence over their lifetime: 34.9 per cent, compared to 20.1 per cent among men with primary education. Men with no formal education were the least likely to report sexual violence over their lifetime (7.8 per cent) (Table 16). Reflecting this latter pattern, sexual violence against men was also more prominent at the top end of the asset ownership distribution. The proportion of men experiencing any type of sexual violence over their lifetime was twice as high for male respondents in the highest asset quintile (29.3 per cent) than for those in the lowest quintile (15.1 per cent) (Table 16). This association was similar for women, but the differences across quintiles were less pronounced (Table 14). These patterns were also observed when considering the incidence of sexual violence in the 12 months prior to the survey (Tables 15 and 17). 84 Table 14: Percentage of women who have experienced domestic or non-domestic sexual violence over their lifetime Sexual Sexual Physically Otherwise Sex Sex Sex Any comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Age group (years) 15–19 20.6 26.4 10.3 3.5 3.3 2.0 0.9 38.2 20–24 25.6 18.6 11.8 1.4 7.9 3.4 5.1 40.4 25–29 19.8 20.2 14.0 2.1 6.6 4.1 6.1 38.3 30–39 17.8 14.2 8.9 1.6 5.9 2.7 4.6 29.5 40–49 13.8 10.5 6.2 1.4 3.5 0.5 4.1 21.4 50–60 13.8 9.6 3.2 1.6 2.1 1.3 2.5 18.8 Employment Self-employed 16.4 13.8 7.9 1.6 4.8 2.2 4.2 27.0 Employed 26.2 21.2 15.2 1.9 6.7 4.4 4.6 41.9 Not working 18.8 16.8 9.2 2.1 5.4 1.9 4.5 32.9 Marital status Never married 14.6 14.7 2.0 0.9 0.9 0.0 0.0 22.6 Married or living together 18.0 14.9 9.2 1.7 5.2 2.3 4.6 30.1 Divorced/separated/widowed 18.9 16.9 10.0 2.3 5.8 3.4 4.3 31.7 Residence Urban 19.0 16.0 10.1 1.6 5.3 2.9 5.2 32.1 Rural 16.9 14.5 7.8 1.9 4.9 1.8 3.2 27.8 85 Sexual Sexual Physically Otherwise Sex Sex Sex Any comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Region Western 15.2 13.0 9.3 2.3 2.9 1.8 1.6 26.7 Central 19.7 19.6 8.8 1.4 8.3 6.2 6.6 35.5 Greater Accra 14.7 12.9 9.3 1.0 6.9 3.0 7.1 28.1 Volta 17.5 13.5 9.0 4.4 5.7 0.6 8.5 29.9 Eastern 26.7 26.8 14.6 2.2 7.0 5.0 6.9 45.5 Ashanti 19.9 14.5 8.6 2.1 4.3 0.9 1.7 30.5 Brong Ahafo 24.7 16.6 9.9 0.9 4.3 1.8 1.9 32.7 Northern 10.9 13.4 3.0 0.6 2.0 0.3 0.6 19.5 Upper East 4.7 4.4 5.0 0.0 1.1 1.7 0.7 8.8 Upper West 6.0 3.6 3.5 2.1 1.1 1.1 3.5 12.9 Education level None 10.6 8.7 3.9 1.3 3.4 1.1 2.6 17.4 Primary 19.9 18.2 10.5 2.0 6.4 2.6 4.6 32.8 Middle/JSS/JHS 20.0 16.6 9.8 1.9 5.3 2.3 4.4 32.0 Secondary 18.9 18.1 11.6 1.9 7.2 3.5 5.8 39.1 Technical 20.1 15.8 11.7 1.2 4.6 3.6 8.3 35.2 Higher 25.7 18.8 12.2 1.5 3.3 4.2 3.9 37.1 Asset quintile Lowest 15.6 15.7 7.5 0.9 4.3 1.8 3.9 27.6 Second 17.0 15.4 8.6 2.1 4.1 2.1 3.8 27.8 Middle 20.1 16.3 10.4 2.6 6.1 2.6 4.3 32.1 Fourth 17.4 14.0 8.5 1.7 5.0 3.1 4.6 31.2 Highest 21.5 15.6 10.7 1.3 7.2 2.5 5.1 33.3 Total 18.0 15.3 9.0 1.7 5.1 2.4 4.3 30.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 86 Table 15: Percentage of women who experienced domestic or non-domestic sexual violence in the last 12 months Sexual Sexual Physically Otherwise Sex Sex Sex Any comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Age group (years) 15–19 11.9 15.4 3.1 0.9 0.4 0.5 1.6 22.1 20–24 13.1 8.1 2.2 0.1 1.6 1.8 1.1 20.2 25–29 7.3 6.1 1.8 0.5 0.9 1.8 1.4 13.7 30–39 7.0 2.5 0.8 0.2 0.7 1.1 0.8 10.1 40–49 2.3 2.0 0.2 0.0 0.1 0.4 0.0 4.2 50–60 1.9 0.9 0.0 0.3 0.0 0.0 0.0 2.1 Employment Self-employed 4.8 3.2 1.1 0.2 0.7 0.7 0.6 8.2 Employed 11.9 5.7 0.9 0.0 0.0 1.7 1.1 16.0 Not working 8.9 7.0 1.0 0.5 0.6 1.4 0.8 14.7 Marital status Never married 7.6 8.5 0.3 0.3 0.0 0.0 0.0 12.8 Married or living together 5.8 3.9 1.3 0.3 0.8 1.2 0.8 10.0 Divorced/separated/widowed 8.9 5.2 0.5 0.2 0.1 0.3 0.7 12.2 Residence Urban 7.6 4.3 1.1 0.1 0.5 1.5 1.0 11.7 Rural 5.4 4.5 1.1 0.4 0.7 0.5 0.4 9.3 87 Sexual Sexual Physically Otherwise Sex Sex Sex Any comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Region Western 5.8 4.2 1.5 0.0 0.3 0.4 0.6 7.9 Central 6.0 5.2 1.6 0.5 1.2 2.0 1.4 13.5 Greater Accra 7.1 4.4 0.9 0.1 1.0 2.2 1.3 11.5 Volta 2.5 2.6 1.7 0.6 0.0 1.0 0.3 7.1 Eastern 6.9 8.0 1.2 0.3 0.6 0.9 0.6 12.5 Ashanti 8.1 3.5 1.3 0.6 0.8 0.7 0.1 12.5 Brong Ahafo 11.7 6.2 0.3 0.0 0.3 0.0 1.0 14.7 Northern 2.3 2.2 0.8 0.0 0.3 0.0 0.0 3.1 Upper East 0.6 0.0 0.0 0.0 0.0 0.0 0.6 1.2 Upper West 3.6 2.4 0.0 0.0 0.0 0.0 1.1 5.9 Education level None 1.8 1.6 0.5 0.4 0.1 0.0 0.1 3.6 Primary 6.6 5.2 1.1 0.0 0.7 1.2 0.1 10.9 Middle/JSS/JHS 7.8 5.1 1.3 0.2 0.9 0.9 0.9 12.2 Secondary 7.8 4.7 1.9 0.6 0.8 2.4 1.5 14.7 Technical 9.0 5.0 1.4 0.0 1.4 3.3 1.4 12.3 Higher 12.6 7.0 0.0 0.0 0.0 0.7 1.2 15.9 Asset quintile Lowest 5.2 4.3 1.3 0.2 0.1 0.5 0.3 9.6 Second 5.9 5.0 1.0 0.1 0.7 0.9 0.6 9.9 Middle 7.9 5.2 1.7 0.4 0.4 0.4 0.7 11.2 Fourth 7.7 4.1 0.9 0.5 1.2 1.9 1.0 12.9 Highest 6.3 3.0 0.7 0.0 0.5 1.2 1.0 9.1 Total 6.6 4.4 1.1 0.3 0.6 1.0 0.7 10.6 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 88 Table 16: Percentage of men who have experienced domestic or non-domestic sexual violence over their lifetime Sexual Sexual Physically Otherwise Sex Sex Sex Any comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Age group (years) 15–19 14.0 5.8 2.6 0.8 0.8 0.0 0.7 19.4 20–24 17.3 15.5 9.9 2.4 2.6 0.5 3.7 29.2 25–29 11.4 10.1 7.0 3.4 3.3 2.0 3.8 25.1 30–39 14.6 14.0 8.0 2.7 2.8 1.4 3.7 29.7 40–49 7.0 7.6 6.4 0.5 1.6 2.2 1.3 16.5 50–60 8.0 5.0 5.6 1.2 1.4 0.2 0.8 13.8 Employment Self-employed 9.3 7.7 5.6 0.9 1.6 1.0 1.8 18.3 Employed 13.6 15.4 10.9 4.8 4.7 1.8 5.7 31.5 Not working 17.1 10.1 5.3 1.3 0.9 0.9 0.8 24.6 Marital status Never married 12.5 7.1 3.8 1.8 2.7 0.0 0.3 16.9 Married or living together 10.9 9.9 7.4 2.0 2.2 1.2 2.9 22.9 Divorced/separated/widowed 15.6 13.9 8.2 2.2 2.4 2.6 3.7 29.4 Residence Urban 16.1 12.8 8.2 3.0 2.9 1.1 2.4 28.5 Rural 8.3 7.8 5.7 1.0 1.7 1.3 2.7 18.1 89 Sexual Sexual Physically Otherwise Sex Sex Sex Any comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Region Western 9.4 15.0 10.6 2.6 1.1 1.1 2.8 26.0 Central 12.8 5.7 7.2 1.7 7.4 4.9 3.7 24.2 Greater Accra 14.1 10.3 7.8 2.8 3.2 1.3 3.6 26.3 Volta 10.0 5.7 4.1 1.7 1.4 0.0 0.9 15.5 Eastern 14.7 15.9 9.0 2.1 4.6 1.4 7.6 29.8 Ashanti 17.8 14.1 9.7 2.3 2.2 1.5 2.0 32.4 Brong Ahafo 13.3 8.4 2.8 2.7 0.6 0.0 0.0 17.6 Northern 4.2 4.9 3.2 0.4 0.0 0.8 1.0 12.6 Upper East 3.0 5.7 2.0 0.0 0.0 0.0 1.1 6.7 Upper West 3.5 3.9 3.5 0.0 0.0 0.0 1.7 7.3 Education level None 1.8 3.5 2.6 0.3 0.9 0.0 0.7 7.8 Primary 10.0 6.6 6.0 1.8 2.8 1.6 0.7 20.1 Middle/JSS/JHS 12.8 8.5 5.5 0.8 0.9 1.0 2.3 22.3 Secondary 16.4 12.2 8.5 3.1 2.7 1.2 4.7 28.7 Technical 13.2 8.3 13.0 4.2 6.0 2.2 5.9 24.4 Higher 15.9 22.3 12.1 4.8 5.2 2.6 3.1 34.9 Asset quintile Lowest 7.6 5.5 4.5 1.3 0.9 1.3 0.8 15.1 Second 10.5 7.9 5.2 1.1 2.9 1.1 1.7 18.8 Middle 13.6 9.2 6.0 0.5 1.7 1.0 3.3 22.5 Fourth 16.3 16.6 11.6 3.5 2.3 1.9 3.4 32.6 Highest 13.6 13.5 8.0 4.2 3.4 0.7 4.6 29.3 Total 12.1 10.3 6.9 2.0 2.3 1.2 2.6 23.1 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 90 Table 17: Percentage of men who experienced domestic or non-domestic sexual violence in the last 12 months Sexual Sexual Physically Otherwise Sex Sex Sex Any comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Age group (years) 15–19 10.8 4.8 1.9 0.8 0.8 0.7 0.0 14.5 20–24 7.9 6.3 1.8 1.6 0.3 1.6 0.5 14.2 25–29 5.8 4.7 1.7 1.8 0.8 1.1 0.4 12.2 30–39 6.0 3.3 0.7 0.9 0.3 1.1 0.7 10.2 40–49 2.6 1.4 2.1 0.0 0.0 0.0 0.2 5.8 50–60 0.0 0.0 0.7 0.0 0.0 0.0 0.0 0.7 Employment Self-employed 3.2 2.0 0.9 0.5 0.1 0.3 0.1 5.9 Employed 6.9 5.6 2.9 1.5 0.7 2.3 0.5 14.8 Not working 8.3 4.0 1.0 1.1 0.4 0.1 0.7 11.9 Marital status Never married 8.4 4.3 1.3 1.3 1.0 0.0 0.0 11.5 Married or living together 4.4 2.9 1.3 0.7 0.1 0.7 0.3 8.2 Divorced/separated/widowed 5.5 4.5 2.1 1.3 0.7 1.8 0.7 12.5 Residence Urban 7.8 4.3 1.7 1.4 0.5 0.7 0.3 12.8 Rural 2.9 2.5 1.1 0.4 0.2 0.9 0.4 6.4 91 Sexual Sexual Physically Otherwise Sex Sex Sex Any comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Region Western 0.3 2.0 2.8 0.6 0.0 0.7 0.7 6.5 Central 5.8 2.0 0.7 0.7 1.2 1.9 1.4 10.9 Greater Accra 7.3 4.7 1.8 2.1 0.4 1.3 0.3 13.2 Volta 2.9 2.7 1.5 0.9 0.5 0.5 0.0 5.7 Eastern 4.3 3.9 0.7 0.7 0.7 2.5 0.0 9.4 Ashanti 9.4 4.7 2.1 0.6 0.4 0.4 0.2 14.2 Brong Ahafo 8.6 5.0 0.9 1.5 0.0 0.0 0.0 10.5 Northern 2.7 1.1 0.0 0.0 0.0 0.0 0.8 4.2 Upper East 0.4 1.4 0.4 0.0 0.0 0.0 0.0 1.8 Upper West 0.0 2.2 1.7 0.0 0.0 0.0 0.0 3.8 Education level None 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Primary 6.0 3.0 1.5 1.0 1.3 0.2 0.0 10.6 Middle/JSS/JHS 4.7 2.5 0.8 0.7 0.0 1.0 0.5 8.9 Secondary 7.8 6.4 2.1 1.7 0.2 1.4 0.3 13.7 Technical 8.4 5.9 3.8 2.7 2.2 0.0 0.0 11.7 Higher 7.2 4.4 3.0 0.5 0.4 0.8 0.8 13.1 Asset quintile Lowest 4.1 2.0 1.3 0.7 0.2 0.6 0.7 6.8 Second 3.7 2.6 0.6 0.7 0.6 1.0 0.2 6.9 Middle 5.0 2.3 0.6 0.0 0.0 0.7 0.2 8.3 Fourth 7.8 5.6 3.1 2.0 0.4 0.7 0.3 13.9 Highest 6.6 4.8 1.8 0.9 0.4 1.0 0.4 13.2 Total 5.3 3.4 1.4 0.9 0.4 0.8 0.3 9.5 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 92 4.1.4. Psychological violence Psychological violence was the most common type of violence experienced by women and men across Ghana. About half of the respondents (51.9 per cent of men and 48 per cent of women) experienced an act of psychological violence at least once in their lifetime (Table 5), while 22.5 per cent of women and 28 per cent of men experienced at least one type of psychological violence in the 12 months prior to the survey (Table 6). Further disaggregated information on the incidence of psychological violence is provided in Tables 18–21 and discussed in the paragraphs below. Dimensions of psychological violence. The most common type of psychological violence reported by both men and women over their lifetime was being insulted, humiliated or belittled. Almost 46 per cent of women (Table 18) and 48.8 per cent of men (Table 20) in the sample reported having experienced this form of violence. Other forms of psychological violence affected only small numbers of women and men in Ghana over their lifetime and in the 12 months prior to the survey (Tables 19–21). Age patterns. The results show that age was less strongly correlated to lifetime experiences of psychological violence than to the other types of violence discussed above, with no clear pattern of association for either women or men (Tables 18 and 20). But, in line with discussions above for other forms of violence, younger women and men (aged 15–19 years) were more likely to have reported at least one incident of psychological violence in the 12 months prior to the survey than all other older age groups: 38.8 per cent and 38.9 per cent of women and men, respectively, aged 15–19 years experienced psychological violence in that period, in comparison to 14.8 per cent of women and 18.8 per cent of men aged 50–60 years (Tables 19 and 21). Employment status. The incidence of lifetime experiences of psychological violence was highest among self-employed women (49.5 per cent) and employed men (56.1 per cent). It was lowest among women who were not working (44.7 per cent) and among self-employed men (49.2 per cent) (Tables 18 and 20). These employment patterns were also observed when considering the incidence of psychological violence in the 12 months prior to the survey (Tables 19–21). Marital status. Men and women who were divorced, separated or widowed were about 6 percentage points more likely to have experienced psychological violence over their lifetime than all other marital statuses on average (Tables 18 and 20). The lifetime incidence of psychological violence across marital status was similar to what was found for the other forms of violence analysed so far, and when considering the 12 months prior to the survey (Tables 19–21). Geographical patterns. Women and men living in rural areas (50.3 and 52.8 per cent, respectively) were more likely to have reported at least one type of psychological violence over their lifetime than those living in urban areas (45.8 and 51 per cent, respectively). As 93 with other forms of violence, the incidence of psychological violence was highest in the Eastern region (73.7 per cent for women and 70.3 per cent for men). The lowest incidence of lifetime psychological violence was observed in the Upper East region for women (13.1 per cent) and in the Upper West region for men (23.9 per cent) (Tables 18 and 20). These geographical patterns were similar when considering the incidence of psychological violence in the 12 months prior to the survey (Tables 19–21). Education levels and asset ownership. Women without formal education and women with higher education were up to 6.8 percentage points less likely to report having experienced psychological violence than women with intermediate levels of education (primary, middle/junior secondary school (JSS)/junior high school (JHS), secondary and technical education) (Table 18). The incidence of psychological violence among men steadily increased with the level of education until secondary/technical education. Men with higher levels of education were as likely as those with primary education to experience psychological violence (Table 20). Women in the highest asset quintile were 7.3 percentage points less likely to report having experienced psychological violence over their lifetime than women in the lowest asset quintile, whereas the incidence of psychological violence was highest among men in the highest asset quintile (54.5 per cent). These patterns were similar when considering the 12 months prior to the survey (Tables 19–21). 94 Table 18: Percentage of women who have experienced domestic or non-domestic psychological violence over their lifetime Insulted, Ignored or Scared or Threatened Threatened to Any humiliated or threatened to be intimidated with the use be hurt psychological belittled abandoned on purpose of a weapon violence Age group (years) 15–19 45.9 11.3 3.3 1.8 0.3 49.3 20–24 46.0 10.1 4.3 1.0 0.8 47.9 25–29 48.2 8.5 4.7 2.1 0.8 50.1 30–39 42.4 7.8 4.3 3.6 1.8 45.0 40–49 48.4 8.5 6.6 2.6 1.1 50.5 50–60 47.0 7.0 7.6 1.9 2.4 47.6 Employment Self-employed 47.5 8.4 5.5 2.6 1.7 49.5 Employed 42.6 9.8 6.6 3.6 0.6 46.2 Not working 43.0 7.9 3.8 1.4 0.9 44.7 Marital status Never married 39.9 4.1 2.6 0.8 0.0 41.0 Married or living together 44.7 7.7 5.0 2.1 1.0 46.7 Divorced/separated/widowed 51.4 12.1 6.7 4.2 3.1 54.0 Residence Urban 44.0 8.4 5.3 2.1 1.0 45.8 Rural 48.1 8.4 5.1 2.9 1.8 50.3 95 Insulted, Ignored or Scared or Threatened Threatened to Any humiliated or threatened to be intimidated with the use be hurt psychological belittled abandoned on purpose of a weapon violence Region Western 54.0 4.7 1.7 1.1 1.0 56.7 Central 45.6 16.3 3.7 2.1 1.8 50.2 Greater Accra 36.4 7.2 5.3 2.3 1.0 37.7 Volta 59.1 10.9 20.7 1.2 2.9 59.5 Eastern 71.1 12.3 7.0 5.6 3.0 73.7 Ashanti 47.6 8.2 1.1 2.6 0.9 48.9 Brong Ahafo 38.1 8.5 5.2 3.7 0.7 41.4 Northern 41.7 2.8 5.2 1.0 1.0 42.8 Upper East 12.0 0.0 1.1 0.0 1.1 13.1 Upper West 14.9 5.3 1.5 1.2 0.0 17.3 Education level None 42.4 7.5 6.0 2.3 1.4 44.1 Primary 49.7 9.5 4.3 2.9 2.7 51.5 Middle/JSS/JHS 47.5 9.1 4.9 2.4 0.8 49.8 Secondary 43.5 7.5 4.3 1.0 0.8 45.4 Technical 49.5 6.1 10.5 8.2 4.0 49.5 Higher 42.0 7.8 7.1 2.4 1.1 45.1 Asset quintile Lowest 51.8 10.2 8.8 3.2 2.4 53.5 Second 44.2 9.0 3.6 2.9 1.3 46.8 Middle 50.9 7.1 3.6 2.1 0.6 52.6 Fourth 40.7 6.3 3.8 1.0 1.1 42.3 Highest 43.7 9.8 7.4 3.2 1.6 46.2 Total 45.9 8.4 5.2 2.4 1.4 48.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 96 Table 19: Percentage of women who experienced domestic or non-domestic psychological violence in the last 12 months Insulted, Ignored or Scared or Threatened Threatened to Any humiliated threatened to be intimidated with the use be hurt psychological or belittled abandoned on purpose of a weapon violence Age group (years) 15–19 37.0 7.7 1.9 1.0 0.0 38.8 20–24 24.1 5.1 0.7 0.0 0.4 26.4 25–29 24.4 3.0 1.4 1.0 0.3 25.6 30–39 19.8 3.0 0.9 1.2 0.8 21.5 40–49 16.6 2.3 1.2 0.6 0.4 19.0 50–60 14.2 0.7 0.8 0.6 0.0 14.8 Employment Self-employed 19.4 2.6 0.7 0.7 0.4 20.9 Employed 18.8 3.2 2.3 1.4 0.0 21.7 Not working 25.8 4.4 1.5 0.9 0.5 27.4 Marital status Never married 28.3 3.5 1.7 0.8 0.0 29.3 Married or living together 19.7 3.0 0.8 0.7 0.3 21.3 Divorced/separated/widowed 23.2 3.4 1.7 1.0 0.9 24.8 Residence Urban 21.1 3.2 1.2 0.6 0.3 22.4 Rural 20.7 2.9 0.8 1.0 0.6 22.6 97 Insulted, Ignored or Scared or Threatened Threatened to Any humiliated threatened to be intimidated with the use be hurt psychological or belittled abandoned on purpose of a weapon violence Region Western 25.8 0.9 0.4 0.4 0.3 27.0 Central 22.0 7.2 0.9 1.1 0.7 25.5 Greater Accra 21.3 3.0 2.0 0.8 0.2 22.3 Volta 22.7 3.4 2.4 0.4 0.0 24.3 Eastern 28.7 4.0 2.5 2.6 1.2 32.1 Ashanti 21.0 3.1 0.2 0.6 0.3 22.7 Brong Ahafo 15.5 3.0 0.3 0.7 0.7 16.8 Northern 16.4 0.7 0.3 0.0 0.3 16.4 Upper East 4.9 0.0 0.0 0.0 0.0 4.9 Upper West 11.1 1.7 0.0 0.0 0.0 12.3 Education level None 16.6 2.1 1.0 0.3 0.4 18.0 Primary 21.6 4.2 1.1 1.4 1.0 24.3 Middle/JSS/JHS 23.6 3.1 0.8 0.7 0.2 24.8 Secondary 22.6 4.1 1.8 0.6 0.3 24.7 Technical 16.4 2.0 0.9 4.8 1.4 17.6 Higher 16.6 2.0 1.5 0.6 0.2 18.4 Asset quintile Lowest 22.1 3.1 1.6 1.2 0.7 25.0 Second 19.0 3.3 0.5 0.8 0.6 20.4 Middle 26.8 3.8 0.9 0.6 0.3 27.9 Fourth 20.5 2.2 1.4 0.2 0.2 21.3 Highest 17.1 3.3 0.9 1.2 0.2 19.0 Total 20.9 3.1 1.1 0.8 0.4 22.5 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 98 Table 20: Percentage of men who have experienced domestic or non-domestic psychological violence over their lifetime Insulted, Ignored or Scared or Threatened Threatened to Any humiliated threatened to be intimidated with the use be hurt psychological or belittled abandoned on purpose of a weapon violence Age group (years) 15–19 50.2 6.8 8.2 2.0 0.8 55.6 20–24 50.6 7.5 6.1 3.2 2.5 53.7 25–29 47.4 4.8 5.9 1.7 3.2 51.0 30–39 45.9 6.8 9.7 6.9 2.7 49.6 40–49 48.3 6.2 6.9 3.2 1.9 49.5 50–60 53.0 7.6 9.1 2.6 2.5 55.4 Employment Self-employed 46.6 5.6 7.4 3.5 2.3 49.2 Employed 53.1 7.7 8.5 4.2 3.1 56.1 Not working 48.7 7.4 7.4 3.1 1.8 53.5 Marital status Never married 40.7 6.4 6.3 2.9 1.1 44.4 Married or living together 49.5 6.4 7.5 3.7 2.3 52.2 Divorced/separated/widowed 53.9 7.5 9.0 3.1 3.5 58.1 Residence Urban 47.5 6.5 7.4 3.8 3.1 51.0 Rural 50.0 6.6 8.0 3.4 1.8 52.8 99 Insulted, Ignored or Scared or Threatened Threatened Any humiliated threatened to be intimidated with the use of to be hurt psychological or belittled abandoned on purpose a weapon violence Region Western 57.8 6.6 7.6 5.7 2.7 61.9 Central 47.1 21.8 9.0 7.1 4.8 53.5 Greater Accra 41.1 3.8 5.9 2.9 2.9 44.6 Volta 57.1 7.8 23.9 1.5 0.0 59.2 Eastern 67.9 7.8 5.2 3.5 3.9 70.3 Ashanti 53.0 5.5 4.5 5.3 3.5 56.5 Brong Ahafo 41.3 6.9 9.4 3.8 2.4 45.2 Northern 42.2 3.1 4.8 0.2 0.0 43.1 Upper East 22.8 1.7 1.7 1.2 0.0 24.6 Upper West 23.9 0.0 0.0 1.5 0.0 23.9 Education level None 39.4 3.9 5.7 1.2 0.3 40.9 Primary 49.4 7.7 8.2 2.9 2.8 53.0 Middle/JSS/JHS 47.7 5.7 6.5 4.0 2.1 51.2 Secondary 53.6 8.6 10.0 4.2 3.7 56.5 Technical 63.7 7.5 11.0 3.8 0.0 70.1 Higher 49.6 6.9 8.3 4.7 3.5 52.0 Asset quintile Lowest 47.1 5.5 10.7 2.3 0.2 49.7 Second 47.9 8.1 6.7 3.3 2.6 50.9 Middle 50.9 7.1 8.3 3.7 4.4 54.3 Fourth 47.7 6.6 5.9 3.9 2.0 51.2 Highest 51.2 4.8 7.0 5.3 3.3 54.5 Total 48.8 6.5 7.7 3.6 2.4 51.9 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 100 Table 21: Percentage of men who experienced domestic or non-domestic psychological violence in the last 12 months Insulted, Ignored or Scared or Threatened Threatened to Any humiliated threatened to be intimidated with the use be hurt psychological or belittled abandoned on purpose of a weapon violence Age group (years) 15–19 36.1 4.0 2.4 1.2 0.8 38.9 20–24 34.0 4.3 2.0 1.5 1.5 38.2 25–29 28.0 1.5 2.5 0.5 1.7 29.5 30–39 22.4 2.9 3.2 1.7 1.1 25.2 40–49 20.9 2.0 0.8 0.6 0.5 22.5 50–60 17.9 1.8 0.6 0.3 0.0 18.8 Employment Self-employed 21.3 2.0 1.4 1.0 0.6 23.0 Employed 30.4 2.9 3.3 0.9 1.8 33.0 Not working 30.7 3.9 2.0 1.2 0.8 34.1 Marital status Never married 25.6 3.6 1.2 1.8 0.2 27.9 Married or living together 24.0 2.5 2.1 0.9 1.0 26.4 Divorced/separated/widowed 32.3 2.5 2.5 0.5 1.6 34.3 Residence Urban 26.1 2.4 1.8 1.2 1.5 28.7 Rural 25.4 2.9 2.2 0.8 0.5 27.4 101 Insulted, Ignored or Scared or Threatened Threatened to Any humiliated threatened to be intimidated with the use be hurt psychological or belittled abandoned on purpose of a weapon violence Region Western 31.0 1.3 1.3 0.7 0.4 32.3 Central 28.2 8.3 5.1 1.9 2.3 36.4 Greater Accra 26.0 2.2 3.0 1.4 1.9 28.2 Volta 24.4 4.6 1.5 0.5 0.0 26.5 Eastern 37.7 3.4 1.7 0.9 2.1 41.5 Ashanti 30.0 2.3 1.4 1.1 0.6 31.7 Brong Ahafo 17.4 1.1 2.7 2.0 1.4 18.8 Northern 17.7 1.1 1.2 0.0 0.0 18.4 Upper East 10.2 1.7 1.1 0.0 0.0 11.9 Upper West 6.8 0.0 0.0 0.0 0.0 6.8 Education level None 16.8 0.8 0.3 0.3 0.3 17.2 Primary 25.3 3.4 3.1 0.3 1.2 28.9 Middle/JSS/JHS 26.0 2.1 1.4 1.5 0.6 28.1 Secondary 31.0 4.8 2.8 1.3 1.1 33.0 Technical 26.4 6.3 0.6 0.0 0.0 32.6 Higher 26.5 1.4 3.6 0.9 2.5 29.4 Asset quintile Lowest 22.2 2.1 0.9 0.2 0.2 23.3 Second 26.7 3.6 2.7 1.2 0.7 29.7 Middle 24.9 3.1 1.0 0.0 1.2 27.3 Fourth 28.2 2.8 2.5 2.0 1.1 30.4 Highest 26.5 1.3 2.9 1.4 1.9 29.2 Total 25.7 2.7 2.0 1.0 1.0 28.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 102 4.1.5. Economic violence About one quarter of respondents – 28.2 and 25.8 per cent of women and men, respectively – experienced an act of economic violence at least once during their lifetime; about one in eight men and women experienced at least one type of economic violence in the 12 months preceding the survey (Table 5). Tables 22–25 provide disaggregated information on the incidence of economic violence across different population groups and socio-economic characteristics, which is discussed in the paragraphs below. Dimensions of economic violence. The most common form of economic violence experienced by both women and men over their lifetime was the denial of chop money: 20.2 per cent of women and 12.2 per cent of men reported this form of violence (Tables 22 and 24). This was also the most prevalent form of violence reported by women and men in the 12 months prior to the survey (11.6 and 5.7 per cent, respectively) (Tables 23 and 25). Age patterns. Younger age is again a risk factor for lifetime economic violence, particularly among men. Around 30 per cent of men aged 15–29 years reported having experienced economic violence in their lifetime, compared to 23.9 per cent of men aged 30–39 years, 21.6 per cent of men aged 40–49 years and 20.1 per cent of men aged 50–60 years (Table 24). The association between age and incidence of economic violence was less pronounced among women. The highest incidence was among women aged 25–29 years (33.9 per cent), while the lowest was among women aged 50–60 years (23.6 per cent) (Table 22). These age patterns were also observed when considering the incidence of economic violence in the 12 months prior to the survey (Table 23–25). Employment status. Just over 30 per cent of women who were not working at the time of the survey experienced economic violence, compared to 28.3 per cent of self-employed women and 23.1 per cent of employed women (Table 22). In contrast, economic violence was more prevalent among employed men: 29.4 per cent of employed men reported having experienced economic violence, compared to 28.3 per cent of men not working and 22.9 per cent of self-employed men (Table 24). These results suggest that employed men were at higher risk of economic violence, perhaps because economic expectations may be higher among these men. On the other hand, economic violence was more prevalent among women who stay at home, perhaps as a result of their economic dependence on men, as discussed in the literature surveyed in Section 2. The same patterns were observed when the survey questions referred to the 12 months prior to the survey (Tables 23–25). Marital status. As with other forms of violence, lifetime experiences of economic violence were also more prominent among men and women who were divorced, separated or widowed at the time of the survey. Thirty-three per cent of divorced, separated or widowed women experienced economic violence, compared to 18.6 per cent of never- married women and 27.6 per cent of married women or women living with a partner (Table 22). Almost 35 per cent of divorced, separated or widowed men experienced economic violence, compared to 25.2 per cent of never-married men and 23.4 per cent of men married 103 or living with a partner (Table 24). These patterns were similar when considering the 12 months prior to the survey (Tables 23–25). Geographical patterns. The incidence of economic violence was higher among women living in rural areas (31.3 per cent) than women living in urban areas (25.5 per cent) (Table 22), whereas there was almost no difference in the lifetime incidence of economic violence among men living in urban or rural areas (Table 24). As with other forms of violence, both women and men living in the Eastern region (48.7 per cent and 51.9 per cent, respectively) were more likely than individuals in other regions to have experienced at least one form of economic violence over their lifetime. This distribution of the incidence of economic violence across locations was similar for the 12 months prior to the survey (Tables 23–25). Education levels and asset ownership. The incidence of economic violence was lower among women with higher education (19.4 per cent) than among women with no formal education or intermediate levels of education (primary, middle/JSS/JHS, secondary, technical). Women with primary education were more likely than women with other education levels to have experienced economic violence (35.8 per cent) (Table 22). In contrast, the incidence of economic violence was lowest among men without formal education (10.9 per cent). Wealthier women in the two highest asset quintiles were around 5–10 percentage points less likely to have experienced economic violence over the course of their lives. In contrast, men in the four highest asset quintiles were more likely to experience economic violence (26.7, 27.3, 27.1 and 26 per cent, respectively) than men in the lowest asset quintile (22 per cent) (Table 24). These education and wealth patterns were similar when considering the 12 months prior to the survey (Tables 23–25). 104 Table 22: Percentage of women who have experienced domestic or non-domestic economic violence over their lifetime Denied Cash Belongings Property Prohibited Forced Denied Any pocket taken controlled damaged from to work food economic money out working violence Age group (years) 15–19 19.4 1.4 3.4 1.6 2.5 6.8 1.1 30.4 20–24 21.3 2.5 1.9 3.6 2.4 6.2 0.1 31.7 25–29 25.9 3.6 2.6 5.2 3.9 4.7 0.0 33.9 30–39 20.4 2.6 2.1 2.6 3.0 3.9 0.2 27.4 40–49 15.8 2.2 1.8 3.2 2.7 2.6 0.1 24.2 50–60 16.8 2.3 1.8 3.6 1.9 3.6 0.1 23.6 Employment Self-employed 20.6 2.8 1.7 3.0 3.0 4.2 0.2 28.3 Employed 15.6 1.8 1.8 4.1 1.5 2.9 0.0 23.1 Not working 21.1 2.3 3.5 4.1 3.0 5.0 0.1 30.2 Marital status Never married 11.5 0.8 1.9 0.8 0.3 4.3 0.1 18.6 Married or living together 19.4 2.8 2.1 3.3 2.8 4.4 0.1 27.6 Divorced/separated/widowed 24.9 2.4 2.4 4.3 3.6 3.7 0.4 33.0 Residence Urban 16.9 2.0 2.6 4.1 2.7 4.3 0.2 25.5 Rural 23.8 3.2 1.7 2.7 3.1 4.3 0.2 31.3 105 Denied Cash Belongings Property Prohibited Forced Denied Any pocket taken controlled damaged from to work food economic money out working violence Region Western 31.8 2.3 2.1 4.4 2.8 1.5 0.7 37.4 Central 25.1 2.2 3.3 4.0 2.6 2.7 0.1 33.8 Greater Accra 11.4 0.4 2.1 4.2 2.1 4.6 0.0 19.1 Volta 13.0 1.4 2.1 0.4 0.2 2.0 0.1 17.4 Eastern 33.2 13.1 2.6 6.8 4.9 7.2 0.2 48.7 Ashanti 20.2 2.3 2.0 2.8 2.2 2.7 0.3 27.3 Brong Ahafo 24.3 1.6 1.9 2.3 5.4 8.7 0.1 34.2 Northern 12.4 0.3 1.1 2.8 4.0 7.2 0.0 20.8 Upper East 6.7 0.0 1.7 1.2 0.0 0.6 0.0 9.6 Upper West 16.6 0.0 2.4 1.2 4.8 2.6 0.0 20.4 Education level None 16.6 1.4 2.2 3.6 3.2 3.5 0.1 24.2 Primary 24.1 4.4 2.6 4.0 2.8 5.8 0.1 35.8 Middle/JSS/JHS 23.5 3.1 1.9 3.2 2.8 4.7 0.3 30.8 Secondary 15.7 1.2 2.0 2.4 2.8 4.0 0.0 22.0 Technical 16.9 2.9 1.1 1.2 5.4 1.4 0.0 25.6 Higher 11.8 1.7 2.4 5.2 0.4 1.9 0.0 19.4 Asset quintile Lowest 21.5 1.8 1.9 2.8 3.8 5.1 0.3 30.7 Second 23.1 3.1 2.6 4.1 2.5 3.3 0.1 30.9 Middle 23.8 1.8 2.7 3.2 2.6 4.5 0.5 31.7 Fourth 17.3 3.4 1.6 3.2 2.5 4.9 0.0 25.5 Highest 13.4 2.3 1.9 3.6 3.1 3.8 0.0 20.6 Total 20.2 2.6 2.2 3.4 2.8 4.3 0.2 28.2 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 106 Table 23: Percentage of women who experienced domestic or non-domestic economic violence in the last 12 months Denied Cash Belongings Property Prohibited Forced to Denied Any pocket taken controlled damaged from work food economic money out working violence Age group (years) 15–19 15.0 1.4 1.6 0.9 0.4 2.8 5.5 20.4 20–24 13.8 1.1 1.0 1.9 1.1 2.2 1.9 17.9 25–29 17.1 1.5 1.0 2.0 1.9 2.0 1.3 20.8 30–39 12.6 1.2 1.1 0.9 1.4 0.3 1.3 15.4 40–49 7.9 0.5 0.3 0.0 0.9 0.5 0.2 8.9 50–60 4.2 0.3 0.1 0.3 0.5 0.0 0.3 5.6 Employment Self-employed 11.4 1.1 0.5 0.6 1.4 0.8 1.1 13.7 Employed 7.6 1.3 1.0 1.2 1.0 0.7 0.3 11.0 Not working 14.2 0.7 1.5 1.7 0.7 1.7 2.4 17.8 Marital status Never married 6.9 0.8 0.9 0.3 0.0 3.3 3.7 12.4 Married or living together 12.4 1.1 0.8 1.0 1.3 0.9 0.9 15.0 Divorced/separated/widowed 10.2 0.6 0.7 1.1 1.1 0.9 1.9 13.0 Residence Urban 9.3 0.6 1.0 1.2 1.0 0.7 0.7 12.1 Rural 14.2 1.5 0.5 0.7 1.5 1.4 1.9 17.0 107 Denied Cash Belongings Property Prohibited Forced Denied Any pocket taken controlled damaged from to work food economic money out working violence Region Western 22.9 0.5 0.4 2.8 0.3 0.0 1.2 23.3 Central 13.6 0.9 1.5 1.1 0.7 1.1 2.6 15.6 Greater Accra 7.2 0.0 1.1 0.8 0.7 1.0 0.3 9.8 Volta 5.7 0.5 0.0 0.0 0.2 1.3 0.7 6.9 Eastern 16.1 5.4 1.1 0.8 1.2 1.1 1.3 23.4 Ashanti 12.4 1.3 0.4 1.0 1.3 0.5 1.4 15.3 Brong Ahafo 11.7 0.0 0.5 0.8 3.6 1.6 1.9 15.0 Northern 6.9 0.3 0.9 0.9 1.8 2.1 2.5 9.4 Upper East 6.1 0.0 1.1 0.6 0.0 0.6 0.6 7.9 Upper West 12.5 0.0 2.4 0.0 2.4 1.5 0.0 13.9 Education level None 8.6 0.3 0.7 1.3 1.1 0.5 1.0 10.2 Primary 15.0 2.2 0.9 0.1 0.9 1.0 0.8 18.6 Middle/JSS/JHS 13.1 1.2 0.6 1.2 1.3 1.3 2.0 16.2 Secondary 10.8 0.4 0.8 1.2 1.5 1.3 0.9 13.9 Technical 9.8 0.0 1.1 0.0 3.3 0.0 1.9 12.1 Higher 6.7 0.7 1.4 0.3 0.0 0.7 0.0 9.3 Asset quintile Lowest 11.2 1.0 0.5 0.8 1.5 1.6 2.2 14.6 Second 14.2 1.5 0.8 1.5 1.4 1.1 1.3 17.0 Middle 15.7 0.4 1.1 0.8 0.7 0.6 2.0 17.3 Fourth 9.7 1.0 1.0 0.8 1.1 1.3 0.5 13.0 Highest 6.2 0.8 0.7 0.6 1.1 0.1 0.5 8.6 Total 11.6 1.0 0.8 1.0 1.2 1.0 1.3 14.4 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 108 Table 24: Percentage of men who have experienced domestic or non-domestic economic violence over their lifetime Denied Cash Belongings Property Prohibited Forced to Denied Any pocket taken out controlled damaged from work food economic money working violence Age group (years) 15–19 21.4 4.7 4.1 4.0 1.3 6.6 0.1 32.3 20–24 17.2 4.2 0.6 6.4 4.5 7.4 0.1 29.2 25–29 13.4 4.5 0.4 5.8 3.1 9.6 0.1 30.0 30–39 8.7 3.8 1.4 9.1 3.8 6.2 0.3 23.9 40–49 9.2 3.7 2.2 7.2 3.7 2.9 0.7 21.6 50–60 8.2 4.2 1.8 6.2 0.8 4.2 0.4 20.1 Employment Self-employed 9.7 4.3 0.9 6.8 3.3 5.3 0.3 22.9 Employed 12.2 5.7 2.5 6.5 3.4 9.5 0.4 29.4 Not working 18.3 1.8 2.0 6.9 2.2 4.5 0.3 28.3 Marital status Never married 16.1 4.5 2.3 4.9 1.1 5.0 0.1 25.2 Married or living together 9.2 3.9 1.5 6.4 3.0 6.4 0.3 23.4 Divorced/separated/widowed 18.7 4.7 1.3 9.3 5.1 6.7 0.1 34.5 Residence Urban 12.4 3.8 2.1 7.8 2.9 6.9 0.3 25.9 Rural 12.0 4.4 1.1 5.8 3.2 5.6 0.3 25.7 109 Denied Cash Belongings Property Prohibited Forced Denied Any pocket taken controlled damaged from to work food economic money out working violence Region Western 18.9 4.0 1.4 6.8 2.5 6.5 0.0 32.8 Central 23.1 6.0 6.4 12.7 3.1 4.9 1.1 40.6 Greater Accra 12.0 0.3 2.1 9.2 2.4 8.9 0.0 23.0 Volta 4.5 2.4 1.6 1.0 0.9 4.9 0.1 13.4 Eastern 23.0 21.1 3.1 10.5 8.1 5.7 0.2 51.9 Ashanti 11.8 4.9 0.2 8.1 2.5 4.3 0.8 25.6 Brong Ahafo 7.4 0.5 0.9 5.5 5.3 7.9 0.0 22.5 Northern 3.9 0.0 0.0 3.2 3.0 8.2 0.4 15.1 Upper East 4.0 0.0 0.0 0.0 0.0 1.6 0.0 5.7 Upper West 6.6 0.0 0.0 0.0 2.2 6.5 0.1 11.0 Education level None 3.4 0.6 0.0 1.6 1.2 5.7 0.0 10.9 Primary 13.9 2.8 2.8 6.1 2.8 6.8 0.1 26.8 Middle/JSS/JHS 13.7 7.1 1.9 8.2 3.3 5.0 0.6 29.4 Secondary 15.6 3.2 1.0 7.1 4.6 8.5 0.1 28.6 Technical 11.0 3.4 1.5 9.3 1.5 8.2 0.1 28.7 Higher 11.1 2.8 2.0 7.2 2.7 5.9 0.3 25.3 Asset quintile Lowest 11.6 2.2 1.7 4.8 2.1 7.1 0.4 22.0 Second 12.5 5.2 1.3 5.0 4.3 6.3 0.2 26.7 Middle 12.7 3.8 1.8 8.3 2.5 6.6 0.1 27.3 Fourth 13.4 4.2 1.8 9.0 3.0 4.4 0.1 27.1 Highest 10.5 5.0 1.3 7.8 3.0 6.7 0.9 26.0 Total 12.2 4.1 1.6 6.8 3.1 6.2 0.3 25.8 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 110 Table 25: Percentage of men who experienced domestic or non-domestic economic violence in the last 12 months Denied Cash Belongings Property Prohibited Forced to Denied Any pocket taken controlled damaged from work food economic money out working violence Age group (years) 15–19 15.1 3.4 3.5 3.1 0.5 3.8 2.9 23.3 20–24 9.5 3.0 0.1 3.8 3.1 2.7 3.2 18.9 25–29 5.5 1.5 0.0 2.2 1.4 5.0 0.9 14.2 30–39 2.2 0.8 0.4 3.2 1.2 1.1 0.3 8.3 40–49 3.9 0.5 0.6 2.2 0.9 0.5 1.0 8.5 50–60 3.4 0.9 0.8 1.2 0.3 0.7 0.6 5.9 Employment Self-employed 3.5 1.4 0.1 1.8 1.1 1.0 0.5 8.1 Employed 4.5 1.6 0.9 2.7 1.8 4.8 2.2 14.2 Not working 12.8 1.6 2.0 4.5 1.1 2.1 2.2 20.0 Marital status Never married 11.0 2.7 1.9 3.1 0.4 2.9 2.9 18.3 Married or living together 3.8 1.0 0.5 1.8 1.0 1.9 0.7 9.4 Divorced/separated/widowed 7.1 2.1 0.3 4.4 3.1 2.6 2.0 16.4 Residence Urban 5.5 1.4 1.2 3.1 1.2 2.9 1.2 13.3 Rural 6.0 1.6 0.2 2.1 1.3 1.5 1.4 11.3 111 Denied Cash Belongings Property Prohibited Forced Denied Any pocket taken controlled damaged from to work food economic money out working violence Region Western 9.3 0.4 0.3 1.2 0.0 2.4 0.5 13.9 Central 12.4 3.8 4.5 4.3 1.3 1.9 9.1 25.7 Greater Accra 6.3 0.0 1.4 4.0 1.3 4.3 0.6 13.0 Volta 2.0 1.7 0.0 0.4 0.9 0.2 0.8 4.6 Eastern 3.1 9.0 0.8 2.6 4.8 2.3 2.1 19.2 Ashanti 7.2 0.6 0.0 4.8 0.9 1.6 0.7 13.4 Brong Ahafo 4.8 0.5 0.0 1.5 1.9 0.9 0.4 9.2 Northern 2.2 0.0 0.0 1.4 0.2 3.0 0.3 6.5 Upper East 2.4 0.0 0.0 0.0 0.0 1.0 0.0 3.4 Upper West 2.2 0.0 0.0 0.0 2.2 3.8 0.0 3.8 Education level None 0.6 0.0 0.0 1.1 0.0 0.6 0.3 2.3 Primary 7.3 0.9 1.0 0.3 2.1 2.4 3.5 12.5 Middle/JSS/JHS 6.2 2.4 1.1 3.1 1.3 1.6 1.2 13.3 Secondary 8.8 2.0 0.1 3.5 2.0 2.9 1.3 16.7 Technical 7.8 0.0 0.0 3.2 1.5 3.8 1.5 15.2 Higher 3.3 0.6 1.1 3.6 0.7 3.8 0.6 12.0 Asset quintile Lowest 7.1 0.5 0.8 1.7 0.8 2.4 1.1 9.4 Second 5.6 1.6 0.5 2.5 2.0 2.1 2.5 13.2 Middle 5.6 1.9 0.3 1.4 1.4 2.0 1.2 11.7 Fourth 6.5 2.1 1.0 3.4 1.0 1.4 0.6 13.7 Highest 3.5 1.3 0.9 4.3 1.0 3.3 0.6 13.3 Total 5.7 1.5 0.7 2.6 1.3 2.2 1.3 12.3 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 112 4.2. Incidence of domestic violence against women and men in Ghana This subsection and the remainder of the report will focus on violence experienced or perpetrated within domestic relations – domestic violence – in the 12 months prior to the survey.24 Domestic violence is defined, as per the Ghana Domestic Violence Act 732, as any form of violence against men or women perpetrated by a domestic relation. Domestic relations are core members of the family, as well as non-family members who live within the household. Table 26 summarises the overall incidence of domestic violence across the five types of violence in the 12 months prior to the survey. The most common form of domestic violence reported by women in the 12 months prior to the GFLHS 2015 was economic violence (12.8 per cent), followed by social violence (11.6 per cent) and psychological violence (9.3 per cent) (Table 26). This order is different from the previous section, where psychological violence had the highest incidence (Table 5). Domestic violence among men was mainly psychological violence (7.9 per cent), followed by social violence (7.7 per cent) and economic violence (7.3 per cent) (Table 26). This order is the same as for the overall levels of violence discussed in the previous section (Table 5). Table 26: Incidence of domestic violence in Ghana 2015 in the 12 months prior to the survey (in percentages) Social Physical Sexual Psychological Economic Total violence violence violence violence violence (at least one type) Women 11.6 6.0 2.5 9.3 12.8 27.7 Men 7.7 2.1 1.4 7.9 7.3 20.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. It is important to note that these types of domestic violence did not occur in isolation. According to the GFLHS 2015, many respondents who reported having experienced domestic violence experienced multiple forms and types of violence: 23.3 per cent of women and 18.9 per cent of men who experienced domestic violence reported having experienced two types of domestic violence, while 9.4 per cent of women and 5.7 per cent of men reported having experienced three different types of domestic violence. In what follows, the analysis focuses on each type of violence for ease of discussion. 24 Qualitative and quantitative data on domestic violence covered only the 12 months prior to the survey. No questions were asked about lifetime experiences due to budget and time constraints, as this would have increased substantially the length of the questionnaires and interviews. 113 4.2.1. Domestic social violence Table 27 compares the proportion of domestic and non-domestic social violence. The results show that almost all types of social violence against men and women were largely perpetrated within domestic relations. The only exception was the spread of rumours, photos and videos without permission: only 23.4 per cent and 13.2 per cent of this form of violence against women and men, respectively, constituted domestic violence. Nearly all (97.4 per cent) of the instances reported by women of being prevented from seeing friends or family were perpetrated by domestic relations. The following forms of violence against women were committed mostly by domestic relations: not being allowed to leave their home (98.5 per cent), required to report on their activities (88.4 per cent), being stalked (62.5 per cent) and being forced to have an abortion (women only) (84.6 per cent). The results for men were similar, with the exception of stalking, which was mostly (59 per cent) perpetrated by non-domestic relations. Table 27: Percentage of domestic social violence over the last 12 months Not Not Required Stalked Rumour, Forced to seeing allowed to photos, have an friends to leave report videos abortion or home activities spread family Women Perpetrated by family members 63.1 65.4 48.8 31.2 2.4 11.5 Perpetrated by non-family members but domestic 34.3 33.1 40.9 32.8 21.2 73.1 % Domestic violence 97.4 98.5 88.4 62.5 23.4 84.6 % Non-domestic violence 2.6 1.5 11.6 37.5 76.6 15.4 Total 100.0 100.0 100.0 100.0 100.0 100.0 Men Perpetrated by family members 65.3 80.0 29.8 17.9 3.3 N/A Perpetrated by non-family members but domestic 27.1 15.0 43.3 23.1 10.3 N/A % Domestic violence 92.4 95.0 73.1 41.0 13.2 N/A % Non-domestic violence 7.6 5.0 26.9 59.0 86.8 N/A Total 100.0 100.0 100.0 100.0 100.0 N/A Source: Ghana Family Life and Health Survey (GFLHS) 2015. Tables 28 and 29 illustrate how domestic social violence was distributed across the socio- economic characteristics and geographical location of women and men, respectively. The incidence of domestic social violence was higher among women (11.6 per cent) than among men (7.7 per cent), and all individual acts of domestic social violence were more prominent among women than among men. The relative importance of each type of domestic social violence was similar among women and men. The most common form of domestic social violence against women and men was being required to report their activities (5.1 per cent 114 for women and 3.9 per cent for men), followed by being stopped from leaving home (4.1 per cent for women and 1.6 per cent for men), being kept from seeing friends and family (4.0 per cent for women and 1.4 per cent for men), being the victim of rumour, photo or video spread (3.4 per cent for women and 1.7 per cent for men), and being stalked (1.2 per cent for women and 1.0 per cent for men). In addition, 0.7 per cent of women in the GFLHS 2015 were forced to having an abortion. Age patterns. As observed with overall levels of social violence, the incidence of domestic social violence was highest among women in the youngest age group (15–19 years): 35.7 per cent of women in this age group experienced at least one type of social violence in the 12 months prior to the survey. The incidence in other age groups was considerably lower: 19.3 per cent for women aged 20–24 years, 15.3 per cent for women aged 25–29 years, 8.6 per cent for women aged 30–39 years, 5.0 per cent for women aged 40–49 years and 3.9 per cent for women aged 50–60 years. The incidence of all acts of domestic social violence was highest among women aged 15–19 years, with the exception of being stalked, which was highest (2.4 per cent) for women aged 20–24 years. These age patterns were similar for men, but differences across the age groups were less pronounced: 13.5 per cent of all men aged 15–19 years experienced at least one form of domestic social violence in the 12 months prior to the survey, in comparison to 13 per cent of men aged 20–24 years, 8.7 per cent of men aged 25–29 years, 5.7 per cent of men aged 30–39 years, 3.8 per cent of men aged 40– 49 and 4.4 per cent of men aged 50–60 years. Employment status. The incidence of domestic social violence was highest for women and men who were not working at the time of the survey (19.8 per cent for women and 10.9 per cent for men). Both women and men who were self-employed were at lower risk of experiencing any act of domestic social violence (8.3 and 5.8 per cent, respectively). All types of domestic social violence were more commonly reported by women and men who were not working. The only exceptions were men who were required to report their activities and men who were victims of a rumour, photo or video being spread. These acts of domestic social violence were highest among employed men (5.6 per cent and 2 per cent, respectively). Marital status. Women who had never been married were at the highest risk of experiencing any form of domestic social violence: 27.1 per cent of these women experienced at least one act of domestic social violence in the 12 months prior to the survey, in comparison with 11 per cent of married women and 9.9 per cent of women who were divorced, separated or widowed. However, being a victim of the spreading of a rumour, photo or video or being forced to having an abortion were more common among women who were divorced, separated or widowed (4.6 per cent and 0.9 per cent, respectively). The patterns for men were different. Men who were divorced, separated or widowed were more at risk of experiencing any form of domestic social violence (8.3 per cent, in comparison to 8.1 per cent for never-married men and 7.4 per cent for married men). However, never-married men were more likely to report being stopped from leaving home (3.8 per cent) and being kept from seeing friends and family (2.4 per cent). These latter 115 results are likely to be explained by the fact that these men lived at home with their parents, an issue that will be explored in the next section, which focuses on the perpetration of violence. Geographical patterns. There was little difference in the levels of domestic social violence experienced by women living in urban and rural areas (11.9 and 11.3 per cent, respectively). The differences were more pronounced for men: 8.5 per cent of men living in urban areas experienced at least one act of domestic social violence, in comparison to 6.9 per cent of men living in rural areas. Women and men living in the Western region were more likely to report having experienced at least one act of domestic social violence (14.8 and 12.2 per cent, respectively). The incidence of domestic social violence was lowest among women living in the Brong Ahafo region (7.4 per cent) and men living in the Upper East region (1.7 per cent). Education levels and asset ownership. Table 29 shows a linear association between domestic social violence and the education and wealth status of men: the higher their level of education and asset quintile, the higher the likelihood of men reporting having experienced at least one act of domestic social violence in the 12 months prior to the survey. Thus 10.4 per cent of men with higher education reported having experienced domestic social violence, in comparison to 2.6 per cent of men with no education. Similarly, 11.9 per cent of men in the highest asset quintile experienced domestic social violence, in comparison to 4.0 per cent of men in the lowest asset quintile. As discussed before, this result may be explained by the fact that better educated and wealthier men may also be more likely to report experiences of domestic violence. The patterns for women were less pronounced. Women with secondary levels of education (15.8 per cent) and in the middle asset quintile (12.9 per cent) were the most likely to report having experienced domestic social violence. The incidence of domestic social violence was lowest among women with no education (5.4 per cent) and women in the highest asset quintile (10.3 per cent). 116 Table 28: Percentage of women who experienced domestic social violence over the last 12 months Not Not Required Stalked Rumour, Forced to Any seeing allowed to to photos, have an domestic friends or leave report videos abortion social family home activities spread violence Age group (years) 15–19 16.5 21.8 12.2 2.1 5.8 1.9 35.7 20–24 7.6 7.1 9.4 2.4 3.6 1.9 19.3 25–29 4.8 4.1 7.8 1.5 4.7 0.9 15.3 30–39 2.4 2.3 4.0 1.1 3.5 0.4 8.6 40–49 1.3 1.5 2.0 0.6 2.0 0.0 5.0 50–60 0.9 0.7 0.7 0.4 2.1 0.0 3.9 Employment Self-employed 2.4 2.4 3.4 1.0 3.3 0.5 8.3 Employed 1.9 2.4 6.1 0.5 3.0 1.5 13.0 Not working 9.2 9.3 9.2 1.9 4.0 0.8 19.8 Marital status Never married 13.1 14.2 8.9 1.6 3.5 0.5 27.1 Married or living together 3.6 3.8 5.5 1.2 3.1 0.6 11.0 Divorced/separated/widowed 3.1 2.7 2.7 1.3 4.6 0.9 9.9 Residence Urban 4.0 3.5 5.7 1.1 3.4 0.7 11.9 Rural 4.0 4.7 4.3 1.3 3.4 0.7 11.3 117 No t No t Requ ired Stal ked Rum our, Force d to An y seeing allowed to to photos, have an domestic friends or leave report videos abortion social family home activities spread violence Region Western 4.7 6.4 3.9 0.6 4.6 2.0 14.8 Central 4.5 4.6 3.9 0.0 1.4 1.1 10.2 Greater Accra 3.0 3.1 5.5 1.9 4.0 0.1 11.1 Volta 5.4 6.2 6.1 1.9 4.6 0.4 12.7 Eastern 4.8 5.5 6.3 1.7 4.5 1.1 14.5 Ashanti 3.2 2.8 5.9 0.6 4.0 0.4 11.9 Brong Ahafo 1.8 1.4 5.0 1.9 1.8 1.0 7.4 Northern 7.2 4.9 2.9 1.4 0.9 0.2 10.9 Upper East 1.8 6.7 0.6 0.0 4.7 0.0 10.6 Upper West 8.0 5.6 9.0 0.0 2.2 0.0 12.5 Education level None 2.3 2.2 1.6 0.6 1.6 0.2 5.4 Primary 4.0 4.9 4.9 1.5 6.0 0.4 13.4 Middle/JSS/JHS 5.0 5.3 5.1 1.5 4.3 1.1 13.7 Secondary 4.2 4.2 10.9 1.8 2.0 0.3 15.8 Technical 0.0 1.3 4.6 0.0 0.9 0.6 6.8 Higher 4.8 2.4 6.8 0.6 1.5 0.8 10.2 Asset quintile Lowest 4.4 4.5 3.9 1.3 4.6 0.3 11.2 Second 4.2 4.5 4.3 1.6 2.8 0.9 11.1 Middle 3.5 4.6 4.9 1.6 4.2 1.6 12.9 Fourth 3.9 4.1 6.8 0.6 4.0 0.4 12.4 Highest 3.5 2.2 5.6 1.0 1.3 0.2 10.3 Total 4.0 4.1 5.1 1.2 3.4 0.7 11.6 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 118 Table 29: Percentage of men who experienced domestic social violence over the last 12 months Not Not Required Stalked Rumour, Forced to Any seeing allowed to to photos, have an domestic friends or leave report videos abortion social family home activities spread violence Age group (years) 15–19 4.6 6.8 3.3 0.0 3.4 N/A 13.5 20–24 2.2 1.0 6.3 1.7 2.9 N/A 13.0 25–29 1.0 1.4 6.0 2.2 1.3 N/A 8.7 30–39 0.5 0.5 4.0 0.7 1.4 N/A 5.7 40–49 0.8 0.7 1.5 0.7 1.1 N/A 3.8 50–60 0.8 1.8 1.7 0.0 1.1 N/A 4.4 Employment Self-employed 1.2 1.4 2.9 1.0 1.5 N/A 5.8 Employed 0.8 0.7 5.6 0.8 2.0 N/A 8.6 Not working 2.5 3.3 4.2 1.1 1.9 N/A 10.9 Marital status Never married 2.4 3.8 2.0 0.0 2.0 N/A 8.1 Married or living together 1.3 1.4 4.4 0.8 1.5 N/A 7.4 Divorced/separated/widowed 0.8 0.5 4.2 2.3 2.4 N/A 8.3 Residence Urban 1.4 1.8 4.9 1.0 1.3 N/A 8.5 Rural 1.3 1.5 3.0 0.9 2.2 N/A 6.9 119 Not No t Requ ired Stal ked Rum our, Force d to An y seeing allowed to to photos, have an domestic friends or leave report videos abortion social family home activities spread violence Region Western 1.8 2.3 5.7 2.6 1.8 N/A 12.2 Central 2.3 1.6 4.7 1.8 5.9 N/A 10.0 Greater Accra 1.3 1.8 6.0 1.9 0.6 N/A 9.1 Volta 0.5 1.3 3.5 0.0 1.9 N/A 6.3 Eastern 1.8 2.1 4.5 0.6 2.7 N/A 9.6 Ashanti 1.1 1.0 4.7 0.6 1.8 N/A 8.5 Brong Ahafo 1.4 1.8 0.0 0.0 0.9 N/A 3.1 Northern 0.8 1.1 1.4 0.3 0.9 N/A 3.1 Upper East 0.7 0.7 1.0 0.0 0.0 N/A 1.7 Upper West 4.4 4.4 0.0 0.0 1.8 N/A 6.1 Education level None 1.0 1.7 0.4 0.0 0.4 N/A 2.6 Primary 2.8 1.9 1.8 1.3 1.8 N/A 7.5 Middle/JSS/JHS 1.9 2.5 3.0 0.3 2.2 N/A 7.5 Secondary 0.5 0.9 4.9 2.2 2.5 N/A 9.3 Technical 2.6 0.0 6.8 1.8 0.6 N/A 9.4 Higher 0.3 0.5 9.2 1.1 0.9 N/A 10.4 Asset quintile Lowest 1.1 1.5 1.2 0.4 1.4 N/A 4.0 Second 1.3 1.2 2.3 0.8 2.4 N/A 6.4 Middle 1.7 1.7 4.0 0.3 2.7 N/A 8.2 Fourth 1.9 2.1 4.5 1.0 1.3 N/A 9.2 Highest 0.8 1.8 9.0 2.5 0.7 N/A 11.9 Total 1.4 1.6 3.9 1.0 1.7 N/A 7.7 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 120 4.2.2. Domestic physical violence All types of physical violence against women were mainly perpetrated by domestic relations (within and outside the immediate family), whereas all types of physical violence against men were mainly perpetrated by non-domestic relations (Table 30). Forms of physical violence against women and men were more likely to be perpetrated by domestic relations outside the immediate family. The exceptions were women being slapped or being threatened with a weapon, which were equally likely to be perpetrated by family or non-family members (36.8 and 38.9 per cent, respectively). It was also more likely that men were burnt by family members (33.3 per cent) than by non-family members (none). Table 30: Percentage of domestic physical violence over the last 12 months Slapped or Pushed Hit Kicked, Choked Burnt Weapon thrown or dragged or used things at shoved or beaten strangled against up Women Perpetrated by family 36.8 24.2 26.9 35.9 25.0 50.0 38.9 members Perpetrated by non- family members but 36.8 38.7 31.3 48.4 55.0 50.0 38.9 domestic % Domestic violence 73.0 62.9 58.2 84.4 80.0 100.0 77.8 % Non-domestic 27.0 37.1 41.8 15.6 20.0 0.0 22.2 vTiootlaeln ce 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Men Perpetrated by family 5.1 0.0 2.0 5.9 0.0 33.3 4.5 members Perpetrated by non- family members but 22.2 6.4 12.0 23.5 10.5 0.0 18.2 domestic % Domestic violence 27.3 6.4 14.0 29.4 10.5 33.3 22.7 % Non-domestic 72.7 93.6 86.0 70.6 89.5 66.7 77.3 Tviootlaeln ce 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Tables 31 and 32 illustrate the variation of incidence of domestic physical violence across the socio-economic and geographical characteristics of women and men, respectively. Women were almost three times more likely to experience physical violence (6 per cent) than men (2.1 per cent). The most common act of domestic physical violence experienced by women and men was being slapped or having objects thrown at them (4.4 and 1.4 per cent, respectively). The second most common form of domestic physical violence for women was being kicked, dragged or beaten up (1.8 per cent), while for men it was being hit (0.3 per cent). These results suggest that women and men experienced domestic 121 physical violence in different ways, an issue that will be further analysed in subsequent sections. Age patterns. As with other forms of domestic and non-domestic violence, the incidence of domestic physical violence was highest among younger women (15–19 years old): 15.2 per cent of these women reported having experienced at least one act of domestic physical violence, in comparison to 9.4 per cent of women aged 20–24 years, 6.9 per cent of women aged 25–29 years, 7.1 per cent of women aged 30–39 years, 2.4 per cent of women aged 40– 49 years and 0.5 per cent of women aged 50–60 years. The incidence of all acts of domestic physical violence was also highest among younger women, with the exception of being pushed or shoved (highest among women aged 25–29 years: 2.4 per cent) and being hit (highest among women aged 20–24 years: 2.2 per cent). The association between age and domestic physical violence was slightly different for men. The incidence of this form of domestic violence was also highest (8.5 per cent) among younger men (15–19 years). However, the lowest incidence of domestic physical violence was observed among men aged 30–39 years (0.5 per cent). In contrast to the data discussed above for women, older men aged 50–60 years were the most likely to report being hit (0.6 per cent) and being choked or strangled (0.3 per cent). Employment status. The incidence of domestic physical violence was highest among women and men who were not working (9.1 and 3.8 per cent, respectively). The lowest incidence was observed among women who were employed (4.7 per cent) but among men who were self-employed (1.5 per cent). Marital status. Women and men who had never been married at the time of the survey were the most likely group to report having experienced at least one act of domestic physical violence in the 12 months prior to the survey (9.0 and 5.6 per cent, respectively). There were not substantial differences in the number of women and men who were married or were divorced, separated or widowed and reported having experienced domestic physical violence. However, women who were divorced, separated or widowed were more likely than married or never-married women to experience acts of violence: being pushed or shoved (2.2 per cent), being hit (1.5 per cent), being kicked, dragged or beaten up (2.9 per cent) or being chocked or strangled (0.9 per cent). These results suggest that these forms of violence were more likely to be perpetrated within intimate partnerships. These patterns were observed for women only. Geographical patterns. Both women and men living in rural areas were more likely to have experienced domestic physical violence (6.4 and 2.6 per cent, respectively), than those in urban areas (5.7 and 1.6 per cent, respectively). The highest incidence of domestic physical violence among women was in the Eastern region (8.7 per cent, in comparison with the lowest level of 0.8 per cent in the Upper East region). The highest incidence levels among men were observed in the Central region (4.4 per cent), and the lowest in the Northern region (0.7 per cent). 122 Education levels and asset ownership. The incidence of domestic physical violence was highest among women with primary levels of education: 9.2 per cent of these women experienced domestic physical violence, in comparison to, for instance, 2.9 per cent of women with no education (who reported the lowest incidence of domestic physical violence). Women in the middle asset quintile (8.2 per cent) reported higher levels of domestic physical violence than women in the highest quintile (5.8 per cent) and women in the lowest quintile (6.3 per cent). These results seem to suggest that women with intermediate levels of education and wealth may be at a higher risk of experiencing domestic physical violence. Men with secondary education were more likely to experience domestic physical violence (3.6 per cent). The lowest levels of incidence were observed among men with technical education (no men reported domestic physical violence), men with no education (0.6 per cent) and men with higher education (0.8 per cent). The pattern for wealth was clearer: men in the two lowest asset quintiles (3.2 and 3.5 per cent, respectively) were more than three times more likely to experience domestic physical violence than men in the fourth quintile (0.8 per cent) or in the top asset quintile (1.0 per cent). 123 Table 31: Percentage of women who experienced domestic physical violence over the last 12 months Slapped or Pushed Hit Kicked, Choked or Burnt Attacked Any thrown or dragged or strangled with physical things at shoved beaten up weapon violence Age group (years) 15–19 10.4 1.5 1.2 4.9 1.3 1.1 1.1 15.2 20–24 7.2 1.5 2.2 2.1 0.0 0.3 0.0 9.4 25–29 5.5 2.4 1.5 2.6 1.1 0.3 1.0 6.9 30–39 4.7 1.2 1.4 2.4 0.6 0.0 1.0 7.1 40–49 2.0 0.8 0.8 0.3 0.2 0.0 0.2 2.4 50–60 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.5 Employment Self-employed 3.5 1.2 1.1 1.8 0.5 0.0 0.4 5.1 Employed 3.2 1.1 1.4 1.4 0.4 0.0 0.7 4.7 Not working 7.3 1.5 1.3 2.0 0.4 0.7 1.0 9.1 Marital status Never married 5.9 0.8 0.0 2.3 0.0 1.3 1.3 9.0 Married or living together 4.5 1.0 1.2 1.4 0.4 0.1 0.5 5.9 Divorced/separated/widowed 3.7 2.2 1.5 2.9 0.9 0.0 0.7 5.7 Residence Urban 4.4 1.0 1.3 1.7 0.4 0.1 0.6 5.7 Rural 4.4 1.5 1.0 1.9 0.6 0.2 0.5 6.4 124 Slapped or Pushed Hit Kicked, Choked or Burnt Attacked Any thrown or dragged or strangled with physical things at shoved beaten up weapon violence Region Western 4.5 0.7 0.5 0.9 0.0 0.0 0.4 5.1 Central 5.0 1.4 1.0 2.2 0.5 0.5 0.8 6.9 Greater Accra 4.2 1.3 1.0 1.3 0.2 0.0 0.8 4.8 Volta 5.4 2.1 1.0 1.9 0.3 1.0 0.0 6.3 Eastern 6.3 1.7 2.5 2.8 1.1 0.0 1.2 8.7 Ashanti 4.4 0.8 1.2 2.1 0.2 0.0 0.4 6.7 Brong Ahafo 4.2 2.0 2.0 1.8 1.2 0.2 0.7 6.8 Northern 3.0 0.5 0.3 1.5 0.5 0.0 0.0 4.2 Upper East 0.0 0.8 0.8 0.8 0.8 0.0 0.0 0.8 Upper West 2.4 0.5 0.5 2.3 1.6 0.0 1.2 6.4 Education level None 2.3 0.7 0.8 0.4 0.6 0.0 0.3 2.9 Primary 6.1 2.3 1.7 4.0 1.0 0.2 0.8 9.2 Middle/JSS/JHS 4.8 1.3 1.3 1.8 0.3 0.3 0.7 6.7 Secondary 5.1 0.9 1.5 1.7 0.5 0.0 0.0 6.6 Technical 4.5 3.5 0.6 4.0 0.0 0.0 3.5 5.1 Higher 3.4 0.2 0.2 0.3 0.0 0.0 0.2 3.8 Asset quintile Lowest 4.0 2.1 1.3 2.0 1.3 0.4 0.3 6.3 Second 4.6 1.3 1.3 2.0 0.5 0.1 0.7 6.6 Middle 5.5 0.6 1.8 2.3 0.1 0.0 0.7 8.2 Fourth 3.5 0.6 0.8 0.9 0.4 0.2 0.2 3.7 Highest 4.8 1.7 0.7 2.0 0.0 0.0 1.1 5.8 Total 4.4 1.2 1.2 1.8 0.5 0.2 0.6 6.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 125 Table 32: Percentage of men who experienced domestic physical violence over the last 12 months Slapped or Pushed Hit Kicked, Choked or Burnt Attacked Any thrown or dragged or strangled with physical things at shoved beaten up weapon violence Age group (years) 15–19 6.4 0.7 0.5 0.7 0.0 0.8 0.0 8.5 20–24 1.9 0.0 1.0 0.3 0.0 0.0 0.3 3.4 25–29 1.2 0.0 0.2 0.4 0.2 0.0 0.4 2.0 30–39 0.3 0.0 0.0 0.0 0.0 0.0 0.3 0.5 40–49 0.4 0.3 0.0 0.0 0.0 0.0 0.0 0.6 50–60 0.7 0.6 0.6 0.3 0.3 0.0 0.3 0.9 Employment Self-employed 0.9 0.3 0.2 0.2 0.2 0.0 0.5 1.5 Employed 1.6 0.0 0.6 0.0 0.0 0.0 0.0 2.0 Not working 2.4 0.3 0.2 0.6 0.0 0.4 0.0 3.8 Marital status Never married 3.8 0.4 0.9 0.4 0.0 0.5 0.0 5.6 Married or living together 0.8 0.2 0.3 0.3 0.1 0.0 0.4 1.4 Divorced/separated/widowed 1.5 0.0 0.0 0.0 0.0 0.0 0.0 1.5 Residence Urban 1.0 0.1 0.3 0.0 0.0 0.2 0.2 1.6 Rural 1.8 0.3 0.3 0.4 0.2 0.0 0.3 2.6 126 Slapped or Pushed Hit Kicked, Choked or Burnt Attacked Any thrown or dragged or strangled with physical things at shoved beaten up weapon violence Region Western 1.5 0.0 0.0 0.0 0.0 0.0 0.0 1.5 Central 1.1 0.7 0.6 0.0 0.6 1.2 1.4 4.4 Greater Accra 0.7 0.0 0.6 0.0 0.0 0.0 0.4 1.6 Volta 1.2 0.0 0.5 1.0 0.0 0.0 0.0 2.2 Eastern 3.6 0.0 0.2 0.0 0.0 0.0 0.0 3.8 Ashanti 1.4 0.0 0.2 0.5 0.0 0.0 0.0 2.1 Brong Ahafo 1.5 0.0 0.0 0.0 0.0 0.0 0.3 1.8 Northern 0.7 0.7 0.0 0.0 0.0 0.0 0.0 0.7 Upper East 1.4 1.1 1.1 1.1 1.1 0.0 0.0 1.8 Upper West 1.5 1.7 1.7 0.0 0.0 0.0 1.7 3.2 Education level None 0.6 0.3 0.6 0.3 0.6 0.0 0.0 0.6 Primary 1.5 0.7 0.3 0.0 0.0 0.7 0.3 3.1 Middle/JSS/JHS 1.6 0.0 0.1 0.2 0.0 0.0 0.3 2.3 Secondary 2.5 0.3 0.8 0.6 0.0 0.0 0.4 3.6 Technical 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Higher 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.8 Asset quintile Lowest 2.7 0.4 0.7 0.5 0.2 0.0 0.2 3.2 Second 1.5 0.0 0.6 0.5 0.2 0.3 0.5 3.5 Middle 1.3 0.4 0.0 0.1 0.0 0.0 0.0 1.5 Fourth 0.5 0.3 0.0 0.0 0.0 0.0 0.0 0.8 Highest 0.8 0.0 0.2 0.0 0.0 0.0 0.5 1.0 Total 1.4 0.2 0.3 0.2 0.1 0.1 0.2 2.1 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 127 4.2.3. Domestic sexual violence Most types of sexual violence against women were predominantly perpetrated by domestic relations, with the exception of sexual comments and sexual touches (3.1 and 5.7 per cent of these two acts, respectively, were committed by non-domestic relations). In contrast, most acts of sexual violence against men – with the exception of sex without consent and sex without protection – were mainly perpetrated by non-domestic relations. This finding suggests that sexual violence against women was more prevalent within domestic relations, whereas for men it was more widespread outside domestic relations (Table 33). Most domestic sexual violence against women was perpetrated by non-family members. This includes sexual comments (2.8 per cent perpetrated by non-family members, and 0.3 per cent by family members), sexual touch (5.7 per cent perpetrated by non-family members, and none by family members), physically forced to have sex (28.6 per cent perpetrated by non-family members, and 22.9 per cent by family members), and having sex without protection (69.6 per cent perpetrated by non-family members, and 21.7 per cent perpetrated by family members). All acts of domestic sexual violence against men were more likely to be committed by non-family members than family members (Table 33). Table 33: Percentage of domestic sexual violence over the last 12 months Sexual Sexual Physically Otherwise Sex Sex Sex comment touch forced to forced to without without because have sex have sex consent protection afraid Women Perpetrated by family 0.3 0.0 22.9 28.6 37.5 21.7 53.6 members Perpetrated by non- family members but 2.8 5.7 28.6 14.3 37.5 69.6 25.0 domestic % Domestic violence 3.1 5.7 51.4 42.9 75.0 91.3 78.6 % Non-domestic 96.9 94.3 48.6 57.1 25.0 8.7 21.4 vTiootlaeln ce 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Men Perpetrated by family 1.4 1.4 0.0 0.0 0.0 0.0 6.7 members Perpetrated by non- family members but 2.2 4.3 12.0 18.8 60.0 75.0 60.0 domestic % Domestic violence 3.6 5.7 12.0 18.8 60.0 75.0 66.7 % Non-domestic 96.4 94.3 88.0 81.2 40.0 25.0 33.3 vTiootlaeln ce 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Tables 34 and 35 illustrate, respectively, how domestic sexual violence was experienced across different socio-economic groups. As with other forms of domestic violence, women 128 were more likely than men to have experienced domestic sexual violence (2.5 and 1.4 per cent, respectively). The most common type of domestic sexual violence experienced by women and men was being forced to have sex because of being afraid of the consequences of refusing. This form of violence affected 0.8 per cent of all women in the GFLHS 2015 and 0.6 per cent of men. The second most common reason for both women and men was being forced to have sex without protection (0.7 per cent for women and 0.3 per cent for men). Age patterns. In contrast to other forms of domestic violence, the incidence of domestic sexual violence was highest among women aged 20–24 years: 4.9 per cent of women in this age group reported having experienced domestic sexual violence, in comparison to 4.6 per cent of women aged 25–29 years, 4.3 per cent of women aged 15–19 years, 2.2 per cent of women aged 30–39 years and 0.5 per cent of women aged 40–49 years. However, the incidence of particular types of domestic sexual violence was highest among young women aged 15–19 years: unwanted sexual touch (0.7 per cent), being physically forced to have sex (1.5 per cent), and being forced to have sex without protection (1.6 per cent). As with other forms of domestic violence, the incidence of domestic sexual violence was lowest among the older women (aged 50–60 years). Men aged 30–39 were at the highest risk of experiencing domestic sexual violence (2.7 per cent) across almost all acts of sexual violence reported. The main exception was being the target of an unwanted sexual comment, which was most prevalent (0.7 per cent) among the younger age group (15–19 years). No men aged 50–60 years reported having experienced an act of sexual violence in the 12 months prior to the survey. Employment status. Women who were either employed or not working at the time of the survey were most likely to report domestic sexual violence (3.0 per cent in both cases). Employed men were most at risk of domestic sexual violence (also 3.0 per cent). The results in Tables 34 and 35 show, however, great variation in the association between employment status and the incidence of domestic violence, with no clear pattern emerging from the data. Marital status. The incidence of domestic sexual violence was highest among married women (2.9 per cent) and divorced, separated or widowed men (2.1 per cent). No never- married women reported having experienced domestic sexual violence. The incidence of domestic sexual violence was lowest among never-married men (0.4 per cent). The results show, however, a variation in how women and men experienced different acts of domestic sexual violence. Notably, women who were divorced, separated or widowed experienced the highest incidence of unwanted sexual comments (0.6 per cent) and acts of forced sex (0.2 per cent). Married men experienced the highest incidence of unwanted sexual touches (0.2 per cent) and acts of physically forced sex (0.3 per cent), while never-married men reported the highest incidence of unwanted sexual comments (0.4 per cent). Geographical patterns. Similar to domestic physical violence, women and men living in urban areas reported a higher incidence of domestic sexual violence (2.8 and 1.6 per cent, respectively) than those in rural areas. The highest level of incidence of domestic sexual 129 violence among women and men was observed in the Central region (5.8 and 2.5 per cent, respectively). The lowest incidence of domestic sexual violence among women was in the Northern and Upper East regions (both 0.6 per cent), and no men in the Brong Ahafo, Upper East or Upper West regions reported domestic sexual violence. Education levels and asset ownership. The incidence of domestic sexual violence was highest among women with secondary education: 4.9 per cent of these women experienced at least one act of sexual violence in the 12 months prior to the survey, in comparison to, for instance, 0.9 per cent of women with no education. Men with technical or higher education were more likely to have experienced domestic sexual violence (2.2 and 2.1 per cent, respectively, in comparison with men with no education). Both men and women in the two highest asset quintiles were more at risk of experiencing domestic sexual violence (3.9 and 2.5 per cent for women, and 1.7 and 2.7 per cent for men) than men and women in the other asset quintiles. 130 Table 34: Percentage of women who experienced domestic sexual violence over the last 12 months Sexual Sexual Physically Otherwise Sex Sex Had sex Any comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Age group (years) 15–19 0.0 0.7 1.5 0.0 0.4 1.6 0.5 4.3 20–24 0.3 0.3 1.1 0.0 1.2 1.1 1.7 4.9 25–29 0.4 0.6 0.9 0.5 0.5 1.0 1.5 4.6 30–39 0.0 0.0 0.6 0.2 0.7 0.8 0.7 2.2 40–49 0.1 0.0 0.1 0.0 0.1 0.0 0.4 0.5 50–60 0.3 0.3 0.0 0.0 0.0 0.0 0.0 0.3 Employment Self-employed 0.2 0.2 0.7 0.1 0.6 0.6 0.6 2.3 Employed 0.6 0.3 0.5 0.0 0.0 0.9 0.8 3.0 Not working 0.0 0.3 0.4 0.2 0.5 0.8 1.4 3.0 Marital status Never married 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Married or living together 0.1 0.3 0.8 0.1 0.7 0.7 1.0 2.9 Divorced/separated/widowed 0.6 0.3 0.1 0.2 0.0 0.7 0.3 1.9 Residence Urban 0.2 0.2 0.5 0.0 0.3 0.9 1.2 2.8 Rural 0.2 0.3 0.7 0.3 0.7 0.4 0.4 2.2 131 Sex ual Sex ual Physi cally Other wise Se x Se x Had sex An y comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Region Western 0.1 0.4 0.2 0.0 0.3 0.3 0.3 1.1 Central 0.0 0.5 1.4 0.5 0.9 1.4 2.0 5.8 Greater Accra 0.2 0.0 0.2 0.0 0.8 1.2 1.6 3.2 Volta 0.0 0.6 1.3 0.6 0.0 0.3 1.0 3.4 Eastern 0.8 0.8 0.6 0.0 0.6 0.6 0.6 2.2 Ashanti 0.3 0.3 0.8 0.3 0.7 0.1 0.6 2.8 Brong Ahafo 0.0 0.0 0.3 0.0 0.0 1.0 0.0 1.0 Northern 0.0 0.0 0.6 0.0 0.3 0.0 0.0 0.6 Upper East 0.0 0.0 0.0 0.0 0.0 0.6 0.0 0.6 Upper West 0.0 0.0 0.0 0.0 0.0 1.1 0.0 1.1 Education level None 0.1 0.1 0.5 0.2 0.1 0.1 0.0 0.9 Primary 0.1 0.0 0.5 0.0 0.7 0.1 1.0 2.2 Middle/JSS/JHS 0.3 0.4 0.8 0.2 0.8 0.9 0.8 2.9 Secondary 0.5 0.3 0.9 0.0 0.3 1.5 1.8 4.9 Technical 0.0 0.0 0.0 0.0 0.0 0.0 1.9 1.9 Higher 0.0 0.8 0.0 0.0 0.0 1.2 0.5 2.5 Asset quintile Lowest 0.0 0.2 0.5 0.2 0.0 0.3 0.3 1.6 Second 0.3 0.5 0.7 0.0 0.6 0.6 0.6 2.3 Middle 0.1 0.0 0.9 0.3 0.4 0.7 0.4 2.2 Fourth 0.2 0.0 0.7 0.2 1.0 0.9 1.7 3.9 Highest 0.3 0.5 0.2 0.0 0.3 0.8 1.0 2.5 Total 0.2 0.3 0.6 0.1 0.5 0.7 0.8 2.5 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 132 Table 35: Percentage of men who experienced domestic sexual violence over the last 12 months Sexual Sexual Physically Otherwise Sex Sex Had sex Any comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Age group (years) 15–19 0.7 0.0 0.0 0.0 0.0 0.0 0.0 0.7 20–24 0.0 0.0 1.0 0.0 0.3 0.5 1.2 2.3 25–29 0.3 0.2 0.0 0.4 0.4 0.0 0.5 1.4 30–39 0.3 0.1 0.2 0.5 0.3 0.7 1.1 2.7 40–49 0.0 0.2 0.0 0.0 0.0 0.2 0.0 0.3 50–60 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Employment Self-employed 0.0 0.1 0.3 0.1 0.1 0.0 0.2 0.8 Employed 0.2 0.1 0.2 0.5 0.4 0.5 1.7 3.0 Not working 0.6 0.0 0.0 0.0 0.0 0.6 0.0 1.0 Marital status Never married 0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.4 Married or living together 0.1 0.2 0.3 0.2 0.1 0.2 0.6 1.5 Divorced/separated/widowed 0.3 0.0 0.0 0.4 0.7 0.7 0.9 2.1 Residence Urban 0.1 0.2 0.1 0.4 0.2 0.2 0.6 1.6 Rural 0.2 0.0 0.3 0.0 0.2 0.3 0.5 1.2 133 Sexual Sexual Physically Otherwise Sex Sex Had sex Any comment touch forced to forced to without without because sexual have sex have sex consent protection afraid violence Region Western 0.0 0.2 0.0 0.5 0.0 0.7 0.7 2.1 Central 0.8 0.4 0.0 0.0 0.0 0.7 1.2 2.5 Greater Accra 0.0 0.1 0.3 0.8 0.4 0.2 1.0 2.3 Volta 0.1 0.0 1.5 0.0 0.5 0.0 0.5 1.6 Eastern 0.0 0.0 0.0 0.0 0.7 0.0 1.8 1.8 Ashanti 0.6 0.0 0.0 0.0 0.0 0.2 0.0 0.8 Brong Ahafo 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Northern 0.0 0.3 0.0 0.0 0.0 0.8 0.0 1.1 Upper East 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Upper West 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Education level None 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Primary 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.2 Middle/JSS/JHS 0.2 0.1 0.0 0.3 0.0 0.3 0.9 1.8 Secondary 0.0 0.2 0.7 0.0 0.2 0.3 0.8 1.7 Technical 0.0 0.0 0.0 2.2 2.2 0.0 0.0 2.2 Higher 0.8 0.2 0.3 0.0 0.4 0.6 0.4 2.1 Asset quintile Lowest 0.3 0.0 0.2 0.0 0.2 0.6 0.2 1.2 Second 0.0 0.1 0.0 0.0 0.0 0.0 0.8 0.9 Middle 0.0 0.0 0.0 0.0 0.0 0.2 0.7 0.9 Fourth 0.5 0.0 0.0 0.7 0.4 0.3 0.4 1.7 Highest 0.1 0.4 0.9 0.3 0.4 0.4 0.7 2.7 Total 0.2 0.1 0.2 0.2 0.2 0.3 0.6 1.4 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 134 4.2.4. Domestic psychological violence Table 36 shows that most psychological violence against women was perpetrated within domestic relations, with the exception of insults (37.1 per cent of this violence was perpetrated by domestic relations, and 62.9 per cent by non-domestic relations). Most psychological violence against men was perpetrated outside domestic relations, with the exception of acts of being ignored or threatened with abandonment (59.6 per cent of this violence was perpetrated by domestic relations) (Table 36). All acts of psychological violence against women and men were more likely to be perpetrated by non-family members than by family members. Table 36: Percentage of domestic psychological violence over the last 12 months Insulted, Ignored, Scared, Attacked Threaten to humiliated, threatened intimidated with be hurt belittled to be weapon abandoned Women Perpetrated by family members 13.4 33.7 12.5 26.1 14.3 Perpetrated by non-family members but domestic 24.2 46.9 37.5 52.2 50.0 % Domestic violence 37.1 80.6 50.0 78.3 64.3 % Non-domestic violence 62.9 19.4 50.0 21.7 35.7 Total 100.0 100.0 100.0 100.0 100.0 Men Perpetrated by family members 7.6 17.3 2.8 0.0 5.6 Perpetrated by non-family members but domestic 20.7 42.3 8.3 15.0 16.7 % Domestic violence 27.6 59.6 11.1 15.0 22.2 % Non-domestic violence 72.4 40.4 88.9 85.0 77.8 Total 100.0 100.0 100.0 100.0 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Tables 37 and 38 disaggregate the information above further and show how the incidence of domestic psychological violence was distributed across different socio-economic characteristics and geographical locations. As with other types of domestic violence, women were more likely to experience domestic psychological violence (6 per cent) than men (2.1 per cent). The most common act of domestic psychological violence was being insulted, humiliated or belittled (7.5 per cent for women and 6.9 per cent for men). The second most common act was being ignored, or threatened with abandonment (2.4 per cent for women and 1.5 per cent for men). Age patterns. The incidence of domestic psychological violence was highest among women and men in the youngest age group (15–19 years old): 15.2 per cent of women and 8.5 per cent of men in this age group experienced at least one act of domestic psychological 135 violence in the 12 months prior to the survey. Levels of incidence of domestic psychological violence reduced as the age of women increased. Among the older age group (50–60 years), only 0.5 per cent of women reported having experienced domestic psychological violence. However, the incidence of certain forms of psychological violence, such as being threatened with a weapon, was highest among women aged 25–29 years (0.9 per cent), whereas being threatened to be hurt was highest among women aged 30–39 years. The lowest incidence of domestic psychological violence among men was among those aged 30–39 years (0.5 per cent). Employment status. Women and men who were not working at the time of the survey were more likely to experience at least one act of domestic psychological violence (9.1 per cent for women and 3.8 per cent for men). The lowest incidence of domestic psychological violence was observed among employed women (4.7 per cent) and self-employed men (1.5 per cent). Marital status. The incidence of domestic psychological violence was highest among women and men who were never married (9.0 and 5.6 per cent, respectively). However, divorced, separated or widowed women were the most likely group to report the incidence of particular acts of domestic psychological violence, such as being threatened with the use of a weapon (0.8 per cent) or being threatened to be hurt (0.7 per cent). Likewise, divorced, separated or widowed men were the most likely to report being insulted, humiliated or belittled (9.6 per cent) or being threatened to be hurt (0.7 per cent), while married men were the most likely to report being scared or intimidated on purpose (0.3 per cent) or being threatened with a weapon (0.3 per cent). Geographical patterns. Women and men living in rural areas were more likely to have reported being a victim of domestic psychological violence (6.4 and 2.6 per cent, respectively) than those in urban areas (5.7 and 1.6 per cent, respectively). The incidence of domestic psychological violence among women was highest in the Eastern region (8.7 per cent) and lowest in the Upper East region (0.8 per cent). Among men, it was highest among those living in the central region (4.4 per cent) and lowest among those living in the Northern region (0.7 per cent). Education levels and asset ownership. The incidence of domestic psychological violence was highest among women with primary levels of education (9.2 per cent), and lowest among women with no education (2.9 per cent) or women with the highest levels of education (3.8 per cent). Men with secondary education were those most likely to report experiences of domestic psychological violence (3.6 per cent). As with women, men with no education (0.6 per cent) or men with the highest levels of education (technical: none, and higher: 0.8 per cent) were the least likely to report having experienced at least one act of domestic psychological violence in the 12 months prior to the survey. The incidence of domestic psychological violence was highest among women in the middle asset quintile (8.2 per cent) and men in the second asset quintile (3.5 per cent). Women and men in the fourth and fifth asset quintiles were the least likely to report experiences of domestic 136 psychological violence (3.7 and 5.8 per cent, respectively, for women, and 0.8 and 1.0 per cent for men). These results suggest that women and men at intermediate levels of education and wealth were those most at risk of experiencing domestic psychological violence. 137 Table 37: Percentage of women who experienced domestic psychological violence over the last 12 months Insulted, Ignored, Scared, Attacked Threaten Any humiliated, threatened intimidated with to be hurt psychological belittled to be weapon violence abandoned Age group (years) 15–19 20.0 6.8 1.4 1.0 0.0 15.2 20–24 10.5 4.2 0.3 0.0 0.2 9.4 25–29 8.5 2.3 0.3 0.9 0.3 6.9 30–39 7.0 2.1 0.5 0.7 0.6 7.1 40–49 3.5 1.9 0.2 0.6 0.1 2.4 50–60 4.3 0.7 0.6 0.6 0.0 0.5 Employment Self-employed 6.5 2.0 0.4 0.5 0.3 5.1 Employed 5.8 2.4 0.4 1.2 0.0 4.7 Not working 11.2 3.7 0.8 0.9 0.4 9.1 Marital status Never married 15.8 3.5 1.2 0.8 0.0 9.0 Married or living together 6.9 2.3 0.4 0.6 0.2 5.9 Divorced/separated/widowed 7.8 2.8 0.8 0.8 0.7 5.7 Residence Urban 6.7 2.7 0.4 0.4 0.2 5.7 Rural 8.5 2.2 0.6 0.9 0.3 6.4 138 Insulted, Ignored, Scared, Attacked Threaten Any humiliated, threatened intimidated with to be hurt psychological belittled to be weapon violence abandoned Region Western 6.8 0.9 0.0 0.4 0.3 5.1 Central 8.8 4.8 0.5 0.0 0.5 6.9 Greater Accra 6.7 2.9 0.2 0.8 0.2 4.8 Volta 8.8 1.2 2.0 0.4 0.0 6.3 Eastern 11.3 4.0 2.2 2.2 0.0 8.7 Ashanti 7.3 2.3 0.0 0.5 0.3 6.7 Brong Ahafo 7.4 3.0 0.0 0.7 0.7 6.8 Northern 5.8 0.7 0.0 0.0 0.3 4.2 Upper East 1.0 0.0 0.0 0.0 0.0 0.8 Upper West 7.2 0.5 0.0 0.0 0.0 6.4 Education level None 5.1 1.6 0.6 0.3 0.3 2.9 Primary 9.4 3.5 0.9 1.4 0.7 9.2 Middle/JSS/JHS 8.2 2.4 0.3 0.5 0.1 6.7 Secondary 7.9 3.5 0.6 0.2 0.3 6.6 Technical 8.9 2.0 0.0 3.5 0.0 5.1 Higher 5.9 1.5 0.2 0.2 0.2 3.8 Asset quintile Lowest 10.1 2.3 1.1 1.2 0.6 6.3 Second 6.1 2.9 0.3 0.6 0.3 6.6 Middle 9.1 2.8 0.2 0.3 0.3 8.2 Fourth 7.5 1.6 0.3 0.1 0.2 3.7 Highest 5.3 2.5 0.6 1.1 0.0 5.8 Total 7.5 2.4 0.5 0.6 0.3 6.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 139 Table 38: Percentage of men who experienced domestic psychological violence over the last 12 months Insulted, Ignored, Scared, Attacked Threaten Any humiliated, threatened intimidated with to be hurt psychological belittled to be weapon violence abandoned Age group (years) 15–19 11.5 3.2 0.7 0.0 0.0 8.5 20–24 10.2 3.3 0.0 0.3 0.4 3.4 25–29 7.8 0.8 0.2 0.0 0.4 2.0 30–39 4.2 1.0 0.0 0.3 0.5 0.5 40–49 6.7 1.1 0.3 0.0 0.0 0.6 50–60 3.8 0.9 0.3 0.3 0.0 0.9 Employment Self-employed 6.1 1.6 0.1 0.3 0.2 1.5 Employed 6.9 1.0 0.4 0.0 0.6 2.0 Not working 8.8 2.2 0.3 0.0 0.0 3.8 Marital status Never married 8.3 3.1 0.0 0.0 0.0 5.6 Married or living together 5.8 1.1 0.3 0.3 0.2 1.4 Divorced/separated/widowed 9.6 1.5 0.0 0.0 0.7 1.5 Residence Urban 6.4 1.5 0.0 0.2 0.4 1.6 Rural 7.4 1.6 0.4 0.2 0.1 2.6 140 Insu lted, Igno red, Scared, Attac ked Thre aten Any humiliated, threatened intimidated with to be hurt psychological belittled to be weapon violence abandoned Region Western 4.2 0.6 0.0 0.0 0.0 1.5 Central 8.2 5.2 0.7 0.7 0.0 4.4 Greater Accra 6.4 0.9 0.0 0.4 0.7 1.6 Volta 9.5 1.5 0.0 0.0 0.0 2.2 Eastern 14.2 2.9 0.0 0.0 0.6 3.8 Ashanti 8.6 1.5 0.0 0.0 0.4 2.1 Brong Ahafo 2.5 1.1 0.4 0.4 0.0 1.8 Northern 5.2 1.1 0.7 0.0 0.0 0.7 Upper East 1.1 0.6 1.1 0.0 0.0 1.8 Upper West 0.0 0.0 0.0 0.0 0.0 3.2 Education level None 3.9 0.8 0.0 0.0 0.0 0.6 Primary 8.1 2.9 0.6 0.0 0.4 3.1 Middle/JSS/JHS 7.2 1.3 0.2 0.2 0.0 2.3 Secondary 8.2 2.7 0.0 0.4 0.0 3.6 Technical 4.2 1.5 0.0 0.0 0.0 0.0 Higher 6.5 0.0 0.3 0.0 1.3 0.8 Asset quintile Lowest 6.1 1.7 0.0 0.0 0.0 3.2 Second 7.2 2.5 0.6 0.3 0.0 3.5 Middle 7.9 1.8 0.3 0.0 0.7 1.5 Fourth 7.0 1.1 0.0 0.0 0.0 0.8 Highest 6.2 0.0 0.0 0.5 0.7 1.0 Total 6.9 1.5 0.2 0.2 0.2 2.1 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 141 4.2.5. Domestic economic violence For most acts of economic violence against women, the majority were perpetrated by domestic relations, with the exception of cash being taken and property damaged: 34.4 and 46.4 per cent of these acts, respectively, were committed by domestic relations (Table 39). A large part of economic violence against men was perpetrated outside domestic relations, with three exceptions: 83.2 per cent of violence in the form of being denied pocket money, 92.3 per cent of violence in the form of having belongings controlled, and 52.2 per cent of violence in the form of being prohibited from working were perpetrated by domestic relations (Table 39). All forms of domestic economic violence against men were more likely to be committed by non-family members than by family members. Most forms of domestic economic violence against women were also committed by non-family members, with the exception of having their belongings controlled (60.9 per cent committed by family members, and 34.8 per cent by non-family members) and being prohibited from working (48.6 per cent committed by family members, and 28.6 per cent by non-family members). Table 39: Percentage of domestic economic violence over the last 12 months Denied Cash Belongings Property Prohibited Forced to Denied pocket taken out controlled damaged from work food money working Women Perpetrated by family 44.1 15.6 60.9 3.6 48.6 30.0 36.4 members Perpetrated by non- family members but 52.3 18.8 34.8 42.9 28.6 33.3 59.1 domestic % Domestic violence 96.0 34.4 91.3 46.4 74.3 63.3 95.4 % Non-domestic 4.0 65.6 8.7 53.6 25.7 36.7 4.6 vTiootlaeln ce 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Men Perpetrated by family 28.7 15.1 38.5 5.7 0.0 7.7 23.1 members Perpetrated by non- family members but 54.5 18.2 53.8 15.1 52.2 25.6 50.0 domestic % Domestic violence 83.2 33.3 92.3 18.9 52.2 33.3 73.1 % Non-domestic 16.8 66.7 7.7 81.1 47.8 66.7 26.9 vTiootlaeln ce 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Tables 40 and 41 illustrate how domestic economic violence was distributed across the socio-economic characteristics and geographical location of women and men in Ghana. Domestic economic violence was experienced by 12.8 per cent of women and 7.3 per cent of men interviewed in the GFLHS 2015. The most common act of domestic economic 142 violence experienced by women and men was being denied pocket money – known in Ghana as chop money. This act of domestic economic violence was experienced by 11.2 per cent of women and 4.9 per cent of men. Less than 1 per cent of the sample reported having experienced any other act of domestic economic violence. Age patterns. The incidence of domestic economic violence overall was highest among women aged 25–29 years (19.7 per cent), in comparison to 17.2 per cent among women aged 15–19 years, 15.9 per cent among women aged 20–24 years, 13 per cent among women aged 30–39 years, 8.1 per cent among women aged 40–49 years and 5.1 per cent among women in the oldest age group (50–60 years old). However, women in the youngest age group (15–19 years) were at a greater risk of particular acts of domestic violence. Notably, being denied food was highest among this group (5.5 per cent) than among any other age group (Table 40). These women also reported the highest incidence of being forced to work (1.5 per cent) and having their belongings controlled (1.6 per cent). The incidence of domestic economic violence among men was highest among the youngest age group (15– 19 years), at 19 per cent, in comparison to 3.4 per cent among the oldest age group (50–60 years). Employment status. Women and men who were not working at the time of the survey were more likely to report having experienced at least one act of domestic economic violence in the 12 months prior to the survey (16.6 per cent of women and 15.3 per cent of men). However, the incidence of certain types of domestic violence, such as being prohibited from working, was highest among self-employed women (1.2 per cent), whereas employed men were the most likely group to report having their cash taken (0.6 per cent) or being denied food (2.1 per cent). Employed women (8.7 per cent) and self- employed men (3.9 per cent) were the least likely to report domestic economic violence. Marital status. The association between marital status and the incidence of domestic economic violence differs between women and men. Married women (13.4 per cent) were more likely to experience domestic economic violence than never-married women (10.1 per cent) and women who were divorced, separated or widowed (11.6 per cent). The incidence of domestic economic violence was highest among never-married men (13.4 per cent), in comparison to 9.9 per cent among men who were divorced, separated or widowed and 4.9 per cent among men who were married. Geographical patterns. The incidence of domestic economic violence was higher among women living in rural areas (15.9 per cent) than among those living in urban areas (10.4 per cent). The incidence of domestic economic violence was also higher among men living in rural areas (7.4 per cent), but the difference in relation to men living in urban areas (7.3 per cent) was small. Women living in the Western region (22.8 per cent) and in the Upper East region (6.7 per cent) were, respectively, the most likely and least likely to report incidence of domestic economic violence. The incidence of domestic economic violence was highest among men in the Central region (16.7 per cent) and lowest among men in the Upper East region (2 per cent). 143 Education levels and asset ownership. Women with primary education were the most likely to report experiences of domestic economic violence (16.7 per cent). The lowest incidence of domestic economic violence was observed among women with either higher education (8.1 per cent) or women with no education (9.6 per cent). Patterns for men were similar: the incidence of domestic economic violence was lowest among men with no education (1.4 per cent) or the highest level of education (5.6 per cent), and highest among men with secondary education (9.6 per cent). Women and men in the second asset quintile were the most likely to report having experienced domestic economic violence (15.3 and 9.1 per cent, respectively), while women and men in the top asset quintile had the lowest incidence of domestic economic violence (6.9 and 5.8 per cent, respectively). These results suggest that individuals with intermediate levels of education and wealth may be at higher risk of economic violence than better-off and more educated individuals. 144 Table 40: Percentage of women who experienced domestic economic violence over the last 12 months Denied Cash Belongings Property Prohibited Forced Denied Any pocket taken controlled damaged from to work food economic money out working violence Age group (years) 15–19 12.9 0.6 1.6 0.2 0.0 1.5 5.5 17.2 20–24 13.2 0.7 1.0 0.5 1.1 1.2 1.9 15.9 25–29 16.7 0.9 1.0 1.1 1.8 1.3 1.2 19.7 30–39 12.1 0.1 0.8 0.5 1.0 0.1 1.2 13.0 40–49 7.8 0.1 0.3 0.0 0.6 0.5 0.2 8.1 50–60 4.2 0.0 0.1 0.1 0.5 0.0 0.3 5.1 Employment Self-employed 10.9 0.4 0.4 0.4 1.2 0.5 1.1 12.1 Employed 7.6 0.0 1.0 0.4 0.4 0.2 0.3 8.7 Not working 13.5 0.5 1.4 0.7 0.5 1.0 2.2 16.6 Marital status Never married 6.1 0.0 0.9 0.0 0.0 1.9 3.7 10.1 Married or living together 12.0 0.5 0.7 0.4 1.1 0.5 0.9 13.4 Divorced/separated/widowed 9.7 0.0 0.7 0.8 0.7 0.6 1.9 11.6 Residence Urban 8.8 0.3 0.9 0.4 0.6 0.4 0.7 10.4 Rural 13.8 0.5 0.5 0.5 1.4 0.9 1.9 15.6 145 Denied Cash Belongings Property Prohibited Forced Denied Any pocket taken controlled damaged from to work food economic money out working violence Region Western 22.4 0.3 0.4 0.7 0.0 0.0 1.2 22.8 Central 10.8 0.9 0.9 1.1 0.4 0.5 2.4 12.8 Greater Accra 6.8 0.0 1.0 0.0 0.1 0.4 0.2 7.7 Volta 5.7 0.5 0.0 0.0 0.2 0.4 0.7 6.9 Eastern 15.7 0.9 1.1 0.8 0.9 0.8 1.3 18.3 Ashanti 12.2 0.4 0.4 0.5 1.3 0.0 1.4 14.2 Brong Ahafo 11.7 0.0 0.5 0.8 3.2 1.6 1.9 14.6 Northern 6.9 0.3 0.9 0.0 1.8 2.1 2.5 8.7 Upper East 6.1 0.0 1.1 0.0 0.0 0.0 0.6 6.7 Upper West 12.5 0.0 2.4 0.0 2.4 1.5 0.0 13.9 Education level None 8.6 0.2 0.7 0.1 0.9 0.5 1.0 9.6 Primary 14.7 0.7 0.9 0.1 0.9 0.3 0.7 16.7 Middle/JSS/JHS 12.3 0.5 0.6 0.8 1.2 0.8 2.0 14.2 Secondary 10.2 0.0 0.4 0.9 1.0 0.6 0.9 11.9 Technical 9.8 0.0 0.0 0.0 1.0 0.0 0.8 10.8 Higher 6.7 0.0 1.4 0.0 0.0 0.5 0.0 8.1 Asset quintile Lowest 11.1 0.3 0.5 0.2 1.4 1.6 2.2 13.4 Second 13.7 0.3 0.8 0.7 1.3 0.7 1.2 15.3 Middle 14.5 0.1 0.8 0.5 0.3 0.0 2.0 15.2 Fourth 9.2 0.7 1.0 0.5 1.0 0.5 0.5 11.6 Highest 6.0 0.2 0.5 0.2 0.5 0.0 0.4 6.9 Total 11.2 0.4 0.7 0.4 1.0 0.6 1.2 12.8 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 146 Table 41: Percentage of men who experienced domestic economic violence over the last 12 months Denied Cash Belongings Property Prohibited Forced Denied Any pocket taken controlled damaged from to work food economic money out working violence Age group (years) 15–19 14.3 1.2 2.9 0.9 0.5 2.4 1.6 19.0 20–24 8.9 1.3 0.1 1.5 1.5 1.8 3.0 13.6 25–29 4.8 0.6 0.0 0.4 0.8 1.2 0.9 7.3 30–39 1.8 0.0 0.4 0.1 0.2 0.0 0.0 2.5 40–49 2.3 0.2 0.6 0.2 0.9 0.3 0.8 4.9 50–60 2.5 0.0 0.8 0.8 0.3 0.0 0.6 3.4 Employment Self-employed 2.4 0.4 0.1 0.1 0.6 0.6 0.3 3.9 Employed 4.2 0.6 0.9 0.6 0.8 1.1 2.1 7.5 Not working 12.0 0.5 1.6 1.7 0.8 1.0 1.4 15.3 Marital status Never married 10.4 0.7 1.5 0.5 0.4 1.4 2.1 13.4 Married or living together 3.0 0.2 0.5 0.1 0.5 0.7 0.5 4.9 Divorced/separated/widowed 5.9 1.1 0.3 1.6 1.6 0.8 1.6 9.9 Residence Urban 4.9 0.2 1.0 0.7 0.5 0.8 1.0 7.3 Rural 4.9 0.8 0.2 0.5 0.9 0.8 1.1 7.4 147 Den ied Ca sh Belon gings Prop erty Prohi bited Forc ed Den ied An y pocket taken controlled damaged from to work food economic money out working violence Region Western 7.7 0.2 0.3 0.2 0.0 0.0 0.5 8.7 Central 8.6 2.3 4.5 0.0 1.3 1.3 5.6 16.7 Greater Accra 5.3 0.0 1.1 0.0 0.9 1.5 0.6 6.3 Volta 2.0 0.0 0.0 0.0 0.0 0.0 0.8 2.3 Eastern 2.3 2.2 0.8 0.6 2.8 1.0 2.1 9.4 Ashanti 6.9 0.2 0.0 2.4 0.0 0.7 0.7 9.1 Brong Ahafo 4.4 0.5 0.0 0.3 1.3 0.0 0.0 6.5 Northern 2.2 0.0 0.0 0.3 0.0 1.8 0.3 4.6 Upper East 2.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 Upper West 2.2 0.0 0.0 0.0 2.2 2.2 0.0 2.2 Education level None 0.6 0.0 0.0 0.2 0.0 0.6 0.3 1.4 Primary 4.7 0.7 1.0 0.0 1.7 0.6 1.7 8.6 Middle/JSS/JHS 5.8 0.7 0.9 0.4 1.0 0.9 1.0 8.7 Secondary 7.7 0.6 0.1 1.2 0.3 1.2 1.3 9.6 Technical 5.0 0.0 0.0 0.0 0.0 2.6 1.5 6.5 Higher 3.3 0.2 1.1 1.0 0.4 0.0 0.6 5.6 Asset quintile Lowest 6.2 0.2 0.5 0.5 0.8 0.9 0.9 7.2 Second 5.0 0.7 0.5 0.7 1.4 1.5 1.7 9.1 Middle 5.3 0.7 0.3 0.7 0.0 0.6 1.0 7.4 Fourth 5.2 0.2 1.0 0.4 0.4 0.6 0.6 6.5 Highest 2.6 0.6 0.9 0.6 0.6 0.0 0.6 5.8 Total 4.9 0.5 0.6 0.6 0.7 0.8 1.0 7.3 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 148 4.2.6. Summary The information discussed in the sections above suggests important patterns in how domestic violence was experienced among men and women, and different socio-economic groups and geographical locations in Ghana. Some key findings include:  Younger women and men were overall at higher risk than older age groups of experiencing any form of domestic violence.  Domestic physical or psychological violence appears to have been more prevalent among non-working, never-married men and women with intermediate levels of education and wealth and living in rural areas.  Domestic economic violence exhibits a similar pattern as above but was more prevalent among married women.  Domestic social violence was more prevalent among non-working and urban women and men with some level of education and wealth.  Domestic sexual violence differs from most other forms of domestic violence: it appears to be most prominent among employed, better-off, less economically dependent individuals living in urban areas. Table 42 summarises these findings, highlighting the categories for which each type of violence was most prominent. These findings are explored in detail in the remainder of Section 4. Table 42: Socio-economic groups with highest incidence of domestic violence Social Physical Sexual Psychological Economic violence violence violence violence violence Age (years) Women 15–19 15–19 20–24 15–19 25–29 Men 15–19 15–19 30–39 15–19 15–19 Employment status Women Not working Not working Employed; Not working Not working not working Men Not working Not working Employed Not working Not working Marital status Women Not married Not married Married Not married Married Men Divorced Not married Divorced Not married Not married Residence Women Urban Rural Urban Rural Rural Men Urban Rural Urban Rural Rural Education Women Secondary Primary Secondary Primary Primary Men Higher Secondary Technical; Secondary Secondary higher Asset quintile Women Third Third Fourth Third Second Men Highest Second Highest Second Second Source: Ghana Family Life and Health Survey (GFLHS) 2015. 149 Qualitative data across all regions supported the findings discussed above: both men and women experienced domestic violence, albeit in different forms and with significant variations across individual, household, community and regional characteristics. The data from both the focus group discussions and the in-depth interviews suggested that women experienced two predominant forms of domestic violence: economic violence (in the form of men’s refusal to provide chop money) and physical violence (in the form of wife-beating). The qualitative data showed, in addition and in line with the quantitative results above, that men experienced domestic violence, by women and sometimes by other men, largely in more public spaces – streets or marketplaces. The forms of domestic violence that men experienced in these spaces were predominantly verbal abuse (such as women shouting at their husbands) and psychological abuse (such as women shaming their husbands in public). This finding was nationally representative, and was illustrated, for example, by the following two respondents in the Gbawe mixed FGD (18–25): Respondent 8: Women in times of violence are physically assaulted, but men are verbally abused. Respondent 5: Women are abused physically and verbally, but men are just verbally abused. 4.3. Perpetration of domestic violence in Ghana The results from the survey, as discussed above, showed that the perpetration of domestic violence varied considerably between men and women, with women being more likely than men to have experienced violence perpetrated within domestic relations. This section discusses in more detail the profiles of domestic violence perpetrators. The section starts by analysing patterns of violence perpetration within domestic relations (subsection 4.3.1). Subsection 4.3.2 discusses the profiles of study respondents who reported having committed social, physical, sexual, psychological and/or economic violence against another domestic relation. Subsection 4.3.3 focuses on the perpetration of violence against children and young people. 4.3.1. Violence perpetration within domestic relations Tables 43 and 44 provide an overview of the main domestic perpetrators of violence for each type of violence – social, physical, sexual, psychological and economic – as reported by female and male respondents with respect to the 12 months prior to the survey. It is important to note that more than one perpetrator could perform different types of violence. For simplicity, this table shows the aggregate figures of respondents’ first response. 150 Table 43: Perpetrators of domestic violence as reported by women (in percentages) Panel A: Respondent is head of household (N=1,304) Social Sexual Physical Psychological Economic Partner/former partner 48.0 92.0 78.0 34.0 74.0 Spouse 0.0 0.0 5.0 3.0 2.0 In-laws 5.0 3.0 3.0 8.0 2.0 Child 2.0 0.0 0.0 0.0 2.0 Parent 12.0 0.0 1.0 11.0 8.0 Brother, sister 17.0 0.0 5.0 22.0 7.0 Uncle, aunt 2.0 3.0 1.0 6.0 4.0 Cousin, niece, nephew 4.0 0.0 6.0 8.0 0.0 Other relative 9.0 3.0 2.0 9.0 2.0 N 174 36 103 158 183 Panel B: Respondent is a spouse of the head of household (N=958) Social Sexual Physical Psychological Economic Head of household 63.0 96.0 81.0 47.0 83.0 Partner/former partner 1.0 2.0 0.0 4.0 2.0 Children 7.0 0.0 1.0 1.0 2.0 Parent of head 11.0 0.0 0.0 17.0 0.0 Parent 3.0 0.0 5.0 0.0 1.0 Brother, sister 4.0 0.0 4.0 5.0 6.0 Uncle, aunt 1.0 0.0 0.0 4.0 2.0 Cousin, niece, nephew 3.0 0.0 4.0 0.0 1.0 Other relative 8.0 0.0 4.0 17.0 2.0 N 123 24 75 78 122 Panel C: Respondent is the daughter/son of the head of household (N=246) Social Sexual Physical Psychological Economic Parent 65.0 18.0 22.0 51.0 50.0 Brother, sister 9.0 0.0 13.0 16.0 16.0 Partner of head 0.0 0.0 3.0 5.0 0.0 Partner/former partner 18.0 64.0 56.0 14.0 29.0 Uncle, aunt 3.0 0.0 0.0 9.0 3.0 Cousin, niece, nephew 3.0 0.0 6.0 2.0 0.0 Other relative 3.0 18.0 0.0 2.0 3.0 N 118 11 32 43 38 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Note: The totals N refer, respectively, to the number of respondents who were the head of the household, the number of respondents who were spouses of the head of the household and the number of respondents who were either the daughter or son of the head of the household. The most predominant perpetrators of domestic violence against women were their partners or former partners: the share of perpetration by (former) partners ranged between 34 per cent (psychological violence) and 92 per cent (sexual violence) against female household heads (Table 43, Panel A). The share of husbands perpetrating violence against their wives ranged between 47 per cent (psychological violence) and 96 per cent (sexual 151 violence) (Table 43, Panel B). Parents, in-laws and brothers and sisters were also identified as perpetrators of domestic violence. Physical strength was one of the main factors used by focus group discussion participants across all regions to describe the prevalence of men’s physical violence against women in the household. In-depth interviews with key community members confirmed this finding – for example, in the description offered by a community leader working in the Eastern region, in the East Akim district: “The difference between how women experience violence and that of men is that the men are stronger and do harm the women when they beat them.” Table 44: Perpetrators of domestic violence as reported by men (in percentages) Panel A: Respondent is head of households (N=1,655) Social Sexual Physical Psychological Economic Partner/former partner 40.0 78.0 11.0 29.0 15.0 Spouse 21.0 2.0 1.0 7.0 9.0 In-laws 2.0 0.0 4.0 3.0 0.0 Child 3.0 0.0 2.0 0.0 3.0 Parent 8.0 0.0 18.0 15.0 28.0 Brother, sister 8.0 0.0 18.0 15.0 25.0 Uncle, aunt 2.0 7.0 10.0 10.0 7.0 Cousin, niece, nephew 5.0 0.0 14.0 7.0 5.0 Other relative 11.0 10.0 18.0 13.0 8.0 N 149 30 49 157 113 Panel B: Respondent is a spouse of the head of the household (N=5) Social Sexual Physical Psychological Economic 0.0 0.0 0.0 0.0 0.0 Panel C: Respondent is the daughter/son of the head of household (N=219) Social Sexual Physical Psychological Economic Parent 56.0 0.0 42.0 62.0 84.0 Brother, sister 24.0 0.0 17.0 23.0 13.0 Partner/former partner 15.0 100.0 0.0 4.0 0.0 Uncle, aunt 3.0 0.0 8.0 0.0 0.0 Cousin, niece, nephew 0.0 0.0 0.0 0.0 3.0 Other relative 3.0 0.0 33.0 8.0 0.0 N 34 2 12 26 31 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Note: The totals N refer, respectively, to the number of respondents who were the head of the household, the number of respondents who were spouses of the head of the household and the number of respondents who were either the daughter or son of the head of the household. Domestic violence against male respondents was also perpetrated largely by (former) partners. Economic or physical violence against male heads of household was mainly perpetrated by parents (28 and 18 per cent, respectively) and siblings (25 and 18 per cent, respectively) (Table 44). Men who experienced violence committed by parents and siblings were generally younger men who lived with their family in the same household. 152 4.3.2. Profile of domestic violence perpetrators The qualitative data from all regions confirmed the range of perpetrators discussed above. Within the domestic setting, the majority of the perpetrators were men, who committed violence against their partners (generally women, in heterosexual relationships). As discussed above, these forms of violence were predominantly economic and physical. Complementing the quantitative findings shown in Tables 43 and 44, the qualitative research revealed that domestic violence was both horizontal (between partners) and vertical (across generations, between caregivers and children). Further, horizontal forms of violence did not only flow one way, exerted by men against their female partners. The focus group discussions and in-depth interviews discussed – sometimes at length, particularly in the Eastern, Ashanti and Volta regions – how men were abused by women, something that is also revealed in Tables 43 and 44. Through careful facilitation of the focus group discussions, particularly in groups of married men and of unmarried men and women, the study found that men also experienced physical violence. This was not a nationally or regionally representative finding, but it was a finding reported in urban areas, where men spoke about psychological and verbal abuse from their female spouses. In the Greater Accra region, one male participant explained to the rest of the mixed male and female group (18–25 years old) that women can be stronger than men, and that they can be physically violent to their spouses, just like men: “The one I know the woman is much stronger than the man. So she does not obey the rules of the man. If the man tries to talk she just beats the man up.” The GFLHS 2015 collected information that allowed the study to explore further the direction of the perpetration of violence between men and women. Table 45 shows the share and number of respondents in the survey who reported having perpetrated acts of domestic violence in the 12 months prior to the survey. Table 45: Percentage and number of domestic violence perpetrators Percentage of respondents reporting perpetration of domestic violence Sex of Social Sexual Physical Psychological Economic Any domestic perpetrator violence violence violence violence violence violence Male 0.12 0.06 0.06 0.15 0.06 0.30 Female 0.12 0.01 0.07 0.15 0.03 0.27 Total 0.12 0.03 0.07 0.15 0.04 0.28 153 Number of respondents reporting perpetration of domestic violence Sex of Social Sexual Physical Psychological Economic Any domestic perpetrator violence violence violence violence violence violence Male 222 120 110 277 97 563 Female 362 26 208 422 79 789 Total 584 146 318 699 176 1,352 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Table 45 reveals that 563 men (0.3 per cent of the sample) and 789 women (0.27 per cent of the sample) reported having perpetrated at least one act of domestic violence. Men were more likely to report perpetrating sexual or economic violence (0.06 per cent in both cases, corresponding to 120 and 97 men, respectively). More women than men reported having perpetrated social, physical or psychological violence. These results must be interpreted with great care because they are likely to be vastly under-reported. The results may also exhibit reporting biases. Given the mismatch between the results in Table 45 and the qualitative data, it is possible that women may have been more prone to report perpetration of violence than men. With these caveats in mind, it is nonetheless noteworthy that violence is perpetrated by both sexes. Table 46 provides further disaggregation of the above results. Table 46: Number and percentage of perpetrators reporting violence against same or another sex Social Sexual Physical Psychological Economic Any violence violence violence violence violence domestic violence Female perpetrators against male or female household members Against men 176 (0.62) 3 (1.0) 90 (0 .60) 106 (0.69) 30 (0.59) 248 (0.67) Against women 193 (0.69) - 92 (0.61) 93 (0.61) 33 (0.65) 233 (0.63) Total number 281 3 150 153 51 370 Male perpetrators against male or female household members Against men 71 (0.55) - 24 ( 0.69) 31 (0.49) 12 (0.24) 84 (0.50) Against women 95 (0.73) 9 (0.90) 17 (0.49) 48 (0.76) 43 (0.86) 126 (0.75) Total number 130 10 35 63 50 169 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Table 46 shows that the largest share of domestic violence committed by women was against men (0.67 per cent), but the difference between this and the share of violence committed by women against other women (0.63 per cent) is small. Women were most likely to commit violence against other women in the cases of social violence, physical 154 violence and economic violence (0.69, 0.61 and 0.65 per cent of women reported having committed these forms of violence, respectively, against other women). Most of the violence committed by men was against women (0.75 per cent, in comparison to 0.50 per cent against other men). It is again important to keep in mind that these estimates may be problematic, as discussed above. However, Tables 45 and 46 illustrate the level of complexity and variety of domestic violence perpetration in Ghana. Episodes of physical violence were also described as dynamic in that they could be driven at first by one partner but ‘ended’ by another. This quote illustrates such dynamics, widely reported at a regional level (but not nationally representative): “The man gave the woman some blows on any part of her body that was available until she fell down. The woman stood up and held the man by the neck, and the man began to give her some blows until both of them fell down. They were so tired that they lay down till everybody left the scene.” (FGD, Ashanti region, Adaase district: male and female group, aged 18+) In some cases, the instance of physical violence followed a period of other forms of violence. For example, some focus group discussions described how women struggled with economic violence (such as men’s refusal to provide chop money) and that this violence was compounded when women were ‘punished’ by their partner for failing to cook food (itself a retaliation against the failure to provide money). Structural violence, and pernicious poverty, might underlie this dynamic of intersecting forms of violence, as men were not always able to secure wage labour and struggled with the pressure to ensure the economic well-being of their household and its members (as is discussed later in the report). The qualitative data found, too, that some women withheld sex or shamed their partners (through verbal abuse) when they did not earn enough money to support the running costs of the household. This suggests an interlinked ‘web of violence’ that requires multiple approaches for redress. The study also collected information on how the perpetration of domestic violence varied across socio-economic characteristics and geographical location of women and men. These data are summarised in Tables 47 and 48, which suggest the following profiles:  The incidence of domestic violence perpetrated by women was highest among self- employed and married women aged 25–29 years who lived in rural areas with no or primary education and in the lowest asset quintile.  The incidence of domestic violence perpetrated by men was highest among men aged 15–19 years who were employed and married, who lived in rural areas, were in the lowest asset quintile but had primary or secondary education. 155 Table 47: Percentage of women reporting perpetration of domestic violence Social Sexual Physical Psychological Economic Any domestic violence Age group (years) 15–19 0.07 0.01 0.08 0.19 0.01 0.29 20–24 0.10 0.02 0.05 0.11 0.01 0.23 25–29 0.14 0.01 0.10 0.20 0.02 0.33 30–39 0.13 0.01 0.09 0.15 0.04 0.29 40–49 0.14 0.01 0.07 0.14 0.02 0.27 50–60 0.09 0.00 0.03 0.09 0.04 0.18 Employment Self-employed 0.14 0.01 0.08 0.15 0.03 0.28 Employed 0.10 0.01 0.09 0.11 0.02 0.24 Not working 0.08 0.01 0.06 0.16 0.03 0.25 Marital status Never married 0.04 0.00 0.09 0.17 0.02 0.25 Married or living together 0.12 0.01 0.08 0.14 0.03 0.27 Divorced/separated/widowed 0.13 0.01 0.05 0.15 0.03 0.26 Residence Urban 0.11 0.01 0.07 0.13 0.02 0.25 Rural 0.13 0.00 0.08 0.17 0.03 0.29 Region Western 0.14 0.01 0.09 0.16 0.04 0.29 Central 0.06 0.01 0.04 0.08 0.04 0.18 Greater Accra 0.10 0.02 0.07 0.11 0.03 0.21 Volta 0.09 0.01 0.11 0.31 0.02 0.39 Eastern 0.23 0.01 0.09 0.18 0.03 0.37 Ashanti 0.12 0.01 0.07 0.15 0.02 0.29 Brong Ahafo 0.08 0.00 0.06 0.11 0.02 0.20 Northern 0.14 0.00 0.13 0.18 0.03 0.33 Upper East 0.14 0.00 0.04 0.04 0.01 0.16 Upper West 0.15 0.04 0.04 0.08 0.02 0.20 Education level None 0.13 0.01 0.08 0.15 0.03 0.28 Primary 0.14 0.00 0.08 0.16 0.04 0.28 Middle/JSS/JHS 0.12 0.01 0.07 0.15 0.02 0.27 Secondary 0.09 0.02 0.08 0.12 0.02 0.25 Technical 0.07 0.03 0.11 0.12 0.00 0.24 Higher 0.08 0.02 0.06 0.11 0.02 0.22 Asset quintile Lowest 0.11 0.01 0.09 0.18 0.04 0.29 Second 0.14 0.01 0.07 0.17 0.03 0.28 Middle 0.10 0.01 0.07 0.16 0.02 0.27 Fourth 0.12 0.01 0.07 0.12 0.02 0.25 Highest 0.12 0.02 0.08 0.09 0.02 0.25 Total 0.12 0.01 0.07 0.15 0.03 0.27 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 156 Table 48: Percentage of men reporting perpetration of domestic violence Social Sexual Physical Psychological Economic Any domestic violence Age group (years) 15–19 0.07 0.06 0.09 0.22 0.04 0.35 20–24 0.12 0.12 0.07 0.18 0.04 0.32 25–29 0.12 0.10 0.07 0.14 0.04 0.33 30–39 0.10 0.04 0.05 0.14 0.07 0.27 40–49 0.16 0.04 0.04 0.12 0.09 0.33 50–60 0.11 0.02 0.04 0.14 0.04 0.25 Employment Self-employed 0.12 0.05 0.05 0.13 0.05 0.28 Employed 0.12 0.07 0.07 0.15 0.08 0.34 Not working 0.11 0.08 0.06 0.19 0.03 0.32 Marital status Never married 0.06 0.05 0.07 0.16 0.03 0.25 Married or living together 0.14 0.06 0.06 0.16 0.07 0.33 Divorced/separated/widowed 0.10 0.09 0.05 0.12 0.03 0.27 Residence Urban 0.10 0.08 0.06 0.15 0.05 0.30 Rural 0.13 0.05 0.06 0.16 0.06 0.31 Region Western 0.12 0.04 0.03 0.17 0.12 0.34 Central 0.09 0.09 0.08 0.13 0.03 0.26 Greater Accra 0.10 0.11 0.05 0.10 0.06 0.26 Volta 0.07 0.07 0.13 0.33 0.03 0.41 Eastern 0.24 0.05 0.08 0.17 0.05 0.40 Ashanti 0.13 0.07 0.05 0.16 0.04 0.33 Brong Ahafo 0.09 0.06 0.04 0.10 0.08 0.23 Northern 0.08 0.02 0.03 0.13 0.02 0.24 Upper East 0.13 0.00 0.07 0.05 0.05 0.23 Upper West 0.14 0.03 0.00 0.06 0.03 0.24 Education level None 0.13 0.02 0.05 0.15 0.05 0.29 Primary 0.14 0.10 0.07 0.16 0.06 0.36 Middle/JSS/JHS 0.10 0.05 0.05 0.15 0.05 0.29 Secondary 0.10 0.10 0.07 0.16 0.05 0.32 Technical 0.09 0.08 0.07 0.15 0.05 0.25 Higher 0.14 0.05 0.05 0.15 0.08 0.31 Asset quintile Lowest 0.10 0.04 0.06 0.16 0.06 0.31 Second 0.11 0.07 0.05 0.16 0.05 0.31 Middle 0.12 0.06 0.06 0.13 0.06 0.28 Fourth 0.13 0.07 0.07 0.16 0.06 0.31 Highest 0.13 0.08 0.07 0.14 0.05 0.30 Total 0.12 0.06 0.06 0.15 0.06 0.30 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 157 4.3.3. Domestic violence against children In addition to the forms of domestic violence discussed in the sections above, the qualitative data noted the occurrence of violence against children by adults who shared the same home. The qualitative data, from all regions, found that women and men employed physical violence to punish children, but there were different views on whether physical violence, in particular, was an appropriate form of punishment for misbehaving children. These views also differed depending on whether the person committing the violence, or enacting the punishment, was male or female. For example, participants in a focus group discussion with unmarried men and women (18–25 years old) in the Greater Accra region said that only the father should beat children “because he is the breadwinner”. In another focus group, also in the Greater Accra region, one respondent linked beating with ‘correction’, saying, “fathers are responsible for correcting children through beating” (FGD, Greater Accra region, Kukurantami district: male and female group, 18–25 years old). Elsewhere, however, participants in the focus groups argued that this responsibility should not be left only to men. In a focus group with married men (18–25 years old) in the Eastern region, one participant asserted, “women should beat the children and the youth in the house whenever they do something bad”. The GFLHS 2015 on domestic violence committed by adults against their biological children found that the main perpetrator was the head of the household (Table 49). Table 49: Perpetration of domestic violence against individuals who reported domestic violence and lived with their parents (in percentages) Social Sexual Physical Psychological Economic violence violence violence violence violence Male head of household 88.0 0.0 100.0 87.0 82.0 Female head of household 87.0 100.0 63.0 70.0 87.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Note: Values were reported by the household respondent and refer to children of the household head. The qualitative data also highlighted that non-biological children may be at particular risk of domestic violence inflicted by stepfathers, particularly when the mother is in a vulnerable economic position. The following statement reflects this widespread finding across regions: “[Domestic violence] is a multi-faceted problem involving couples, children and at times in- laws as well. Yes, at times it involves adults and their non-biological children. You sometimes hear of men making advances towards the daughter of his wife who was brought into the marriage. You can’t believe this, but it is true. When the man is supposed to treat that child as his own daughter, he rather decides to make amorous advances towards her.” (IDI, Upper East region, Garu Tempane district assembly, Garu opinion leader) 158 In addition to physical violence, controlling behaviour, such as fathers refusing to let their daughters attend school, was noted as a form of domestic violence across all regions in the qualitative data. The focus group participants were in significant agreement that this form of violence was unacceptable. Given the nature of the qualitative data, it was not possible, as with the quantitative survey, to gauge the full extent of these different forms of violence. However, the data offer important insights into some of the underlying social norms, beliefs and attitudes that individuals, families and communities hold around certain forms of violence and their acceptability. These findings are discussed in further detail in the section below. 4.4. Attitudes towards domestic violence in Ghana The second research question asked how attitudes and social norms may influence the level of acceptance of domestic violence in Ghana. Attitudes and beliefs – even if held at a personal level – are usually shaped by interactions with people within the family; with people in the immediate environment, including communities, neighbourhoods, social networks and workplaces; and within economic or political structures. Understanding how these reproduce or challenge gender inequality, gender norms and domestic violence is crucial for designing interventions and policies that will effectively decrease the incidence of domestic violence. For this reason, this component of the study sought to explore, first, how individuals perceived and experienced domestic violence in their own lives and families (the individual and community in the social ecological model) and, second, how perceptions and experiences of individuals were informed by social norms that may be collectively reinforced at the level of the community and broader social contexts. During the quantitative data collection, men and women were asked to discuss five different scenarios that involved acts of violent behaviour. The aim of this exercise was to understand to what extent these violent acts were perceived as acceptable, to what extent victims of domestic violence were blamed for the violence (and under which circumstances), and what respondents felt would be acceptable punishments for different acts of violence. This information was complemented with qualitative data from focus group discussions and interviews. To generate data to complement the survey, the focus groups were also asked to respond to and discuss similar scenarios of social, physical, sexual, psychological and economic violence. The focus group discussions and in-depth interviews with opinion leaders, activists and legal experts explored in addition whether and why certain types of behaviour were considered acceptable. An important finding from the research centred on the variation in people’s understanding of what constitutes: (i) violence in general; and (ii) domestic violence in particular. As a result, the study offers a set of nuanced findings that complicate a straightforward reading of domestic violence as per the Domestic Violence Act. Importantly, these nuanced findings suggest that successful interventions need to be carefully tailored to raise 159 awareness around each of the categories of violence that are included under the general term of ‘domestic violence’. Given the scale of the qualitative and quantitative research, this section has been structured to reveal findings for each of the five main types of violence explored in the study – namely: social, physical, sexual, psychological and economic. While data have been disaggregated against these forms of violence, it is important to recognise that individuals can experience multiple forms of violence simultaneously; wherever possible, the multiplicity of domestic violence is highlighted in the discussion below. 4.4.1. Attitudes towards domestic social violence Both the quantitative and qualitative data collection made use of vignette scenarios to understand individual and group-level attitudes towards domestic social violence. The GFLHS 2015 included a range of vignette scenarios, and survey respondents were randomly assigned scenario questions. In doing so, the survey generated findings about different attitudes towards domestic violence by sex and other socio-economic characteristics of the respondents. One of the first scenarios intended to find out about the acceptability of jealous behaviour by a girlfriend or a boyfriend. Examples of jealous behaviour in the vignette included: checking messages on their partner’s phone; following them around; showing up at their workplace unannounced; or demanding to know their whereabouts all the time. A second vignette scenario assessed respondents’ attitudes towards women who stalked men, and compared those to attitudes towards men who stalked women. Results for this analysis are included in Tables 50 and 51, respectively. Overall, the majority of respondents thought that neither men nor women should ever resort to the two forms of controlling behaviour hypothesised in the scenario questions. The results for the first scenario were very consistent (Table 50), with around 70 per cent of respondents on average across all socio-economic groups and sex answering that ‘he should never do that’. On average, around 20 per cent of the sample across all groups answered ‘I sympathise with him, but he should not do this’. There were interesting patterns among those who answered ‘Yes, he has the right to do this’. Fifteen per cent of the sample overall answered the question in this way, but men and individuals aged 15– 19 years, living in rural areas, in the Upper East and Upper West regions, with no education and in the lowest asset quintile were more likely than others to answer ‘yes’ to this question (Table 50). The reactions to the second scenario question were very similar (Table 51). 160 Table 50: Responses to first social violence scenario (in percentages) Yes, he has I sympathise with He should never Total the right to him, but he should do that do this not do that Sex Male 12.1 19.5 68.4 100.0 Female 10.9 19.7 69.4 100.0 Age group (years) 15–19 14.4 19.8 65.8 100.0 20–24 8.6 17.8 73.6 100.0 25–29 9.6 18.3 72.0 100.0 30–39 13.0 18.4 68.6 100.0 40–49 11.0 22.7 66.3 100.0 50–60 11.8 21.4 66.8 100.0 Employment Self-employed 11.5 19.3 69.1 100.0 Employed 11.7 20.6 67.7 100.0 Not working 10.7 19.7 69.6 100.0 Marital status Never married 9.4 22.0 68.6 100.0 Married or living together 11.5 19.2 69.3 100.0 Divorced/separated/widowed 11.9 20.2 67.8 100.0 Residence Urban 9.9 19.2 70.9 100.0 Rural 13.0 20.1 66.9 100.0 161 Yes, h e has I sympath ise with He shou ld never To tal the right to him, but he should do that do this not do that Region Western 7.0 25.6 67.5 100.0 Central 12.2 30.9 56.9 100.0 Greater Accra 8.6 13.3 78.1 100.0 Volta 4.5 32.3 63.1 100.0 Eastern 14.5 15.9 69.6 100.0 Ashanti 10.7 21.5 67.8 100.0 Brong Ahafo 11.7 17.7 70.6 100.0 Northern 7.5 10.7 81.9 100.0 Upper East 32.3 14.0 53.7 100.0 Upper West 49.0 8.5 42.5 100.0 Education level None 14.3 21.3 64.5 100.0 Primary 12.2 21.0 66.8 100.0 Middle/JSS/JHS 9.5 19.5 71.0 100.0 Secondary 12.9 16.5 70.6 100.0 Technical 5.3 12.6 82.1 100.0 Higher 10.4 21.5 68.1 100.0 Asset quintile Lowest 14.6 21.9 63.5 100.0 Second 12.0 18.4 69.6 100.0 Middle 8.0 22.6 69.4 100.0 Fourth 10.5 17.8 71.7 100.0 Highest 11.1 18.5 70.4 100.0 Total 11.4 19.6 69.0 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 162 Table 51: Responses to second social violence scenario (in percentages) Yes, she has I sympathise with She should never Total the right to her, but she should do that do this not do that Sex Male 9.7 18.9 71.4 100.0 Female 10.1 19.3 70.6 100.0 Age group (years) 15–19 11.6 19.5 68.8 100.0 20–24 12.3 19.7 67.9 100.0 25–29 11.2 16.9 72.0 100.0 30–39 10.1 17.2 72.6 100.0 40–49 6.8 22.6 70.6 100.0 50–60 8.2 20.9 71.0 100.0 Employment Self-employed 9.3 19.2 71.5 100.0 Employed 8.8 18.0 73.2 100.0 Not working 12.4 19.8 67.8 100.0 Marital status Never married 15.1 15.9 69.1 100.0 Married or living together 9.1 20.3 70.6 100.0 Divorced/separated/widowed 10.2 16.9 73.0 100.0 Residence Urban 9.7 18.9 71.4 100.0 Rural 10.2 19.4 70.4 100.0 163 Yes, she has I sympathise with S he should never Total the right to her, but she should do that do this not do that Region Western 9.4 21.9 68.7 100.0 Central 12.9 27.9 59.2 100.0 Greater Accra 9.5 18.5 72.1 100.0 Volta 1.1 29.3 69.6 100.0 Eastern 6.0 14.7 79.3 100.0 Ashanti 10.1 16.7 73.2 100.0 Brong Ahafo 11.1 14.9 74.0 100.0 Northern 12.6 14.9 72.5 100.0 Upper East 21.5 10.5 67.9 100.0 Upper West 28.4 20.9 50.7 100.0 Education level None 12.6 17.2 70.2 100.0 Primary 9.4 18.4 72.3 100.0 Middle/JSS/JHS 8.6 19.2 72.2 100.0 Secondary 11.7 20.6 67.8 100.0 Technical 7.8 19.2 73.0 100.0 Higher 8.4 22.1 69.6 100.0 Asset quintile Lowest 7.9 21.2 70.9 100.0 Second 11.4 17.8 70.8 100.0 Middle 12.4 20.1 67.5 100.0 Fourth 10.4 17.7 71.8 100.0 Highest 6.9 19.5 73.5 100.0 Total 9.9 19.1 70.9 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 164 As noted above, the qualitative research also included scenarios to understand the participants’ attitudes towards social violence. These scenarios were discussed in all 20 focus groups. The findings from this component of the research were nationally representative. They are discussed in detail below in relation to the quantitative findings, and specific quotes are used selectively to illustrate certain findings. Two complementary vignette scenarios on attitudes towards social violence were discussed in the focus group discussions, whereby participants were asked to discuss: (i) whether it was acceptable for men to prevent women from working (outside the home); and (ii) whether it was acceptable for men to prevent their daughters from attending school. The qualitative findings show that in most regions (with the exception of the three northern regions) respondents found that it was unacceptable for men to prevent their wives from working – particularly when the women contributed extra income to the household. Results for all regions also showed widespread disapproval of men preventing their daughters from attending school. Focus group discussions in the three northern regions indicated higher variation in the response to whether or not it was acceptable for men to prevent their partners from working. This finding is in line with a recently published study on gender inequality in the Northern region (Alhassan and Odame, 2015), where the authors outlined a set of household and contextual factors that worked against government efforts to promote gender equality. These factors included limited availability of household resources and poverty. Further socio-cultural practices, such as polygyny, fostering, child marriage and boy-child preference were also shown to collectively undermine women’s status relative to men, and to reinforce their economic dependence on men in shared households. In the focus group discussions in the Northern region, preventing women from working was viewed as acceptable because their work outside the house would be at the cost of their domestic chores. The following quote illustrates this view: “[F]rankly speaking, juggling motherhood and wifely duties with a profession is no mean joke. It’s really difficult. So if the man thinks his wife is suffering too much and tells her to stay at home and take care of the house whiles he works to support them, I don’t see any problems with that.” (FGD, Northern region, Sagnerigu district: female group, 18–50 years) In contrast, the following dialogue, drawn from a focus group discussion with married men (25–50 years old) in the Volta region, is an example of the prevailing response across all regions to the scenario asking whether it is acceptable for men to prevent their daughters from attending school: Interviewer: Is it acceptable for fathers to stop daughters from attending school? Respondent 7: It is not right for the father to deny his girl child education. Women are now doing what some men cannot do. There are women who are taking care of their families. Respondent 6: It is very wrong for the father to stop his girl child from attending school. Women also need education to survive just like men. Respondent 5: It is wrong to deny a girl child her right to education. Your daughter may marry an educated person. The problem comes when the husband marries another woman 165 and the second wife happens to be educated, the man will definitely prefer the educated woman to your daughter. Respondent 4: It is not good for a man to ask his daughter to stop schooling. In this modern era, women are equally doing what men can do. Respondent 10: It is not right for a father to deny his girl child her right to education. The reason is that capability is not based on gender. Combined, the qualitative and quantitative findings drawn from the vignette scenarios suggest that social violence was consistently viewed as unacceptable. However, social violence between partners may also be shaped by gender norms – prevalent in the northern regions – in which women were discouraged from engaging in paid labour. All respondents consistently argued that it was unacceptable to prevent girls from acquiring an education. 4.4.2. Attitudes towards domestic physical violence As noted above, an important finding from the qualitative and quantitative data relates to the variation in people’s understanding of what constitutes domestic violence. This section offers a detailed analysis of attitudes towards domestic physical violence. The research team again used scenario vignettes to elicit attitudes towards physical violence. The survey introduced the following scenario: ‘After a marital dispute the wife refused to talk to her husband, and refused to make his dinner and any other household chores. The husband hit his wife, but without leaving a mark.’ The responses to this scenario are provided in Table 52. Seventy-five per cent of male and female respondents found it completely unacceptable for a man to beat his wife because she refused to talk to him and do household chores after their dispute. In line with previous findings, this answer was most common among respondents who were employed (82.3 per cent), living in urban areas (80.2 per cent) and in the Greater Accra region (89.9 per cent), with higher education (83.8 per cent) and in the highest asset quintile (82.6 per cent). 166 Table 52: Responses to first physical violence scenario (in percentages) Completely Somewhat Somewhat Completely Do not Total unacceptable unacceptable acceptable acceptable know Sex Male 75.0 13.4 4.8 6.4 0.4 100.0 Female 75.1 10.6 4.9 9.2 0.2 100.0 Age group (years) 15–19 74.3 7.7 5.6 11.0 1.3 100.0 20–24 70.5 14.7 4.3 9.7 0.7 100.0 25–29 77.4 10.1 4.9 7.6 0.0 100.0 30–39 75.5 11.9 4.9 7.6 0.0 100.0 40–49 75.5 10.4 5.5 8.3 0.3 100.0 50–60 75.3 14.2 4.0 6.2 0.3 100.0 Employment Self-employed 72.1 13.1 5.1 9.5 0.1 100.0 Employed 82.3 7.2 4.0 6.0 0.5 100.0 Not working 77.4 11.2 4.8 5.9 0.6 100.0 Marital status Never married 77.4 10.7 2.9 7.8 1.1 100.0 Married or living together 73.9 11.9 5.4 8.7 0.1 100.0 Divorced/separated/widowed 78.1 11.5 4.0 6.3 0.2 100.0 Residence Urban 80.2 9.9 3.5 6.0 0.3 100.0 Rural 69.7 13.5 6.3 10.2 0.3 100.0 167 Completely Somewhat Somewhat Completely Do not Total unacceptable unacceptable acceptable acceptable know Region Western 86.6 6.2 3.3 4.0 0.0 100.0 Central 64.8 18.0 7.9 8.4 0.9 100.0 Greater Accra 89.9 5.5 3.1 1.2 0.3 100.0 Volta 69.9 22.8 2.6 4.0 0.7 100.0 Eastern 82.3 10.4 1.8 5.5 0.0 100.0 Ashanti 82.6 6.6 3.8 6.6 0.5 100.0 Brong Ahafo 71.7 12.1 2.6 13.6 0.0 100.0 Northern 44.5 22.2 13.9 19.4 0.0 100.0 Upper East 54.0 17.7 8.1 20.2 0.0 100.0 Upper West 32.0 19.9 16.7 31.5 0.0 100.0 Education level None 58.6 17.5 8.6 15.3 0.0 100.0 Primary 70.8 11.7 5.3 11.2 1.0 100.0 Middle/JSS/JHS 80.0 9.9 4.1 5.7 0.4 100.0 Secondary 81.9 9.1 3.6 5.4 0.0 100.0 Technical 80.7 13.6 1.5 4.2 0.0 100.0 Higher 83.8 10.2 2.5 3.4 0.0 100.0 Asset quintile Lowest 62.8 16.2 6.5 14.0 0.5 100.0 Second 73.3 13.0 6.4 7.2 0.0 100.0 Middle 76.6 10.9 3.7 8.5 0.3 100.0 Fourth 82.3 8.0 3.4 6.2 0.2 100.0 Highest 82.6 9.3 3.3 4.1 0.6 100.0 Total 75.1 11.7 4.9 8.1 0.3 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 168 Table 53: Responses to second physical violence scenario (in percentages) Yes, fully Yes, partly Not at all Do not know Total Sex Male 48.2 32.7 18.5 0.6 100.0 Female 50.9 30.3 18.5 0.3 100.0 Age group (years) 15–19 55.9 20.5 22.3 1.3 100.0 20–24 49.6 30.5 19.2 0.7 100.0 25–29 49.9 29.3 20.6 0.3 100.0 30–39 50.3 30.9 18.6 0.2 100.0 40–49 47.4 33.4 18.8 0.4 100.0 50–60 48.4 38.8 12.6 0.2 100.0 Employment Self-employed 52.4 30.6 16.8 0.3 100.0 Employed 40.9 36.4 21.8 0.9 100.0 Not working 49.7 29.3 20.5 0.4 100.0 Marital status Never married 47.2 30.4 21.4 1.1 100.0 Married or living together 50.3 30.8 18.5 0.4 100.0 Divorced/separated/widowed 49.5 33.1 17.4 0.0 100.0 Residence Urban 43.8 36.7 19.0 0.4 100.0 Rural 56.1 25.5 18.0 0.4 100.0 169 Yes, f ully Yes, p artly Not a t all Do not know To tal Region Western 57.6 22.6 19.8 0.0 100.0 Central 50.1 40.1 8.9 0.8 100.0 Greater Accra 25.9 52.6 21.2 0.3 100.0 Volta 39.6 36.4 23.2 0.8 100.0 Eastern 51.1 30.4 18.1 0.4 100.0 Ashanti 56.2 24.0 19.3 0.5 100.0 Brong Ahafo 56.5 21.5 21.5 0.5 100.0 Northern 64.4 22.0 13.7 0.0 100.0 Upper East 62.7 23.3 13.3 0.7 100.0 Upper West 67.1 14.1 18.7 0.0 100.0 Education level None 59.0 27.3 13.7 0.0 100.0 Primary 54.4 28.2 16.5 0.9 100.0 Middle/JSS/JHS 49.8 31.4 18.1 0.6 100.0 Secondary 43.9 30.6 25.5 0.0 100.0 Technical 34.8 47.5 17.7 0.0 100.0 Higher 38.0 39.3 22.7 0.0 100.0 Asset quintile Lowest 52.6 28.7 18.1 0.6 100.0 Second 58.2 25.6 16.1 0.2 100.0 Middle 51.4 26.9 21.4 0.4 100.0 Fourth 43.4 36.3 19.9 0.4 100.0 Highest 39.5 41.5 18.4 0.6 100.0 Total 49.8 31.2 18.5 0.4 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 170 The qualitative data from all regions similarly found that participants considered wife- beating to be unacceptable. Wife-beating linked to women’s refusal to have sex was less frequently mentioned in the focus group discussions, but when it was mentioned, both female and male respondents iterated a consistent attitude, condemning men’s treatment of women ‘as slaves’. In the excerpt below, for example, married men (aged 18–50) in the Volta region explained why they believed that physical violence against women was unacceptable. Their views were representative of the region’s findings (these issues were not picked up in other focus group discussions), and they reflect the national-level data from the interviews and other regional focus groups, condemning men as ‘weak’ for beating women: Respondent 2: No, it’s not acceptable because she’s his wife and deserves to be treated with respect. She’s not a slave. Respondent 7: No, it’s not acceptable. This is because the bible entrusts us to treat each other with respect and love. After marriage, the two individuals become one, and so the man ought to treat her as he would like to be treated. Respondent 10: No, I also think it’s not acceptable because conflicts and disagreements are integral parts of every relationship and so we can’t fall on beating to drive home our points. Beating usually doesn’t solve anything so we must employ other means to get our wives to agree with us. Analysis of the qualitative data indicated that views about physical violence differed depending on the age and relationship of the two (or more) people involved in the violent incident or series of interactions. The data (in all regions) revealed in particular that physical violence by parents against children was appropriate, and sometimes even necessary. This statement, by a participant in a focus group discussion with married men (aged 18–50) in the Brong Ahafo region, indicates this distinction about the acceptability of violence against children by adults, on the one hand, and between men and women, on the other hand: Respondent 4: No, there is no justification whatsoever to beat your wife, because it is not a father and child relationship, but you’re coming together as friends, and so whatever the case may be, both of them should be able to iron their differences peacefully. In research on attitudes to violence, predominantly in horizontal relationships (between partners), scholars have found that the perceived culpability of the victim (i.e. whether they are perceived to be at fault) plays a significant role in men and women’s perceptions of the violence (Garcia-Moreno et al., 2005). Seeking to explore this factor in relation to physical violence, the survey asked whether the respondents thought it was the woman’s fault when she was beaten by her husband – in relation to the scenario discussed above. The results of this analysis were striking and are presented in Table 53. Around half of the sample on average across all groups answered that they fully agreed that the beating was the woman’s fault – even though the majority of the respondents found the husband’s behaviour unacceptable. This answer was more common among women: 50.9 per cent of all women agreed that the woman in the scenario was at fault, compared to 48.2 per cent of male respondents. The answer was also more common among self-employed 171 individuals (52.4 per cent), married (50.3 per cent), living in rural areas (56.1 per cent), with no education (59.0 per cent) and in the lowest two asset quintiles (52.6 and 58.2 per cent, respectively) (Table 53). The survey also asked respondents under what circumstances wife-beating would be acceptable. The five most accepted reasons for wife-beating are shown in Table 54, while Table 55 expands this analysis across different socio-economic groups. The main reasons why men and women justified wife-beating included disobedience and neglect of children: 11.2 per cent of male respondents and 17.2 per cent of female respondents mentioned that wife-beating was acceptable if the wife had disobeyed the husband, and 10.2 per cent of male respondents and 16.5 per cent of female respondents agreed that wife-beating was acceptable if the wife neglected the children (Table 54). If the woman refused sex, a form of violence largely condemned in all focus group discussions where this issue was raised, 5.7 per cent of men and 9.8 per cent of women thought this was an acceptable reason for wife-beating (Table 55, Column 3). Table 54: Percentage of respondents agreeing with wife-beating if the wife… Male Female respondents respondents …disobeys her husband 11.2 17.2 …neglects the children 10.2 16.5 …goes out without telling her husband 7.3 13.2 …contracts a sexually transmitted disease 7.8 13.2 …argues with her husband 7.0 11.8 Source: Ghana Family Life and Health Survey (GFLHS) 2015. As noted above, women were consistently more likely to find wife-beating acceptable than men (Tables 54 and 55). This finding echoes a similar result discussed in Takyi and Mann (2009) and may imply that victims internalise social norms once they are in abusive relations, as discussed in Section 2. In line with findings discussed above, Table 53 shows that condoning wife-beating was also more common, on average, among younger individuals, either self-employed or not working, married, living in rural areas, with no education and in the bottom income quintiles. There were also large regional differences: respondents from the three northern regions were more likely to reveal attitudes supporting wife-beating than other regions. 172 Table 55: Percentage of respondents who think that a man has a good reason to hit his wife if: 1 2 3 4 5 6 7 8 9 10 11 Sex Male 4.6 11.2 5.7 2.0 5.6 0.4 7.8 7.3 10.2 7.0 2.4 Female 7.7 17.2 9.8 3.2 8.7 0.5 13.2 13.2 16.5 11.8 4.3 Age group 15–19 10. 21.7 11.7 2.8 9.6 1.6 15.5 17.0 18.4 13.7 4.9 20–24 5.69 15.3 7.8 4.0 9.4 0.4 12.1 10.9 16.5 9.4 5.1 25–29 5.5 14.5 7.7 2.3 6.0 0.3 10.1 10.0 12.9 10.3 2.8 30–39 5.5 12.9 6.4 2.2 6.3 0.3 10.2 9.2 12.7 8.5 3.2 40–49 6.4 15.3 10.0 3.0 9.1 0.3 11.9 12.0 14.9 10.6 3.8 50–60 7.6 13.7 8.4 2.8 6.8 0.5 9.5 10.1 12.0 9.5 2.8 Employment Self-employed 7.0 16.4 9.2 3.3 8.2 0.4 12.5 11.6 15.5 11.1 3.9 Employed 3.2 6.8 4.6 0.9 4.2 0.4 5.7 6.0 8.7 4.9 2.0 Not working 7.4 16.5 8.1 2.6 8.0 0.7 11.2 12.3 14.0 10.2 3.8 Marital status Never married 5.9 12.8 6.6 1.4 6.1 0.9 9.6 8.6 10.0 7.2 2.6 Married or living together 6.6 15.5 8.6 3.1 7.9 0.3 11.4 11.2 14.3 10.5 3.7 Divorced/separated/widowed 6.0 13.2 7.6 2.1 6.7 0.6 10.6 10.5 15.0 8.9 3.6 Residence Urban 3.4 9.3 4.8 1.3 4.7 0.4 7.2 6.9 9.9 6.0 1.6 Rural 9.6 20.5 11.7 4.2 10.5 0.5 15.1 15.0 18.3 13.9 5.6 173 1 2 3 4 5 6 7 8 9 10 11 Region Western 4.4 14.1 8.2 0.9 10.0 0.2 8.9 13.0 13.2 8.3 4.0 Central 5.9 16.6 8.1 0.5 4.7 0.6 13.3 9.4 10.5 4.9 1.3 Greater Accra 1.2 3.1 1.7 1.0 2.6 0.1 3.0 2.6 3.9 1.7 0.8 Volta 4.9 7.3 5.2 3.3 4.7 1.6 9.9 7.2 9.5 8.0 4.5 Eastern 5.3 11.8 7.0 1.6 3.6 0.3 15.4 7.7 18.2 8.6 2.4 Ashanti 5.7 12.4 4.8 1.0 6.5 0.3 6.1 8.5 9.5 7.3 2.3 Brong Ahafo 3.6 16.1 7.3 2.7 7.7 0.5 10.1 8.6 12.4 12.8 2.6 Northern 14.8 42.7 24.4 11.3 15.0 0.6 21.3 28.5 38.6 29.1 6.7 Upper East 25.8 32.4 24.0 10.8 26.4 0.4 37.4 36.9 36.4 27.0 17.1 Upper West 31.5 38.6 29.4 8.9 27.8 0.4 34.9 30.2 37.4 29.3 16.7 Education level None 12.6 28.6 17.5 7.0 13.5 0.4 19.9 21.9 25.5 20.9 7.6 Primary 8.1 17.7 9.9 3.8 9.4 0.9 14.6 12.3 17.0 12.4 4.6 Middle/JSS/JHS 5.5 12.6 6.3 1.7 6.4 0.4 9.5 8.9 12.2 7.9 3.1 Secondary 3.1 8.2 3.5 0.9 4.9 0.3 6.2 6.7 8.5 4.8 1.2 Technical 2.0 4.3 5.1 0.6 4.3 0.1 5.2 2.6 3.9 1.3 0.7 Higher 1.6 4.6 2.5 0.3 1.8 0.1 3.4 2.9 5.1 2.5 0.5 Asset quintile Lowest 10.6 22.5 13.4 4.7 11.2 0.9 18.5 17.7 20.5 15.5 6.0 Second 8.2 18.7 11.0 4.2 8.9 0.4 13.3 13.4 17.3 12.6 5.3 Middle 5.0 13.8 6.8 1.4 7.8 0.2 8.8 10.1 12.5 9.6 2.5 Fourth 4.4 10.1 4.6 1.7 5.6 0.3 7.8 6.7 10.6 5.8 2.1 Highest 2.4 5.9 3.1 0.7 2.5 0.4 4.6 4.1 6.4 4.0 0.7 Total 6.4 14.8 8.2 2.7 7.5 0.4 11.1 10.8 14.0 9.9 3.6 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Note: 1 = She does not complete her household work to his satisfaction; 2 = She disobeys him; 3 = She refuses to have sex with him; 4 = She asks him whether he has other girlfriends; 5 = He suspects that she is unfaithful; 6 = He finds out that she has been unfaithful; 7 = She contracts a sexually transmitted disease; 8 = She goes out without telling him; 9 = She neglects the children; 10 = She argues with him; 11 = She burns the food. 174 Fewer women in 2015 found wife-beating acceptable than in the 2008 DHS. Table 56 includes comparable questions asked in the GFLHS 2015 and in the Ghana DHS 2008. The results show a reduction of 13.5 percentage points in the number of women justifying wife- beating (from 36.6 per cent in 2008 to 23.1 per cent in 2015), and 8 percentage points in the number of men condoning wife-beating for at least one of the reasons provided (from 21.8 per cent in 2008 to 13.8 per cent in 2015). Even though this attitude is still more prevalent among women, it also seems to be changing more rapidly among women (the reduction is larger among women). Table 56: Percentage of respondents stating that wife-beating is acceptable if the woman… …refuses … goes out … neglects … argues … burns At least to have sex without children with the food one asking husband reason 2015 GFLHS Men 5.7 7.3 10.2 7.0 2.4 13.8 Women 9.8 13.2 16.5 11.8 4.3 23.1 2008 DHS Men 6.6 10.7 13.7 11.0 4.6 21.8 Women 12.3 22.3 26.3 21.3 8.3 36.6 Source: Ghana Family Life and Health Survey (GFLHS) 2015 and DHS 2008. While the GFLHS 2015 focused predominantly on wife-beating, the qualitative focus group discussions were able to explore in more detail men’s experiences of physical violence by women – even though few numbers of respondents considered this a form of domestic violence. This finding was regionally representative, with respondents in the Central, Eastern, Volta and Greater Accra regions iterating the importance of ‘fairness’: just as women should not be beaten, so too should women not beat men. However, in the Northern and Upper East regions, the focus group discussions about ‘husband-beating’ – particularly among young men and women (aged 18–25) and among married women (aged 18–50) – elicited a great deal of laughter, and sometimes even disbelief. These responses to some of the group members’ accounts of men experiencing physical violence were overlaid with gendered assumptions that these men were ‘weak’ and not able to maintain their superiority over women. For example, the excerpt below from the Gbolo district in the Northern region offers a window into one such discussion, where the possibility of a man being beaten by his wife was met with disbelief: Respondent 2: If it is the woman beating the man, I would not separate them. I will pass by because why should a man allow himself to be beaten by a woman? Respondent 7: I will go and separate them, but if I could, I will support the man to sack the woman/wife because it is disrespectful. Why should you beat up your husband? Respondent 6: I will separate them. 175 Respondent 4: I will not mind them. I will pass by because why a woman should beat her husband is a taboo in our community. (FGD, Northern region, Gbolo district: male and female group, 18–25 years) However, sometimes the laughter had a serious undertone, revealing discontent with gender norms among the respondents, particularly in the Northern and Upper East regions. In some groups, the question revealed a desire for revenge by women who had themselves experienced multiple forms of domestic violence. In this case, this finding suggests that violence may be deemed ‘acceptable’ when it constitutes, or is viewed to constitute, a form of retribution. This sentiment, while widely reported mostly in the Northern and Upper East regions, is powerfully captured in this discussion among married women (aged 18–50). Respondent 1: I’ll stand by and watch as the woman beats him, I’ll neither call for help nor try to separate them because he’s beaten his wife and so he should also have a taste of what it takes to be beaten. Respondent 2: If I meet such a scenario, I’ll simply let the woman beat the man because he’s a wife-beater and so he must be taught a lesson in respecting women, so I’ll cheer the woman on. Respondent 7: As for me, I’ll help the lady to give the man a good beating. I’ll even referee the fight. (FGD, Northern region, Sagnerigu district: female group, 18–50 years) This dynamic, however, did not apply to men’s views of the acceptability of violence against women. The qualitative and quantitative data have also allowed the research team to examine attitudes towards physical violence against children. This form of violence was predominantly considered unacceptable. However, some study participants expressed certain contexts when physical violence against children was acceptable or even desirable. The qualitative data revealed how, across all regions, parents often struggled to discipline their children when they misbehaved. Focus group discussions in all ten regions revealed that both older and younger respondents of both sexes agreed that there were certain circumstances when physical violence against children was acceptable. This finding suggests that not only do gender norms shape the perceived acceptability of certain forms of violence between partners, but that age is an important feature for distinguishing what kinds of physical violence are and are not appropriate. This finding is poignantly illustrated in the following excerpt from a focus group discussion in the Northern region, which represents views expressed across the whole country: Respondent 2: Yes, it is acceptable somehow because if they do what is not right, the father has to talk to them and if they persist, then the father can beat them. Respondent 4: Yes, it is acceptable somehow because the man has to talk to the children for the third time and if they persist then the man can beat them. Respondent 7: Yes, but it is very necessary for the men to talk to the children for the first time and if they prove stubborn, then they have to beat them. 176 (FGD, Northern region, Gbolo district: male group, aged 50+) To explore the circumstances under which physical violence against children was deemed acceptable, the GFLHS 2015 included a scenario question asking respondents whether they would intervene if they witnessed an act of physical violence against a child. Ninety per cent of respondents said they would intervene through a variety of means, including their own direct intervention or by approaching religious authorities, community leaders, a relative of the child, a neighbour or other community members. 53 per cent of men and 50.8 per cent of women in the GFLHS 2015 mentioned that they would intervene directly by going to check what happened, trying to rescue the child themselves or trying to stop the dispute. Physical punishments of children seemed to be at least partially acceptable if they intended to discipline the child when non-physical attempts were ineffective. The GFLHS 2015 asked respondents whether they fully agreed, somewhat agreed, somewhat disagreed or fully disagreed with the following statement: ‘To bring up, raise or educate a child properly, the child sometimes needs to be physically punished.’ 50 per cent of women and 43.3 per cent of men fully agreed with this statement (Table 57). Younger respondents were more likely to fully disagree with the physical punishment of children (over 24 per cent among those aged 15–24 years, compared to 19.9 per cent among respondents aged 50–60 years), as were employed respondents, living in urban areas and in the Western and Ashanti regions, and those with higher levels of education and assets (Table 57). There has been, however, a significant change in attitudes towards physical violence against children: in 1998, 8 per cent of respondents thought it was never acceptable to beat a child (Coker-Appiah and Cusack, 1999). In 2015, this number has risen to 25.3 per cent among men and 18.8 per cent among women (Table 57). 177 Table 57: Views about statement ‘To bring up, raise or educate a child properly, the child sometimes needs to be physically punished’ (in percentages) Fully Somewhat Somewhat Fully Do not Total agree agree disagree disagree know Sex Male 43.3 18.9 12.3 25.3 0.1 100.0 Female 50.6 18.7 11.9 18.8 0.0 100.0 Age group (years) 15–19 48.1 15.2 11.8 24.8 0.0 100.0 20–24 45.7 17.6 12.3 24.3 0.0 100.0 25–29 47.6 19.7 10.8 22.0 0.0 100.0 30–39 48.1 18.3 12.4 21.1 0.1 100.0 40–49 49.6 18.9 12.7 18.7 0.1 100.0 50–60 46.4 21.4 12.2 19.9 0.0 100.0 Employment Self-employed 49.2 19.1 12.7 19.1 0.0 100.0 Employed 44.2 17.8 12.2 25.5 0.3 100.0 Not working 46.4 18.8 10.4 24.4 0.0 100.0 Marital status Never married 50.8 15.6 11.6 22.1 0.0 100.0 Married or living together 47.0 19.2 12.1 21.7 0.0 100.0 Divorced/separated/widowed 48.5 19.0 12.3 20.2 0.0 100.0 Residence Urban 45.8 19.5 11.4 23.2 0.0 100.0 Rural 49.7 18.0 12.8 19.5 0.1 100.0 Region Western 50.6 10.9 8.5 30.0 0.0 100.0 Central 46.5 17.7 22.1 13.7 0.0 100.0 Greater Accra 49.2 20.8 10.3 19.8 0.0 100.0 Volta 40.7 32.6 13.4 13.3 0.0 100.0 Eastern 67.2 18.2 6.7 7.9 0.0 100.0 Ashanti 45.4 8.8 8.2 37.3 0.2 100.0 Brong Ahafo 42.1 23.0 13.3 21.6 0.0 100.0 Northern 31.2 28.0 24.7 16.2 0.0 100.0 Upper East 54.2 18.7 10.1 17.1 0.0 100.0 Upper West 63.8 21.8 6.8 7.6 0.0 100.0 Education level None 45.2 23.3 15.3 16.2 0.0 100.0 Primary 53.7 18.0 11.5 16.8 0.0 100.0 Middle/JSS/JHS 49.2 17.0 12.0 21.6 0.1 100.0 Secondary 44.8 17.3 9.5 28.5 0.0 100.0 Technical 38.6 25.5 12.1 23.8 0.0 100.0 Higher 45.0 18.4 11.0 25.7 0.0 100.0 Asset quintile Lowest 45.3 21.9 14.1 18.8 0.0 100.0 Second 50.3 18.1 13.4 18.3 0.0 100.0 Middle 48.6 16.7 12.0 22.7 0.0 100.0 Fourth 47.4 18.3 10.7 23.4 0.1 100.0 Highest 45.8 19.0 9.2 25.9 0.1 100.0 Total 47.7 18.8 12.1 21.4 0.0 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 178 4.4.3. Attitudes towards domestic sexual violence The GFLHS 2015 revealed nuanced attitudes of different individuals and social groups towards domestic sexual violence, with many exhibiting a higher degree of acceptance of some dimensions of sexual violence than other forms of violence. The research team explored attitudes towards domestic sexual violence by asking respondents their views in relation to two statements: (1) ‘If a woman wears revealing clothes, it is okay for men to make some sexual comments as she walks in her neighbourhood’; and (2) ‘If a woman wears revealing clothes, it is her fault if she is raped.’ Tables 58 and 59 show, respectively, the reactions of the respondents to these two statements. Findings show that a very substantial part of the sample – 36.7 per cent of men and 40.7 per cent of women – fully agreed with the first statement (Table 58), without much variation in responses across socio-economic groups. Individuals with higher levels of education, in the top asset quintile and living in the Eastern and Ashanti regions and in urban areas were, however, more likely to fully disagree with the first statement (Table 58, Column 4), even though ‘fully agree’ was the most common answer across most socio- economic groups. More than three in every five respondents (61.6 per cent) fully agreed that the woman was to be blamed for rape if she wore revealing clothes. As with other forms of violence, a higher proportion of women than men displayed attitudes condoning this: 65.3 per cent of women interviewed agreed with the second statement, compared to 56.2 per cent of men. Respondents from rural areas were more likely than those in urban areas to fully agree with the second statement (65.8 and 57.7 per cent, respectively). The data show no other substantial differences across other group divisions. Employed individuals with higher levels of education, in the top asset quintile and living in urban areas were more likely to fully disagree with the second statement (but ‘fully agree’ was the most common answer among most socio-economic categories). 179 Table 58: Views about statement ‘If a woman wears revealing clothes, it is okay for men to make some sexual comments as she walks in her neighbourhood’ (in percentages) Fully Somewhat Somewhat Fully Do not Total agree agree disagree disagree know Sex Male 36.7 17.2 16.6 28.9 0.6 100.0 Female 40.7 16.9 14.4 27.4 0.6 100.0 Age group (years) 15–19 40.4 15.0 13.6 28.6 2.4 100.0 20–24 40.6 14.8 13.7 30.8 0.1 100.0 25–29 38.9 17.4 14.9 28.5 0.3 100.0 30–39 38.2 17.5 16.3 27.5 0.5 100.0 40–49 40.8 17.4 15.1 26.0 0.6 100.0 50–60 36.9 18.3 16.4 27.7 0.7 100.0 Employment Self-employed 39.5 17.9 15.9 26.2 0.5 100.0 Employed 38.3 15.7 14.2 31.4 0.5 100.0 Not working 38.5 15.9 14.5 30.1 0.9 100.0 Marital status Never married 36.1 17.5 14.4 29.9 2.1 100.0 Married or living together 38.9 17.0 16.1 27.6 0.4 100.0 Divorced/separated/widowed 41.2 17.1 13.0 28.4 0.3 100.0 Residence Urban 38.2 17.0 15.0 29.1 0.7 100.0 Rural 40.0 17.1 15.6 26.8 0.5 100.0 Region Western 52.4 9.7 5.3 32.3 0.2 100.0 Central 49.4 16.2 12.8 21.6 0.0 100.0 Greater Accra 38.8 17.2 14.7 28.2 1.1 100.0 Volta 28.6 23.8 35.5 10.2 1.9 100.0 Eastern 27.8 18.1 18.3 35.1 0.7 100.0 Ashanti 41.4 12.6 10.6 35.1 0.4 100.0 Brong Ahafo 41.4 16.3 11.1 31.0 0.2 100.0 Northern 49.5 24.8 26.4 28.9 0.4 100.0 Upper East 47.9 27.6 11.4 13.1 0.0 100.0 Upper West 60.6 16.4 5.0 18.0 0.0 100.0 Education level None 36.7 22.0 17.8 23.1 0.4 100.0 Primary 42.2 18.7 14.2 23.2 1.8 100.0 Middle/JSS/JHS 42.0 15.2 14.7 27.7 0.4 100.0 Secondary 37.4 15.7 14.4 32.1 0.4 100.0 Technical 37.8 14.8 23.5 23.1 0.8 100.0 Higher 31.0 14.8 13.1 40.7 0.4 100.0 Asset quintile Lowest 35.1 21.6 17.3 25.0 1.1 100.0 Second 40.4 16.2 15.2 28.1 0.2 100.0 Middle 42.7 16.9 14.7 25.4 0.4 100.0 Fourth 39.1 16.0 14.5 29.6 0.8 100.0 Highest 38.1 14.3 14.6 32.3 0.6 100.0 Total 39.1 17.0 15.3 28.0 0.6 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 180 Table 59: Views about statement ‘If a woman wears revealing clothes, it is her fault if she is raped’ (in percentages) Fully Somewhat Somewhat Fully Do not Total agree agree disagree disagree know Sex Male 56.2 11.0 8.8 23.7 0.4 100.0 Female 65.3 9.5 6.7 18.2 0.3 100.0 Age group (years) 15–19 63.5 9.8 8.5 17.6 0.7 100.0 20–24 64.0 9.2 5.8 20.5 0.5 100.0 25–29 62.9 11.0 6.9 18.5 0.7 100.0 30–39 61.7 11.3 7.8 18.9 0.2 100.0 40–49 59.8 9.2 9.0 21.9 0.1 100.0 50–60 58.8 8.6 7.5 24.9 0.1 100.0 Employment Self-employed 63.3 9.6 8.3 18.6 0.2 100.0 Employed 54.8 11.7 6.7 26.3 0.6 100.0 Not working 62.3 10.1 6.4 20.7 0.5 100.0 Marital status Never married 57.2 13.2 9.6 19.2 0.8 100.0 Married or living together 62.2 9.8 7.5 20.3 0.2 100.0 Divorced/separated/widowed 62.1 9.6 6.8 21.0 0.4 100.0 Residence Urban 57.7 10.1 8.8 23.0 0.5 100.0 Rural 65.8 10.0 6.3 17.7 0.2 100.0 Region Western 78.6 5.8 1.9 13.5 0.2 100.0 Central 71.3 7.0 6.9 14.8 0.0 100.0 Greater Accra 55.8 10.1 10.7 22.8 0.6 100.0 Volta 45.0 12.1 15.5 26.7 0.7 100.0 Eastern 62.8 10.7 9.3 16.9 0.3 100.0 Ashanti 62.7 10.1 3.7 23.1 0.4 100.0 Brong Ahafo 66.7 8.9 4.9 19.5 0.0 100.0 Northern 48.7 12.9 11.0 27.1 0.3 100.0 Upper East 67.7 15.8 4.1 12.3 0.0 100.0 Upper West 64.1 16.6 7.2 12.1 0.0 100.0 Education level None 59.5 12.4 8.8 18.9 0.4 100.0 Primary 68.9 9.4 6.1 15.5 0.2 100.0 Middle/JSS/JHS 65.4 8.3 8.1 17.7 0.4 100.0 Secondary 59.6 10.3 6.1 23.7 0.3 100.0 Technical 58.5 7.3 7.7 26.5 0.0 100.0 Higher 43.8 14.0 7.5 34.4 0.3 100.0 Asset quintile Lowest 59.4 11.7 7.9 20.8 0.2 100.0 Second 64.2 10.1 7.2 18.2 0.4 100.0 Middle 65.6 7.9 7.6 18.4 0.5 100.0 Fourth 61.2 10.0 7.3 21.3 0.3 100.0 Highest 56.7 10.4 8.1 24.4 0.4 100.0 Total 61.6 10.1 7.6 20.4 0.3 100.0 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 181 There was limited discussion about the incidence of sexual violence in the focus group discussions and individual interviews. Qualitative data on attitudes towards sexual violence revealed (in the groups that discussed this form of violence), however, that this form of violence was unambiguously perceived as unacceptable, across all regions. This view is illustrated, for example, in the following focus group discussion with married men (18–50 years) in the Brong Ahafo region, in the Ahafo Ano North municipal district: Respondent 3: No, it’s not acceptable. Sex is based on understanding and affection and is only enjoyable if both partners willingly participate. The woman may not be in the mood so if she’s not in the mood, how do you enjoy it? Besides how do you beat your partner and have sex with her the next moment? It’s really cruel. Respondent 4: No it’s not acceptable. Marriage is based on love and not sex; sex is just one aspect of marriage so it shouldn’t dictate how you treat your partner. The man should be able to understand and empathise with the woman in those times that she’s unable for some reason to give him sexual gratification. There would definitely be other opportunities. The previous section discussed how patterns of domestic sexual violence differed from other types of domestic violence. In particular, the results revealed that experiences of sexual violence were more common among employed, wealthier and more educated individuals living in urban areas, particularly among men. The research team postulated that this finding may have emerged because these groups are also more likely to have reported experiences of sexual violence. The findings in this section go some way towards supporting this. First, there was very limited discussion about sexual violence in the qualitative data. Second, most survey respondents showed high levels of acceptance of behaviours defined by sexual violence. Taken together, these findings suggest that sexual violence may be understood largely as something that takes place outside domestic relations. Perceptions of what may constitute sexual violence within domestic relations seem very limited – or hidden. This latter hypothesis is partially substantiated by the additional finding that larger percentages of female respondents accept forms of sexual violence, suggesting that women may internalise gender norms and norms about the acceptance of certain types of behaviour, as discussed in Section 2 and found in other studies. 4.4.4. Attitudes towards domestic psychological violence As discussed in the sections above on attitudes towards domestic social and physical violence, the qualitative and quantitative data found that there were circumstances where certain forms of domestic psychological violence were viewed as defence mechanisms and as a way of women claiming some limited power in highly unequal domestic partnerships. This finding was nationally representative in the qualitative data. The following quote, from a participant in a focus group discussion among young men and women aged 18–25 years in the Eastern region, captures this finding succinctly: “Most of the fights are physical but some are verbal. Most of the verbal fights are from the women and they use them as a defensive mechanism.” 182 Echoing the findings from the qualitative data, data from the GFLHS 2015 revealed that 31 per cent of acts of insulting, belittling or humiliation of male heads of households (in private or in front of other people) were perpetrated by their female spouse, partner or former partner. The remainder were perpetrated by other relations. The incidence of psychological abuse of the same type against female household heads was lower, around 25 per cent, according to the GFLHS 2015. The qualitative data provided important explanations for these levels of incidence. Notably, focus group participants often perceived the exercise of psychological pressure in the form of verbal abuse as a means of defence and a coping strategy (for women, in particular) to ward off physical violence, or to claim some autonomy or agency in a context where women may not have the ability to exert much power beyond the use of words (insults), their bodies (withholding sex) or their housework (not cooking). The qualitative data also offered some striking (but not nationally representative) findings that reveal how men also use psychological violence as a form of retribution. For example, in a mixed focus group discussion with youth in Kukurantumi (FDG, Eastern region, East Akim district: male and female group, 18–25 years), a respondent illustrated this finding as follows: “Let us assume that my wife has offended me but I don’t want to beat her. I can abuse her by not talking to her for a couple of months.” In a focus group discussion among young men and women (aged 18–25) in the Ashanti region, in Obuasi district, a respondent outlined an additional factor that may lead to abuse of women by men as a form of retaliation: “Women who are rich or self-made. Who doesn’t respect or appreciate their husbands, especially if the man is not rich like her, a case in point is when the woman doesn’t appreciate the man’s gifts, the man feels unappreciated resulting in drunkenness, psychological abuse, etc.” To explore attitudes towards psychological violence further, the GFLHS 2015 included the following scenario: ‘A woman repeatedly makes demeaning comments to her husband about him not making enough money, and not being able to provide for a nice home and gifts for her. She constantly compares him to his brother, who is a successful businessman.’ Respondents were asked whether this was a scenario that they were familiar with. Nearly half (48.8 per cent) of respondents stated that this situation was ‘very common’, and 22.9 per cent that it was ‘somewhat common’. The survey then repeated this scenario with the variation that the woman was beaten by her husband following the remarks and reported the incident to the police. Just over half of the respondents (55.8 per cent) agreed that she should have reported the abuse to the police. Of a total of 1,098 respondents (412 men and 686 women) who thought the woman should not have reported the incident to the police, 25.2 per cent of men and 32.1 per cent of women mentioned that she should not have reported the beating to the police because the woman was “the one to blame” for the incident. This finding is in line with the discussion in the previous sections, where attitudes condoning domestic violence seem more prevalent among women. 183 4.4.5. Attitudes towards domestic economic violence Throughout the qualitative data, respondents across all regions expressed strong attitudes against domestic economic violence. However, the qualitative data drew out interesting paradoxes. While most forms of domestic violence against women were generally seen as unacceptable, the data have also shown that men – but not women – were seen to be the main providers of material well-being to the household and its members. This argument was forcefully made, in particular, by married women (aged 18–50) in the focus group discussions that were conducted across all ten regions of the country. The flipside of the argument, which was made less often by married men, was that the pressure to secure a job and be the sole breadwinner was also a form of economic violence exerted by women who had not been restricted from earning an income but who had chosen not to. The following excerpt illustrates some of the different attitudes women expressed about economic violence. In this particular conversation, the women articulated one major form of economic violence that was expressed across the qualitative data – namely, men’s refusal to give women chop money for household costs: Facilitator: Is it acceptable if a man refuses to give chop money to his wife? Respondent 8: No. From the Bible, God said husband should take care of their wives for that reason any man who marries and doesn’t take care of the woman is not doing right. Respondent 5: It is not the right thing to do. You have to give her chop money. Facilitator: Can you explain your answer? Respondent 2: Because you the man brought the woman to your house so every morning, you must give her chop money. If you refuse to give me chop money, there will be a fight in the house. Respondent 3: No. Because he brought me to live with him, so he must take care of me well. The following extract from a focus group discussion with married women (18–50 years) in the Brong Ahafo region, in the Ahafo Ano North municipal district, similarly illustrates how gender norms – and expectations that men should be the ‘breadwinner’ – affect women’s perception of economic violence. The views expressed by the respondents below were common across most of the sample and across all regions: Respondent 3: No, it is not acceptable because it is the man’s responsibility to take care of the whole family. Since the man cannot starve himself throughout the whole day, he should not also attempt to starve the rest of the family members. The man cannot say that I cannot refuse giving chop money to the family. In some instance where the man is not financially sound, he can tell the woman politely to support him if she has but not to intentionally refuse to give the wife chop money. It is only a wicked person who will refuse to give the woman chop money. Respondent 4: It is a big no, because the man is the breadwinner of the family and hence his responsibility to take care of the family. If for one reason or the other the man does not have money, he can call the woman to support for that period but not to intentionally refuse to give her chop money. Considering the way our economy is now, the women are also liable to support but the responsibility lies on the man to take care of the household. Respondent 5: It is not acceptable. There are times that the man may genuinely not have money due to loss of job. Within that period the man can ask the woman to support the house. But, if the man has money, it not acceptable to refuse to give the woman chop money. 184 Respondent 8: No, it is not acceptable because considering the main responsibilities of a man at home, providing food is one of their core responsibilities. So if by God’s grace you have money I don’t see the reason why you should refuse to give chop money. What will happen then is that, for days that you will refuse to give chop money we will not also leave food for you to come and eat. Men felt the strain of these expectations. This excerpt from an interview with an activist in the Ashanti region describes this view, which was expressed across discussions with men: “Men tend to experience violence if they are not playing their role well. I am talking about roles because in Africa, we expect the man to provide. At least make the petty provision, not the essential. At least to maintain the family. So if you are not maintaining the family, definitely, the lady turns hostile. When the lady becomes hostile, it rolls unto you. The lady will not allow you to make love with her, and then she will try to fight you, especially if you are not strong. And then that is where the men turn to suffer. At times some of the men come here to complain about the women slapping them, but because of the scriptures and the fact that they are God fearing, they cannot do anything.” The denial of chop money also came out very strongly in the quantitative survey. It was the single most important act of economic violence mentioned against women and men (Tables 23–26): 11.6 per cent of women and 5.7 per cent of men reported having been denied chop money in the 12 months prior to the survey, despite believing there was enough money available for those expenses. 4.5. Determinants of domestic violence in Ghana This section discusses a number of factors the study found to be associated with different forms of violence against women, men, girls and boys, including where people live, and the social and economic characteristics of individuals, households and communities. In line with the sections above, the analysis combined insights from both quantitative and qualitative data. The sections also include information from multivariate regression analysis, as described in the methodology section, which has attempted to identify the main determinants of domestic violence in Ghana in 2015. To conduct the regression analysis, the research team created a variable for each category of domestic violence (social, physical, sexual, psychological and economic). This variable takes the value 1 if the person reported having been exposed to any act of violence within each of these broad categories in the 12 months prior to the interview; otherwise the variable takes the value 0. In line with the study’s conceptual framework, the incidence of domestic violence was regressed on a set of variables reflecting individual, household and community factors, as well as a set of regional variables.25 25 Each of the ten regions of Ghana is represented by a dummy variable with value 1 if the respondent lives in the region, and 0 otherwise. The inclusion of regional variables – or regional fixed effects – allowed us to account for some of the unobserved variation of domestic violence across regions that would otherwise be 185 The variables used in the regressions were derived from the conceptual framework discussed in Section 2, as follows:  Intergenerational dimension of domestic violence: measured using a binary variable indicating whether the respondent witnessed each type of domestic violence in their home during childhood.  Marital status: measured with a variable indicating whether the respondent: (i) was not married ever; (ii) was in a monogamous relationship; (iii) was in a polygamous relationship; or (iv) was divorced, separated or widowed.  Socio-economic status: captured by an asset ownership index. The assets included were: radio, TV, mobile phone, bed, other furniture, cooking utensils, washing machine, fan, air conditioner, refrigerator, iron, sewing machine, kerosene stove, gas/LPG/biogas stove, electric stove, open stove, dish/cable TV, wall clock, landline telephone, digital camera, tablet, computer and mosquito net. The research team chose to measure socio- economic status through ownership of assets, as this measure is generally recognised as including less measurement error than income (Deaton, 1997).  Education: measured using a variable that indicated the different levels of schooling attained by each respondent.  Empowerment: following the existing literature (see review in Justino et al., 2012), empowerment was measured using data on decision-making. The research team constructed a decision-making index to identify the involvement of the respondent in making decisions about the following domains of household life: food purchases, daily purchases, major purchases and health care (for self).  Patriarchal norms: following the discussion on attitudes towards domestic violence in the previous sections, the prevalence of patriarchal norms was measured using a patriarchal gender norm index,26 a standardised index of acceptance of wife-beating27 and an index of tolerance of women’s sexual autonomy.28 included in the error term. Since observations are clustered at enumeration area level, the regressions show corrected standard errors. 26 This index was constructed based on the number of the following statements for which the respondent fully or somewhat agrees: ‘A good wife obeys her husband even if she disagrees with him’, ‘Family problems should only be discussed with people in the family’, ‘To bring up, raise or educate a child properly, the child sometimes needs to be physically punished’, ‘If a women wears revealing clothes, it is her fault if she is raped’ and ‘It is allowed for men to stare at women they don’t know’. The index was standardised to have a mean of 0 and a standard deviation of 1. 27 This index correspond to the number of the following cases for which the respondent agrees that a husband has a good reason to hit his wife: “she does not complete her household work to his satisfaction”“, “she disobeys him”, “she refuses to have sex with him”, “she asks him whether he has other girlfriends”, “he suspects she has been unfaithful”, “he finds out she has been unfaithful”, “she contracts a sexually transmitted disease”, “she goes out without telling him”, “she neglects the children”, “she argues with him” and “she burns the food”. The index was standardised to have a mean of 0 and a standard deviation of 1. 28 This index corresponds to the number of the following cases for which the respondent agrees that a married woman can refuse to have sex with her husband: “she does not want to”, “he is drunk”, “she is sick”, “he mistreats her”, “she is menstruating”, “he does not want to use a condom/contraception” and “she finds out that he has other girlfriends”. The index is standardised to have a mean of 0 and a standard deviation of 1. 186  Exposure to other forms of violence: measured using an index that included how many of the following types of violence the respondent had witnessed in their community: theft or robbery, vandalism, arson, kidnapping or abduction, extortion, bullying, political violence, highway attacks, harassment, rape, defilement and seduction.29 In addition to the variables above, the pre-analysis of the qualitative and quantitative data suggested the inclusion in the regression analysis of an additional variable – drinking habits of the respondents, measured by a binary variable taking the value 1 if the respondent reported drinking, even occasionally. The regressions control also for the region of residence and whether the household lives in urban or rural areas. The regression results are presented in tables F1–F4 in Appendix F, and discussed below, in conjunction with the findings of the qualitative data.30 4.5.1. Age and intergenerational effects The literature discussed in Section 2 revealed that experience of domestic violence is often found to be closely correlated to the age of the victim. In particular, younger women are generally more likely to report domestic violence than older women (Aizer, 2011; Garcia- Moreno et al., 2005). The data collected in this study for Ghana are in line with this international pattern: younger women – and men – were found to be more likely to have experienced any form of domestic violence than older age groups, as discussed in subsection 4.1. The results of the multivariate regressions confirmed these results. Older women were less likely to report experiencing domestic violence than younger women (Table E1, in Appendix E). Women aged 50–60 years were 84 per cent less likely to have experienced domestic violence than women aged 15–19 years (Table E1, in Appendix E). The same figures for physical, sexual, social and economic violence, respectively, were 99 per cent, 98 per cent, 89 per cent and 73 per cent. Even women aged 20–24 years were significantly less likely than women aged 15–19 years to experience all forms of domestic violence, except economic violence, for which there is no statistically significant difference (Table 29 The sample mean for this variable is 0.56, and the median is 0. This indicates that violence tends to concentrate in a few areas. We, therefore, use a binary variable taking the value 1 if the index of violence exposure is positive, and 0 otherwise. Similar results in the regression analysis were obtained when we used the continuous index. 30 The tables in Appendix E display the odds ratios associated with every risk factor, alongside its standard errors. The odds ratio shows how much more (or less) the odds are of experiencing domestic violence when a certain risk factor is present, in relation to respondents who did not experience the risk factor. An odds ratio greater than 1 means that the variable increases the risk of domestic violence. As an example, an odds ratio of 1.5 for individuals with no education indicates that the odds of experiencing domestic violence are 50 per cent higher among individuals with no education in relation to those with education. An odds ratio of 2 for individuals with no education indicates that the odds of experiencing domestic violence are 2 times (100 per cent) higher among individuals with no education than those with education. If instead the odds ratio is 0.2, this indicates that people with no education are 80 per cent less likely than people with other levels of education to experience domestic violence. 187 E1, in Appendix E). The odds of men experiencing psychological, physical, social or economic domestic violence were also reduced as age increased (Table E2, in Appendix E), though the statistical significance of this association is more variable than the results shown for women. The odds of men being victims of sexual violence did not vary significantly with age. The literature reviewed in Section 2 also discussed how domestic violence is typically related to strong intergenerational effects (Clark et al., 2010; Gallegos and Gutierrez, 2011; Bobonis et al., 2013; Pollak, 2004), a result that was also revealed in the Ghana 2008 DHS study (GSS et al., 2009). This study found similar results for Ghana in 2015. Except for sexual violence (which may be more hidden from the view of children than other forms of domestic violence), women who witnessed any form of domestic violence at home as children were significantly more likely than women who did not witness violence in their childhood to have been exposed to the same type of domestic violence in the 12 months prior to the survey (Table E1, in Appendix E). The variations were large: women who witnessed psychological violence as a child were 2.2 times more likely to be exposed to psychological violence than women who did not witness this form of violence during their childhood. The corresponding figures for social and economic violence were almost two and three times higher. The effect for physical violence was more modest but remained large: women who witnessed physical violence as a child in their homes were 73 per cent more likely to be exposed to physical violence than women who did not witness this form of violence during their childhood. Men who witnessed social or economic violence in their homes as children were 2.2 times more likely to have experienced these forms of violence than men who did not witness these forms of violence in their childhood. Men who witnessed psychological violence as a child were 80 per cent more likely than men who did not witness psychological violence in their homes during childhood to have experienced this form of violence as an adult in the 12 months prior to the survey. The effect was not statistically significant for physical or sexual violence (Table E2, in Appendix E). Witnessing violence as a child, therefore, has important generational effects, as these children were more likely than others to have experienced domestic violence themselves as adults. The effect also applies to the perpetration of domestic violence. According to the GFLHS 2015, about 28 per cent of women and men who admitted committing social violence witnessed social violence as children, compared to 14–15 per cent of those women and men who did not experience violence as children. Similarly, among men who witnessed sexual violence as children, almost 30 per cent committed sexual violence over the 12 months prior to the survey, compared to 2.6 per cent of men who did not experience this form of domestic violence during their childhood. Across psychological, physical and economic violence, rates of perpetration are between two and three times higher for women and men who witnessed domestic violence as children than among individuals who did not experience domestic violence during their childhood. 188 4.5.2. Socio-economic effects Section 2 discussed in detail how domestic violence was generally found to be associated with low levels of income, wealth and education in many parts of the world. These socio- economic factors were found to be some of the most important determinants of domestic violence in Ghana in 2015. The vast majority of the research participants identified fights over money, assets or property as the leading determinants of domestic violence across all regions. Economic tensions were mentioned as leading to multiple forms of domestic violence among adults – including between partners, siblings and in-laws. Other factors reported frequently included education, poverty and the timing of certain economic events. These findings are discussed in turn below. Assets, property and inheritance. The data for Ghana discussed in the previous section showed that women and men in households with lower levels of assets were more likely than individuals in higher asset quintiles to have experienced physical, psychological and economic violence. Women in the lower asset quintiles were also less likely to have experienced social violence than women in higher asset quintiles. However, social violence among men and sexual violence among both men and women was more concentrated among those in the top asset quintiles. The regression analysis in Table E1 (in Appendix E) confirms that the asset ownership index is negatively and significantly associated with psychological and economic violence among women: for each of these forms of violence, a one unit increase in the asset index is associated with a five per cent reduction in the odds of a given woman experiencing domestic violence. The results in Table E1 indicate that an increase of one standard deviation of the index is associated with a reduction in psychological and economic violence among women of almost a third (32.5 per cent) – a very large effect. Asset ownership is not statistically associated with any other form of domestic violence against women – indicating that low levels of wealth are only significant risk factors for psychological and economic violence among women. The results among the male sample are different. Asset ownership is not statistically associated with the odds of men experiencing psychological, physical or economic violence. However, asset ownership is positively correlated with the odds of men experiencing sexual or social violence (as discussed in subsection 4.1.2): the likelihood of men experiencing sexual or social violence increased by 3 and 2 per cent, respectively, when the asset index increases by one unit. This result confirms that men of higher economic status were more likely than men in lower asset quintiles to have experienced social and sexual violence in the 12 months prior to the survey. Even though higher levels of asset ownership lowered the likelihood of some forms of domestic violence (particularly among women), the qualitative data showed that one of the major causes of fights in families who experienced domestic violence were disputes over land assets. This finding was particularly strong in the northern regions, where 189 gender norms played a major role in land conflicts in polygamous contexts that were arbitrated by customary law. Tensions were particularly high when, for instance, the first wife of a deceased man had only girls and the second wife had boys: the family of the second wife could in principle push for the girls not to inherit any of their deceased father’s property. If properties were not registered, families would have to divide the property among themselves, which often raised suspicions and mistrust. A female key informant, who works in litigation in the Northern region, explained when these conflicts occur: “When one person tries to cheat other family members in sharing family property or when one person tries to dispose of a family property without the approval and consent of other family members or to the detriment of the other family members.” In addition: “If the man has more than two wives when he dies, the sharing of his property becomes a problem. Because everybody wants to inherit the man’s property and that brings disintegration into the family.” (FGD, Upper West region, Jipara district: female group, 18–50 years) Focus group discussions across all ten regions also referred to fights over assets between siblings, and within the wider kinship network, when parents died as determinants of domestic violence. The following excerpt from an interview with an opinion leader in the Ashanti region further captures some of the dynamics of conflict among siblings that were frequently cited as causing violence across households in the region: “Land matter is one of the disagreements and fights in families. The sharing of the land for farming purposes becomes a problem.” The same person added: “There is fighting in the family because of cheating in the sharing of the family properties like houses, lands, cattle, sheep, goats or trying to take what rightfully belongs to the other family member(s).” Education. Section 4.2 discussed how women with primary education were more likely than women with other levels of education to have experienced physical, psychological or economic violence. Social and sexual violence was more predominant among women with secondary education. There was a much less clear pattern of association between education and domestic violence among men, but men with higher levels education were more at risk of social and sexual violence. The multivariate regressions show very varied patterns of association between education levels and domestic violence among women: (i) women with primary school education are 82 per cent more likely to have experienced physical domestic violence than women with no education; (ii) women with secondary education are 2.5 times more likely to have experienced sexual violence than women with no education; (iii) women with primary, middle and secondary education are, respectively, 82, 89 and 71 per cent more likely to have experienced social violence than women with no education; and (iv) women with 190 primary education are 50 per cent more likely to have experienced economic violence than women with no education. There was no statistically significant association between education levels and the incidence of psychological domestic violence among women (Table E1, in Appendix E). In general, however, it appears that women with no education are less likely to experience any form of domestic violence. These results may, however, be due to the fact that women with more education are also more likely to report having experienced domestic violence, a factor highlighted in the international literature and in previous studies in Ghana, as discussed in Section 2. Education is largely unrelated (or only weakly statistically associated) to the likelihood of men experiencing domestic violence. The exceptions are economic and sexual violence. Men with primary, middle, secondary, technical or higher education are, respectively, 2.8, 2.9, 2.9, 3.7 and 2.5 times more likely to experience economic violence than men with no education. Men with no education seem to be at lower risk of experiencing this form of violence. Education also matters significantly for the incidence of sexual violence among men: men with any level of education are substantially more likely to experience sexual violence than men with no education, for whom the likelihood of experiencing sexual violence is very close to zero (Table E2, in Appendix E). Poverty and non-compliance with economic responsibilities. Several respondents in the focus group discussions and in-depth interviews identified the failure of men to provide for their family as a major cause of conflict within families, as discussed in the previous section. Verbal insults and public humiliation by women often followed the failure of men to provide adequate economic resources, and, in turn, were seen as triggering the use of physical violence by men against women. This view was widely held across all regions and most discussion groups, as exemplified in the following statement: “When a woman nags or complains a lot if her husband is unable to provide her needs, it obstructs the man from having peace of mind at home. And this makes him go out of the house often and returns late in the night.” (FGD, Upper East region, Garu-Tenpane district: female group, 18–50 years) The GFLHS 2015 data confirmed similar dynamics: 9.4 per cent of men who never denied chop money suffered from psychological domestic violence; this rate increased to 19.9 per cent for male respondents who ever denied chop money. Across all ten regions, the respondents – in particular, older married women who participated in the focus groups – made links between these forms of domestic violence and the overall structural economic situation beyond the household, which has left men (and women) unable to provide for their households. These structural constraints included lack of education and business skills, as well as the availability of few paid jobs in the wider economy. In a context characterised by few formal employment opportunities, many men and women were often compelled to set up their own business, many without having acquired adequate skills. Frustration and the inability to cope with these wider structural economic problems were identified as factors that fuelled domestic tensions and resulted 191 in psychological violence committed by women against men, and in physical violence committed by men against women. The quantitative data allowed some of these findings to be explored in detail. In particular, the research team analysed the association between employment status and domestic violence, because employment opportunities are typically a good indicator of the overall wealth of the economy, and a determinant of households’ economic welfare. We discussed in section 4.2 how, overall, women and men who were not working at the time of the survey were more likely to experience any type of domestic violence than women and men who were employed or self-employed. The exception was sexual violence – as discussed before, this form of violence was more prominent among women and men who were employed. However, the regression analysis in Tables E1 and E2 (both in Appendix E) did not support these results. These tables show that there is almost no correlation between the employment status of women or men and their experiences of domestic violence, suggesting that other factors associated with employment status may be driving the results (for instance, asset ownership or education). The main exception is domestic economic violence: (i) women who were not working were 43 per cent more likely to have experienced economic violence than self-employed women; and (ii) men who were employed or not working were, respectively, 2 times and 2.5 times more likely to have experienced economic violence than self-employed men. In line with previous discussions, employed men are at greater risk of sexual violence (Table E1, in Appendix E). Timing of economic events and expenditures. Related to the findings above, domestic violence was also found to be linked to specific moments, festivals, events and seasons, such as Christmas, the beginning of the school year and pay day. An opinion leader in the Eastern region summarised this issue, widely reported across all regions: “Yes, normally some of these fights happen during funeral times, festivals and also when schools resume and parents have to pay school fees, sew new school uniform and other stationery.” In addition, conflicts about expenses at the start of school were reported across Ghana. Respondents from all regions mentioned that women usually buy the teaching and learning materials for their children with financial support from their husband. When children are not properly clothed or are missing items, they might be sent home, bringing shame to their family. Fighting about school expenses predictably takes place, as illustrated in this quote, which reflects nationally representative findings: “When schools reopen, usually within the second and third weeks. It is during this period that we need the children’s learning materials. So this information is put on paper to be given to their parents, and this can cause fights if fathers refuse to follow that.” (IDI, opinion leader, Northern region, Sagnerigu district) 192 In farming communities, harvesting or planting seasons were often mentioned as a cause of conflicts between couples, particularly when views diverged as to whether to send children to school or keep them at home to help on the farm. The following respondent expresses a view often encountered in rural areas: “Mostly in the rural and farming communities, it happens between the months of May to October, when the first rains fall and the farmers plant their crops. Often the man may ask the children to stay home from school and help with the farm work, the woman may object to this and it often results in violence.” (IDI, opinion leader, Northern region, Sagnerigu district) Spending monthly salary payments on personal pleasure – especially on drinking – was found to lead to domestic conflicts. A community leader articulated this nationally representative finding, saying: “When workers are paid at the end of the month and the man does not cater for the family, but is seen at drinking spots, it can also cause fights within the families.” (IDI, Eastern region, East Akim district) Alcoholism was mentioned in every region, and almost every district, as a particularly problematic determinant of domestic violence, with men who consume excessive levels of alcohol being more likely to commit violence against their wives. This was more frequently articulated to be a problem in communities where pito or akpeteshi (local beers/spirits) consumption was high in the poorer northern regions. For example, a community leader explained that: “Some of the men force the women to go and work on farm, whiles they sit at home, sometimes drinking ‘pito’ – that is the local alcohol drink.” (IDI, Upper East region, Garu Tempane district) The regression analysis in Table E1 (Appendix E) showed that alcohol consumption was significantly associated with 39 per cent higher odds of women experiencing economic violence. Alcohol habits are associated with a 66 per cent increase and a doubling of the odds of men experiencing psychological and sexual violence, respectively (Table E2, in Appendix E). These results do not indicate a causal relationship between alcohol consumption and domestic violence but strongly suggest that the risks of domestic violence incidence increases when alcohol consumption is high. 4.5.3. Marital status Several international and Ghana-based studies reviewed in Section 2 have shown that marital status is an important factor associated with domestic violence. The discussion in Section 4.2 on incidence of domestic violence showed mixed patterns: married women (or those living with a partner) were at higher risk of sexual or economic domestic violence than unmarried women, whereas divorced, separated or widowed men were more at risk of social or sexual violence than married or never-married men. All other forms of violence were more prominent among non-married women and men. 193 These results change substantially in the regression analysis in Tables E1 and E2 (Appendix E). The results show that, in general, not being married reduces the likelihood of having experienced domestic violence among women and men. Furthermore:  Divorced, separated or widowed women were twice as likely to have experienced physical violence than women who were never married.  Married women or divorced, separated or widowed women were, respectively, 1.8 times and 2 times more likely to have experienced economic violence than non- married women.  Married or divorced, separated or widowed men were, respectively, 6 times and almost 8 times more likely to have experienced sexual violence than non-married men, and almost 3 times and 1.7 times more likely to have experienced social violence than men who were never married. Table E1 (Appendix E) shows that women in polygamous relationships were almost twice as likely to experience psychological violence than women who were never married. These women were also 3.5 times more likely to have experienced physical violence and 4.2 times more likely to have experienced economic violence. Among men, polygamy increases considerably the odds of sexual or social violence against them (Table E2, in Appendix E). However, the association between the respondent’s marital status and experiences of different types of domestic violence could reflect reverse causality, as divorce and separation, for instance, may result from domestic violence. Disentangling these issues through quantitative data alone is very challenging. The qualitative data (and some additional quantitative data) suggest some plausible mechanisms that may explain these correlations, including access to sex within marital relations, extra-marital affairs and forced marriage, as discussed in more detail below. Access to sex within marital relations. The work of Ellsberg et al. (2001), cited in Section 2, showed how, in many parts of the world, marriage is seen as guaranteeing men unconditional sexual access to their wives. Conflict over sex was a ubiquitous finding in the qualitative data, prevalent across all regions, and in both rural and urban areas. The different compositions of the focus groups enabled frank discussions about these sex- related conflicts, particularly in the single-sex groups of married women and of married men (aged 18–50, and aged 50+). The principle source of conflict, across all regions, centred on men’s desire for sex at times when they were unable, or unwilling, to provide for the members of the household, including giving their wives chop money. This nationally representative finding was highlighted as one of the major causes of violence – particularly psychological violence and controlling behaviour. This finding suggests a close link between economic and marital determinants of domestic violence: women described how they expressed anger against their partner by refusing sex, while men responded or initiated marital conflicts by refusing to fulfil their economic obligations. In particular, most respondents across all regions agreed that both women and 194 men use sex as a weapon that has many forms, including being forced to have sex or being denied sex. Sex is part of an intricate power play between men and women, as explained in this interview with an education professional: “Fear can also make women feel [that] they need to accept the dominant or violent ways of the men. She is afraid to be beaten, afraid to be denied food and afraid of being driven out of the matrimonial home, and some even fear being denied sex, especially in the polygamous marriage. The man can proudly say ‘I will not have sex with you.’” (IDI, Northern region, Sagnarigu district) The quantitative survey included a number of questions about this important issue. Opinions diverged widely about whether women have a right to refuse sex. For instance, 8.2 per cent of respondents (9.8 per cent of women and 5.7 per cent of men) felt that the refusal of sex was a justifiable reason for wife-beating (Table 55). Employed respondents (4.6 per cent) were less likely to agree that the denial of sex was a reason for wife-beating than self-employed (9.2 per cent) and non-working respondents (8.1 per cent); as were respondents living in urban areas (4.8 per cent, compared to 11.7 per cent in rural areas), in the Greater Accra region (1.7 per cent, compared to 24.4 per cent in the Northern region, 24.0 per cent in the Upper East region and 29.4 per cent in the Upper West region), with higher levels of education (2.5 per cent, compared to 17.5 per cent among those with no education) and in the highest asset quintile (3.1 per cent, compared to 13.4 per cent in the lowest quintile). Extra-marital affairs. Men and women who engaged in extra-marital affairs were perceived to be at risk of domestic violence. The qualitative data indicated that this violence differed by sex, but that it was ubiquitous across all regions. Women who had extra-marital affairs were more likely to experience physical violence perpetrated by their partners, whereas men were more likely to experience psychological violence and controlling behaviour from their partners. There were, however, some striking anomalies in the qualitative data, which suggested that this gendered dichotomy of experiences of violence might not always hold true. In a focus group with young men and women (aged 18–25) in the Greater Accra region, this young person recalled a case of a man who found out that his wife was cheating: “What happened was that the man was called by the friend that he saw the wife somewhere with another man. Instead of the man asking the woman where she went, upon her return, he pounced on her and started beating her and insulted her. The woman also went into the room and got herself an iron and hit the man with it.” (Mixed FGD, Gbawe, unmarried/youth (18–25 years) The findings from the survey differed somewhat from the qualitative findings but are equally paradoxical: 0.4 per cent of the survey respondents agreed that unfaithfulness would be an acceptable reason to hit women, but 7.5 per cent mentioned that the suspicion of unfaithfulness was an acceptable reason for wife-beating (Table 55). This finding requires further future research, as the data collected are not sufficient to fully understand these findings: unfaithfulness emerges as a risk factor for domestic violence, but 195 perceptions and responses vary substantially across types and perpetrators of domestic violence. Forced marriage. Forced marriage may itself be considered a form of domestic violence perpetrated by families against girls with the consent of their parents. While not widely reported in the qualitative data, respondents in the Northern region reported that parents would be more likely to arrange an early or forced marriage when girls were found to be underperforming in school and paying attention to boys. While non-representative, this narrative was iterated by a number of community leaders and activists in interviews. For example, according to a faith leader in the Northern region: “As we sit now, there is a girl whose parents have arranged with an old man who is already married to two wives to support their daughter in school and marry her. The girl is now in senior high (SHS) and this is creating a serious fight between the girl and her parents.” (IDI, Northern region, Sagnarigu district) Similarly, an activist working on education explained that early marriages are a form of violence that reflect structural poverty and absent fathers: “Some of the children who are female may look for support from outside the house from men. No man will give out money free. Therefore, these girls may end up being impregnated. This is where early marriage comes in. She may be forced to be given out to marriage at that tender age, if she attends school, she may drop out of school making her situation worse and her future bleak.” (IDI, Northern region, Sagnerigu district) This issue was only raised in the discussion groups in the northern regions of Ghana. 4.5.4. Geographical effects The results showed several differences in how domestic violence and attitudes towards domestic violence were reported across the different regions of Ghana. The discussion in Section 4.2 highlighted that women and men living in urban areas were at greater risk of experiencing social or sexual violence, whereas women and men living in rural areas were at greater risk of experiencing physical, psychological or economic domestic violence. The experiences of women and men across the ten regions were very different, and no immediate pattern stood out. The regression analysis in Tables E1 and E2 (Appendix E), which control for a range of risk factors, showed no statistically significant difference in the likelihood of domestic violence experienced by women in urban and rural areas (the coefficients were not statistically significant). The only exceptions are that: (i) women in rural areas were 39 per cent more likely to suffer from economic violence than women in urban areas; and (ii) men in rural areas were 2.2 times more likely to have experienced physical violence than men living in urban areas. In all other cases, the effect of living in a rural or in an urban location was not found to be associated with domestic violence. 196 Most of the regional differences observed in women’s experiences of physical and social violence disappear in the multivariate analysis (Tables E1 and E2 in Appendix E). The only associations found between regions and the incidence of domestic violence included: (i) psychological violence against women in the Central, Eastern, Ashanti and Upper regions, and against men in the Eastern, Brong Ahafo and Upper East regions; (ii) physical violence against women in the Eastern and Upper East regions; (iii) sexual violence against women in the Central and Greater Accra regions; (iv) social violence against women in the Eastern region and against men in the Brong Ahafo, Northern and Upper East regions; and (v) economic violence against women in the Central, Greater Accra, Volta, Ashanti, Northern, Upper East and Upper West regions, and against men in the Volta and Upper East regions. Table 60 summarises this information. Table 60: Regions significantly associated with increased experience of domestic violence (by type of domestic violence and sex) Social Physical Sexual Psychological Economic violence violence violence violence violence Women Eastern Eastern Central Central Central Upper East Greater Eastern Greater Accra Accra Ashanti Volta Upper East Ashanti Upper West Northern Upper East Upper West Men Brong Ahafo Eastern Volta Northern Brong Ahafo Upper East Upper East Upper East Source: Ghana Family Life and Health Survey (GFLHS) 2015 and DHS 2008. Note: Based on results reported in Tables E1 and E2 (Appendix E). However, for both the male and female samples, regional effects were not collectively statistically significant (Tables E1 and E2, in Appendix E).31 These results suggest that, once the regressions control for a range of variables, there is no regional exceptionalism in patterns of domestic violence. In other words, the differences in domestic violence incidence discussed in Section 4.2 are due to a range of factors – possibly levels of poverty or economic development – but not to intrinsic geographical characteristics. 4.5.5. Patriarchal and gender norms Section 2 discussed in detail how patriarchal norms and perceptions of masculinity that support the persistence of certain gender norms may be closely associated with experiences of domestic violence. The association of domestic violence in Ghana with patriarchal 31 With the exception of economic violence, in the regression for men, for which regional effects are marginally significant. 197 cultural norms was evident in the qualitative and quantitative data collected across all regions and cut across the various socio-economic backgrounds of the respondents. The qualitative and quantitative data found that social norms concerning appropriate behaviour for women were a significant trigger for domestic violence by men against women in all regions. In the qualitative interviews and focus groups, the research team asked which expectations of masculinity make men behave in certain dominant or violent ways. The male and female respondents, in every region, discussed in detail which societal gender norms and expectations shaped appropriate and inappropriate behaviour for both men and women. The dominant view was that domestic violence was associated with non- compliance with these gender norms. As discussed in the section above on attitudes to violence, the quantitative and qualitative data found that different forms of domestic violence were used to ‘punish’ men and women, and young people, when they had been found to contravene normative expectations of appropriate behaviour. For instance, the qualitative data found that men were more likely to become violent when their expectation of being obeyed was not fulfilled by their wives, or when wives failed to perform expected domestic roles, such as ensuring their meals were ready on time, or refused to have sex. A community elder in the Ashanti region explained a widely held view across all regions: “Men, like children, expect to be pampered at home. Thus, if they do not get this from their wives, it makes them act violently in the house. Also when they are refused sex it makes them go wild in the house. The other main reason why men act violently at home is when women do not [fulfil] their wifely duties. For instance, you know the man sometimes come back from town in the company of friends and when the wife always leave the house untidy, he will fight with you.” The respondent added: “The man expects that he would be accord[ed] the needed respect and dignity at home by both his wife and children so, if he is not getting it, it becomes a problem. It also happens when the man sees that the woman is inciting the children or trying to organise the children against him. The man also expects that his instructions must be obeyed at home.” The qualitative data from all regions, and especially from the Northern region, also found that societal norms governing expectations about women made them more accepting of abusive relationships. Both male and female respondents pointed to the ‘expected’ submissiveness of women within conjugal relationships, women’s roles as primary caregivers (for children) and their economic dependence on men as reasons for them to accept violence. For example, a health practitioner in the Ashanti region discussed these gender dynamics in an interview, saying: “At times they can say I am the head of the house. Whatever I instruct you to do, you do it. I give you money daily for you to eat. I clothe you. That is why at times they do that. It could also be our bringing up as Ghanaians and also what they tell married couples on the day of their marriage. Whatever your husband tells you to do, do it. Even if your husband beats you, stay in the marriage. The marriage counselling she received says that the husband is the head. Even in our churches they say the husband is the head. Maybe it could be that the 198 woman does not wash the man’s clothes, or not giving the man food to eat. When you also deny the man sex, he can behave in this violent way.” This view emerged across the focus group discussions in all ten regions, and is illustrated by the following exchange in the Central region: Respondent 8: We are told from the word go that we have to be submissive so we seem to have no other choice. Sometimes is because of the children so we don’t even care what happens to us. I will plead with men to have patience for women. Respondent 7: Sometimes is because the man takes care of all finances so we can only be submissive no matter what the case may be but in cases where at least the woman is engaged in some occupation and can also support financially then the domination will be minimised. Respondent 2: The men take care of us, so we cannot even go anywhere without their permission. (FGD, Central region, Anomabo district: male and female group, 18–25 years) The quantitative data allowed the research team to analyse the effect of gender norms on domestic violence using four distinct variables: a patriarchal gender norms index; an index that measured tolerance towards wife-beating; an index that measured acceptance of women’s sexual autonomy; and a measure of decision-making power within the household. The results in Table E1 (Appendix E) show that women who exhibited higher levels of acceptance of wife-beating were 14 per cent more likely to have experienced economic violence than women with lower acceptance of wife-beating. As mentioned before, this is not a causal effect and may well suggest (as discussed in Section 4.2 on attitudes) that victims of violence internalise and accept gender norms that perpetuate or justify domestic violence. Women who expressed a higher acceptance of women’s sexual autonomy were 16 per cent more likely to have experienced psychological and social violence, 17 per cent more likely to have experienced physical violence, and 11 per cent more likely to have experienced economic violence than women with a lower acceptance of women’s sexual autonomy. This latter finding suggests, as discussed in the qualitative data analysis above, that women who are more likely to challenge accepted gender norms may be at risk of domestic violence because they may be punished for their non-compliance with established norms of behaviour. The results in Table E1 (Appendix E) show no statistically significant association between domestic violence against women and the patriarchal gender norms index or the levels of decision-making within the household. The results in Table E2 (Appendix E) show that men who expressed adopting stronger patriarchal gender norms were 26 per cent less likely to report social violence than men who did not accept such norms. However, men who displayed higher levels of tolerance towards wife-beating were 21, 47 and 30 per cent more likely to have experienced psychological, physical or economic violence, respectively, than men who reported lower levels of tolerance towards wife-beating. This result may potentially reflect retaliation by women who are beaten by their husbands, as discussed in the previous section. Men who 199 exercised more decision-making power within their households were 34 per cent less likely to experience social violence than men who exercised less levels of decision-making within the household. Taken together, and in conjunction with the qualitative data discussed above, these results suggest that gender norms shape the extent of domestic violence in Ghana. This finding is in line with recent research by Heise and Manji (2016), who have shown how social norms play important, albeit different roles, for actors functioning at different levels of the social ecological model. 4.5.6. Exposure to other forms of violence Some recent evidence from conflict-affected countries has suggested that domestic violence may be associated with other forms of violence that take place across countries and within communities. The survey questionnaire included questions on various forms of violence that may occur in the Ghanaian society, which were aggregated using an index that included how many of the following types of violence the respondent had witnessed in their community: theft or robbery, vandalism, arson, kidnapping or abduction, extortion, bullying, political violence, highway attacks, harassment, rape, defilement and seduction. The regressions in Tables E1 and E2 (Appendix E) show that exposure to violence in the community was significantly associated with higher levels of domestic violence. The odds of victimisation among women increased by 77 per cent for psychological violence, 49 per cent for physical violence and 57 per cent for economic violence when the respondent lived in a community that experienced at least one form of violence. The effect was stronger for sexual violence and social violence, for which the odds of having experienced this violence were 3.6 times and 1.9 times higher, respectively, for women exposed to violence in their communities than for women in communities that did not experience other forms of violence. The results showed an even stronger association between violence in the community and incidence of domestic violence among men. Men who reported some violence in their communities were between two and three times more likely than men living in more peaceful communities to report having experienced any form of domestic violence. The qualitative data emphasised, in addition, that in the northern regions of Ghana chieftaincy disputes between extended royal family members can erupt into community- wide violence if, for instance, the members of the extended family or even siblings fight over the right to the throne, or disputes and fights erupt between other family members who may support different candidates: “Sometimes too if the man and woman’s family are fighting over the right to the throne, it often results in fights and conflicts between them and sometimes a break of the union by the two families even when they love each other.” (IDI, Northern region, Sagnerigu district, Gbolo opinion leader) 200 Chieftaincy disputes were only mentioned in the northern regions, where they are described as “a major cause of disagreements and fights in this community.” (FGD, Upper East region, Garu-Tempane district: male group, 18–50 years) Despite the strong association between domestic violence and other forms of community violence, the study is not able to indicate the direction of causation. On the one hand, communities that experience high levels of community violence may also experience high levels of domestic violence because violence becomes an accepted way of behaviour. On the other hand, it is possible that experiencing violence within the family and immediate social relations may predispose individuals to other forms of violence. This result merits further research in future analyses of the GFLHS 2015. 4.6. Consequences of domestic violence in Ghana This section addresses the research question about the consequences of domestic violence in Ghana. The main consequences of domestic violence emphasised in the literature reviewed in Section 2 included effects on physical health, effects on mental health and emotional well-being, effects on concentration during and take-up of day-to-day tasks, and indirect effects on children. The qualitative and quantitative data collected in Ghana for this study revealed similar patterns, with a strong emphasis on the negative physical and mental health impacts of domestic violence. As with other sections, this analysis draws on complementarities between these different data. Multivariate regression analyses were also undertaken to explore the physical and mental health consequences of domestic violence. The study used three self-reported measures of physical health: health status, illness and chronic condition. Health status was captured by the following question: ‘In general, would you describe your general health as very good, good, fair, poor or very poor?’ ‘Very good’ and ‘good’ categories were grouped together into a binary variable indicating ‘good health’. Illness was captured by the following question: ‘In the last 30 days, have you felt sick or unwell?’ Chronic condition was measured by the following question: ‘Do you have any chronic condition?’. The study also used the K6 scale of mental illness developed by Kessler et al. (2003) to assess mental health. The K6 scale is constructed from a survey module that asked how often in the past 30 days the respondent felt any of these six symptoms: nervous, hopeless, restless, so depressed that nothing could cheer them up, that everything was an effort, and worthless. For each of these six questions (K6), possible answers included: ‘all the time’, ‘most of time’, ‘some of the time’, ‘a little of the time’ and ‘none of the time’. A respondent who answered ‘all the time’ across the six items was attributed a maximum K6 score of 24 (6*4). Respondents with K6 scores exceeding 13 were considered to be suffering from serious mental illness (SMI) (Kessler et al., 2010). The regression analysis used both the continuous (K6 scale) and the binary (SMI) variables in the analysis. Regressions, in addition, control for age, education, marital status, socio-economic status, residence and violence in the community, as well as food security (which takes the value 1 if household members went to bed hungry, even if only rarely, during the last lean season), source of 201 drinking water (distinguishing between piped water into the dwelling; public taps, wells and springs; surface water; and bottle/sachet water) and access to toilet (distinguishing between private toilet; shared toilet between a few families; public toilet; and no toilet). 4.6.1. Physical health effects To measure the impact of domestic violence on respondents’ health, the quantitative survey asked respondents who reported having experienced any act of sexual or physical violence whether the act had resulted in injuries. A total of 182 men and 315 women across the whole GFLHS 2015 experienced an act of sexual violence in the 12 months preceding the survey, of which 32 incidents were domestic acts of violence against men, and 78 against women. Out of these, nine women and one man reported having received injuries. Seven women were injured once or twice, and the man and two women more than five times. Four of the respondents needed health care following their injuries; one of the women needed health care three times in the 12 months prior to the survey. Fifty men and 192 women survey respondents reported having experienced physical domestic violence in the 12 months that preceded the survey. Injuries were reported by 114 of these respondents: 14 men and 77 women were injured once or twice during that time frame, one man and 16 women were injured between three and five times, and one man and five women were injured more than five times. Forty-six women and ten men were hurt badly enough to have needed health care: 32 once, and 24 more often. Other studies conducted in Ghana (e.g. Pool et al., 2014) have shown that violence against women is associated with serious health consequences in terms of pregnancy outcomes. The GFLHS 2015 found that 30 of the 192 women who reported having been physically assaulted were hit while pregnant (15.6 per cent), resulting in a miscarriage rate of 10 per cent (three women) and neurological complications to the foetus in 6.7 per cent (two women) of the cases. Another 6.7 per cent reported other, non-specified, effects on the child or mother. Domestic violence also appears to have exerted an adverse impact on general health outcomes: 43.8 per cent of women who had been physically assaulted declared having been ill in the 30 days prior to the survey – a rate that was one third higher than among women who were not physically assaulted (31.2 per cent). Also, 42.3 per cent of women who had experienced psychological violence reported ill health, compared to 30.9 per cent of women who did not experience this form of domestic violence. The social ecological model adopted by this study understands the consequences of domestic violence within the complex interplay of individual, relationship, social, cultural and environmental factors. This understanding was evidenced in all focus group discussions and in-depth interviews that addressed the consequences of domestic violence, and particularly emphasised in the younger (mixed unmarried, 18–25 years) and older groups (mixed, 50+ years). These two groups mentioned, in addition, that the health 202 consequences of domestic violence were felt not just by the individuals involved (victim or perpetrator) but also by witnesses of the violence within the household, family and community. At the individual level, domestic violence was described as affecting several health complications such as increased blood pressure, mental disorders, depression, eating disorders, sleeplessness, alcoholism and suicidal thoughts. These responses sum up the range of views on the health consequences of domestic violence: “Psychological violence is also likely to cause the victim stress and pressure because continuously he or she would have to be thinking and feel isolated. These could also lead to several health complications such as increase in blood pressure, mental disorders, depression, eating disorders, sleeplessness, alcoholism, suicidal thoughts and others. Sexual violence also results in unwanted pregnancy, chronic pelvic and abdominal pain and contraction of STIs.” (IDI, opinion leader, Kukurantumi/Koforidua East Akim district, Eastern region) Physical violence was commonly noted as affecting the physical health of women and men. For men in particular, and in all regions, alcoholism was consistently mentioned by all categories of focus group discussions and in-depth interviews as both a consequence and a cause of violence (as discussed also in the previous section). This nationally representative view is well expressed by a respondent in a focus group discussion in the Greater Accra region: “It [domestic violence] affects the man and woman mentally, emotionally, psychologically, to the extent that the men start drinking alcohol and the woman suffer from depression.” (FGD, mixed 50+, Tomefa, Ga South district, Greater Accra region) Focus group discussions across all regions mentioned that household and community members who witnessed these outbursts of violence were also negatively affected by the violence, as illustrated in these two poignant quotes: “[The] aged in such environments could have their blood pressures rising upon hearing of a fight.” (FGD, mixed 50+, Gbawe, Ga South district, Greater Accra region) “[Domestic violence] causes emotional distress to neighbours or family members who may be in the household.” (FGD, married/living together women 18–50, Gbolo, Sagnerigu district, Northern region) These responses reveal the relationship between individual and contextual factors and demonstrate how domestic violence affects multiple social spheres. Table 61 shows the regression results for the effect of each form of domestic violence – psychological, physical, sexual, social and economic – on the physical and mental health of women. The results show that:  Women who experienced psychological violence were 29 per cent less likely to assess their health as very good or good than women not affected by violence.  Women who experienced physical violence were 67 per cent more likely to have had an illness in the month prior to the survey than women not affected by violence. 203  Women who experienced social violence were 32 per cent less likely to have assessed their health as very good or good than women not affected by violence.  Women who experienced economic violence were 40 per cent less likely to assess their health as very good or good, 36 per cent more likely to report having been ill in the month prior to the survey and 66 per cent more likely to have reported suffering from at least one chronic condition than women not affected by violence. Table 61: Relationship between experience of domestic violence and physical health (women) (1) (2) (3) Health status Illness (1=ill Chronic conditions (1=very during last (1=yes) good/good) month) Logit Logit Logit Psychological violence 0.71* 1.19 1.10 (0.13) (0.19) (0.23) Physical violence 0.71 1.67*** 1.38 (0.16) (0.29) (0.34) Sexual violence 0.97 1.30 0.80 (0.34) (0.39) (0.29) Social violence 0.68** 1.41** 1.20 (0.13) (0.20) (0.21) Economic violence 0.60*** 1.36** 1.66*** (0.091) (0.19) (0.30) Observations 2,971 2,971 2,963 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Note: The table shows the odds ratios. Standard errors in parentheses are clustered within enumeration areas (EA). * p < 0.1, ** p < 0.05, *** p < 0.01. Each column represents the results of a regression equation that includes all controls included in the tables in Appendix E. The results in Table 61 show that sexual violence among women did not have any statistically significant effects on their health. In stark contrast with the results for women, Table 62 shows that none of the forms of domestic violence were statistically related to the physical health status of men. 204 Table 62: Relationship between experience of domestic violence and physical health (men) (1) (2) (3) Health status Illness (1=ill Chronic conditions (1=very during last (1=yes) good/good) month) Psychological violence 0.54 0.76 0.94 (0.26) (0.29) (0.46) Physical violence 0.67 1.55 0.99 (0.71) (0.80) (0.69) Sexual violence 0.57 2.23 (0.48) (1.67) Social violence 0.43 1.00 1.63 (0.26) (0.33) (0.78) Economic violence 0.55 1.37 1.32 (0.25) (0.41) (0.64) Observations 695 695 686 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Note: The table shows the odds ratios. Standard errors in parentheses are clustered within enumeration areas (EA). * p < 0.1, ** p < 0.05, *** p < 0.01. Each column represents the results of a regression equation that includes all controls included in the tables in Appendix E. The higher incidence of health effects of domestic violence among women may be due to women being more vulnerable to physical assault than men. As a respondent poignantly pointed out: “The abusers are natural[ly] stronger. So if it comes to the physical attack very few women will be able to outwit men. Most times the men overpower the women.” (IDI, opinion leader, Jirrapa district, urban Jirrapa) As discussed in Section 4.1, this view was common across all regions, as illustrated by this quote: “There is a difference in the violence that men and women experience because a man is stronger than a woman. As a result, women are mostly hurt in violence than men.” (FGD, Upper West Jirrapa district, rural Wuling) And, when “…the violence is physical, it can cause injury to the victim which could lead to permanent deformity in the long term.” (FGD, married/living together men, Gbolo: Sagnerigu district, Northern region) 205 The qualitative data showed that the consequences of physical abuse, including slapping, pushing, kicking, strangling and using weapons, varied from minor injuries to death. Physical injuries often led to hospitalisation and incapacitated victims: “Women experience a lot of domestic violent cases more than men. A lot of women end up losing their lives because they are most often at the receiving end of these brutish attacks.” (IDI, opinion leader, Goaso, Ahafo North district, Brong Ahafo region) The majority of respondents interviewed who work in health services mentioned having seen permanent disability, scars and even death as a result of physical violence, particularly against women. 4.6.2. Mental health and emotional effects The survey data reported a very strong correlation between exposure to domestic violence and mental health. Using the K6 scale for depression and anxiety, Table 63 shows that the likelihood of suffering from serious mental illness (i.e. having K6 scores exceeding 13) was considerably higher for women and men exposed to domestic violence in the 12 months prior to the survey than for those who were not. Women who reported serious mental illness were 26.7, 24.2, 21.9, 19.3 and 26.3 per cent more likely to have experienced, respectively, psychological, physical, sexual, social or economic violence than women who did not report serious mental illness. The association between serious mental illness and domestic violence was generally lower among men, but still significant across all categories, with the exception of sexual violence. Table 63: Percentage of respondents who have experienced domestic violence reporting serious mental illness (by type of domestic violence and sex) Psychological Physical Sexual Social Economic violence violence violence violence violence Women Yes 26.7*** 24.2*** 21.9*** 19.3*** 26.3*** No 11.0 11.7 12.3 11.6 10.6 Men Yes 11.7* 19.0*** 6.9 12.7** 15.7*** No 7.9 7.9 8.2 7.8 7.6 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Note: Pearson Chi-squared test of equality of means between ‘yes’ and ‘no’ categories: *** p<0.01; ** p<0.05; *p<0.1 Table 64 illustrates further, using regression analysis, the consequences of domestic violence on mental health among women. Column 3 shows the effect of each form of domestic violence on their K6 score (women exceeding a K6 score of 13 were considered to be suffering from serious mental illness). Column 4 shows the same results using a 206 binary variable (which takes the value 1 if the woman was considered to be suffering from serious mental illness, and 0 otherwise). The results were as follows:  Women who experienced psychological violence had a K6 score that was 5.6 percentage points higher than women who did not experience psychological violence, and were twice as likely to suffer from serious mental illness.  Women who experienced social violence had a K6 score that was 2.4 percentage points higher than women who did not experience social violence, and were almost 1.5 times more likely to suffer from serious mental illness.  Women who experienced economic violence had a K6 score that was almost 10 percentage points higher than women who did not experience economic violence, and were 94 per cent more likely to suffer from serious mental illness. Table 64: Effect of domestic violence on men and women’s mental health (1) (2) (3) (4) Men: K6 Men: Serious Women: K6 Women: score mental illness score Serious (K6 > 13) mental illness (K6 > 13) OLS Logit OLS Logit Psychological violence 1.22 0.35 5.63*** 2.07*** (0.73) (0.23) (2.58) (0.43) Physical violence 4.12 5.86*** 2.07 1.18 (5.39) (3.75) (1.10) (0.27) Sexual violence 0.77 n.a. 2.95 0.94 (1.26) (2.45) (0.38) Social violence 1.04 0.80 2.41*** 1.47** (0.75) (0.50) (0.81) (0.24) Economic violence 3.05 1.64 9.95*** 1.94*** (2.09) (0.90) (4.44) (0.34) Controls YES YES YES YES Observations 704 687 2,972 2,972 R-squared 0.12 0.16 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Note: The table shows the odds ratios. Standard errors in parentheses are clustered within enumeration areas (EA). * p < 0.1; ** p < 0.05; *** p < 0.01. The results show that physical and sexual violence had no statistically significant effects on the mental health status of women. In addition, Columns 1 and 2 of Table 64 show that (as with physical health) domestic violence did not have statistically significant effects on the mental health status of men. The only exception was for physical violence, where those who have experienced physical violence were 5.8 times more likely than men who did not experience physical violence to suffer from serious mental illness. Some results from the qualitative analysis may explain this latter finding. Although this view was not nationally 207 representative, an opinion leader in the Volta region mentioned that some men took to drinking because of violence at home, which affected their mental capacity to work on the farm or at any workplace. 4.6.3. Effects on work and daily activities In line with the discussion in Section 2, the survey included a number of questions about how each act of physical or sexual violence experienced in the 12 months prior to the survey affected the daily lives of those who had experienced violence, in terms of their absence from or ability to work, go to school or do domestic work, their ability to concentrate on daily activities, their levels of confidence and on reporting feelings of living in fear. Thirty-two men and 78 women across the whole GFLHS 2015 reported consequences of sexual violence. The majority of respondents affected by domestic sexual violence (90.6 per cent of men and 87.2 per cent of women) reported no impact of sexual violence on their daily activities. Ten women and four men felt that domestic violence had adversely affected their ability to concentrate on work. Others (less than three individuals) reported that they could not focus at school, were unable to concentrate on their domestic chores, lost confidence in their own abilities and lived in constant fear. The GFLHS 2015 also asked about the effects of physical violence on the daily lives of individuals who experienced it. Fifty men and 192 women across the whole survey reported effects of physical violence. The most prominent consequence for men was inability to concentrate at work (four men). Among women, 39 were unable to concentrate while doing their domestic work, 37 were unable to concentrate while at work, 30 lost confidence in their own abilities, 22 lived in constant fear, and 21 missed work. The qualitative data also showed that women became economically more vulnerable as a result of domestic violence. This view was widely mentioned across all regions by participants in both focus group discussions and in-depth interviews, who discussed the extra burden placed by vengeful perpetrators as they refused to provide housekeeping or chop money following other acts of violence. Reflecting this view, an opinion leader stated: “The man will decide not to provide for the family. He will refuse to give chop money and also stop paying the children’s school fees.” (IDI, opinion leader, Gbi Kpoeta, Hohoe municipality, Volta region) As a result, women often came to accept abusive relationships because of their economic dependence on men. This economic motivation for the acceptance of violence was not nationally representative but was more visible in highly patriarchal sites and areas with greater levels of poverty in the northern regions, as illustrated by the following quote: “The victims, especially the married women, get scared that when they report their husbands to the police, they will be jailed and who takes care of them and their children.” (FGD, Married women, Gbolo, Sagnerigu district, Northern region) 208 4.6.4. Consequences for children Across all regions of Ghana, respondents in both focus group discussions and in-depth interviews mentioned the wide-ranging consequences of domestic violence for children and young people in the short and long term. A prominent effect discussed by respondents was the psychological impact of domestic violence on children who are deprived of parental care and socio-economic rights. This nationally representative view was well illustrated by this quotation: “Lack of peace in the family can cause the minds of children to retard. Their confidence level is also reduced as they grow in domestic violence homes.” (IDI, opinion leader, Gbi Kpoeta, Hohoe district, Volta region) Respondent after respondent emphasised how care for, and attention to, children may be compromised in homes that experience domestic violence. A respondent pointed to his own experience, in a view shared by many in focus group discussions and in-depth interviews across Ghana: “It will affect the children emotionally when their daddy is groaning and their mum is moody. I want to use my experience as an example. I had a little quarrel with my wife, the child was crying, I was expecting my wife to go for the child and my wife was also expecting me to go for the child, and because we were angry with each other, we didn’t know who should go for the child. I just had to be a man and go for the child because she was not ready to attend to the child.” (IDI, opinion leader, Garu, Garu Tempane district, Upper East region) Another way in which domestic violence may affect children was through the loss of their mother. This effect was only mentioned in areas dominated by patrilineal culture, where men retain possession of children upon separation, such as in the Upper East and Upper West regions. This point was illustrated by an interview with an opinion leader: “When it persists, the woman will seek separation or divorce and will become free from this burden. When the woman leaves, the children will not have motherly care because in this area, the children are considered to belong to the man.” In areas of matrilineal inheritance in the Obuasi district of the Ashanti region, responses typically pointed to women assuming responsibility for the care of children upon separation, with children being deprived of paternal care and the economic support that comes with it: “Most often it is the children and the woman that suffer especially when they are left in the care of an unemployed mother. Some men refuse to give house-keeping money for the use of the mother and the children.” (IDI, opinion leader, Obuasi, Obuasi municipality, Ashanti region) 209 Children growing up in abusive homes manifest their traumatic experience in school. Many respondents from the field of education pointed out this link between abusive homes and challenges in school: “Yes. Even just this morning, which is a holiday, a woman drove her child to come to class. Meanwhile the woman knows well that today is a holiday. To provide the child with just GHc1 is a problem. So she must find ways and means to escape that burden of providing the child with money, hence the decision to send her away from home. When the child comes to school, as her friends are eating, she might be lucky to get something to eat. Even today, one of my teachers gave a child GHc1 because the child was hungry. The moment she collected the money, she set off for home. When I enquired where her mother was, she replied she was at home. I enquired further and was told the parents are divorced and for that matter, when she was coming to school, she wasn’t given any money. At times we are compelled to sacrifice to feed some of these pupils, providing token money to these young children. Even we have declared some of these children ‘HIPC’, because anything you ask, they won’t bring it. We have about 8 pupils in this HIPC category in this school. At times too some of the teachers buy dresses for these pupils. These are problems we go through. Thus some parents are not treating their children well at all. This innocent child can be somebody in the near future.” (IDI, opinion leader, Goaso, Ahafo-Ano district, Brong Ahafo region) As a result of these experiences, children’s education is also affected, as they start: “absenting themselves from school or neglect their studies altogether” (IDI, opinion leader, Tomefa, Ga South district, Greater Accra region) In addition, echoing a view mentioned across all ten regions: “Some of the parents refuse to send their children to school. Some of the parents use the children to do some kind of jobs, which children cannot do. […]Sometime too, threaten them with dangerous weapons.” (IDI, opinion leader, Boakyekrom, Ahafo North district, Brong Ahafo region) These forms of neglect were discussed in particularly strong terms among focus group discussions and in-depth interviews in areas with high levels of poverty, as this quote illustrates: “Some parents abandon their children to the point where some children go to school on empty stomachs. Parents neglecting their children are a paramount situation here.” (IDI, opinion leader, Jirapa, Jirapa district, Upper West region) Children in turn may seek comfort in friends or resort to risky behaviours. The outcome in the long term may be waywardness and social deviance among the youth, as most opinion leaders involved in child protection services mentioned: “The children are not spared under these conditions. They are also negatively affected. They tend to look for happiness out of home due to the fighting and sometimes fall into bad company. Some begin to smoke wee, drink and even steal. The girls also look for boyfriends to support them because their fathers are not caring for them well. This often results in 210 pregnancies which have bad repercussions on their future or career.“ (IDI, opinion leader, Gbi Kpoeta, Hohoe municipality, Volta region) In essence, the results from the qualitative data highlighted how domestic violence experienced in the home, and the resulting divorce or separation, had both short- and long- term impacts on children’s educational, health, economic and social well-being. This is an important finding because, as discussed in subsection 4.3.1, experiences of domestic violence during childhood may themselves be reflect in vicious cycles of victimisation or perpetration of violence in adulthood. 4.7. Use of and access to victims’ support services and institutions This subsection addresses the final research question: Which institutional structures support victims of domestic violence, and what proportion of victims are able to effectively access available services? The analysis starts by examining the factors that may explain why women and men who have experienced domestic violence seek (or not) external help (subsection 4.7.1). It then looks at the public services and institutions available in Ghana to support those affected by domestic violence (subsection 4.7.2), how many women and men who participated in the study were aware of these services and institutions (subsection 4.7.3), and the levels of satisfaction with services among those who have used them (subsection 4.7.4). The main aim of this analysis is to uncover information about the current use and knowledge of and access to public services among individuals and communities, which will be useful for future interventions aimed at preventing and mitigating domestic violence in Ghana. 4.7.1. Seeking external help A large international study found that most victims of domestic violence did not report their abuse because domestic violence was often considered ‘normal’, or for fear of further violence, losing their children or causing shame to the family (Garcia-Moreno et al., 2005). Previous studies in Ghana had similar findings (Amoakohene, 2004). According to the GFLHS 2015, 853 women and 419 men across the whole sample interviewed in the survey reported having experienced at least one type of domestic violence in the 12 months preceding the survey. Around 35 per cent of these women (297) and men (146) tried to approach a person or organisation for help or support after experiencing violence. These numbers were very similar to the DHS 2008 survey, which showed that 37 and 34 per cent of women and men, respectively, who experienced violence had sought help. Table 65 shows the share of women and men who sought external help after experiencing any form of domestic violence. It reveals a wide variety of patterns among women and men’s help-seeking behaviour:  Women were more likely to have sought help if they experienced physical violence (52.8 per cent) or psychological violence (44.5 per cent) than for other forms of violence. 211  Men were more likely to have sought help if they experienced psychological violence (40.1 per cent) or economic violence (39.6 per cent) than for other forms of violence.  Women were more likely than men to have sought help if they experienced sexual, physical or psychological violence.  Men were more likely than women to have sought help if they had experienced social or economic violence. Table 65: Percentage of respondents who experienced domestic violence who reported seeking external help (by sex) Social Sexual Physical Psychological Economic violence violence violence violence violence Women 33.9 39.5 52.8 44.5 34.3 Men 35.6 35.1 34.3 40.1 39.6 N 531 110 242 487 508 Source: Ghana Family Life and Health Survey (GFLHS) 2015. When asked what made them seek help, the majority of women (58.5 per cent) and men (53.3 per cent) answered that they “could not endure more”. The second most common reason was the encouragement of friends and family (24.5 per cent for women and 28.5 per cent for men), followed by the extent of injuries sustained (7.5 per cent for women and 8.0 per cent for men) (Table 66). These findings suggest that domestic violence is still largely considered a private family matter: the GFLHS 2015 reported that 81.1 per cent of respondents believed that “family problems should only be discussed with people in the family”. Women and men who experienced domestic violence seemed to only resort to external help in more extreme circumstances. Table 66: Reasons for approaching external help after exposure to any form of domestic violence by sex (in percentages) Men Women Encouraged by friends/family 28.5 24.5 Could not endure more 53.3 58.5 Badly injured 8.0 7.5 Afraid person/persons would kill me 0.6 1.7 The person/persons who did this to me 0.9 0.5 threatened or tried to kill me Saw that children are suffering 0.0 0.7 Thrown out of the home 0.0 0.2 Encouraged by organisation 0.3 0.2 Other 7.4 6.3 N 146 297 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 212 Table 67 shows the reasons for not seeking help. The most common reason given was that they did not think they needed help (44.1 per cent of men and 32.5 per cent of women). In line with Garcia-Moreno et al. (2005) and the findings discussed above, 22.7 per cent of men and 17 per cent of women did not seek help because they considered violence experienced as ‘normal’ or ‘not serious’. In addition, 10.7 per cent of men and 16.4 per cent of women did not seek help because they did not know who to ask – an issue that will be explored further in the next subsections. Table 67: Reasons for not seeking external help after exposure to any form of domestic violence by sex (in percentages) Men Women Did not know who to ask 10.7 16.4 Person/organisation too far 0.2 0.1 Nobody can help me 6.9 7.6 I don’t need help 44.1 32.5 If my partner finds out it will cause trouble 0.1 1.7 If other family member finds out it will cause trouble 0.7 2.0 I don’t trust anyone 2.4 5.0 Violence normal/not serious 22.7 17.0 Embarrassed/ashamed 2.5 4.5 Afraid would not be believed 0.5 0.5 Afraid would be blamed 0.8 1.2 Afraid would end relationship 0.0 1.4 Afraid would lose children 0.0 0.3 Bring bad name to family 3.0 3.0 Other 4.6 5.7 Don't know or refused to answer 0.8 1.1 N 268 550 Source: Ghana Family Life and Health Survey (GFLHS) 2015. The qualitative data allowed the research team to explore the quantitative results further. Data collected through focus group discussions and in-depth interviews in all ten regions in Ghana revealed that distrust of, and poor experiences with, state authorities contributed to individuals not seeking help. A large number of the respondents mentioned three key challenges: the costs of the services, lengthy or ineffective delays in the handling of cases, and corruption. Corruption, in particular, was said to affect access to justice when survivors were given bribes and when perpetrators paid bribes or used their networks of influence to avoid prosecution. This view was well illustrated by the following statement: “There are times the assemblymen go to the police station to ask for the discharge of a culprit. Normally when they have some small money on them they give it to the police in charge and the matter will be dropped.” (FGD, married/living together men 18–50, Gbawe, Ga South district, Greater Accra region) 213 There was a widespread view that particularly the police and the courts collected money from those who approached them, but did not deliver appropriate care in return. This view is exemplified by this respondent: “For the police, they will collect money from you and would not help you solve the problem.”(FGD, unmarried youth, Sagnarigu district, Northern region) An opinion leader interviewed in another region added: “People only take their case to the chief because they cannot afford the monies the police will demand from them.” (IDI, opinion leader, Wuling, Jirrapa district, Upper West region) Respondents also mentioned that victims were often sent back by the police to their families to find a solution, and prosecution was discouraged: “If a police staff knows the victim, they will ask you to take the matter to the family level for solution.” (FGD, married/living together men 18–50, Gbawe, Ga South district, Greater Accra region) Concurring with this widespread perception of domestic violence as a private matter, even within public services set up to deal with domestic violence, the majority of respondents in the GFLHS 2015 who sought help did so in the first instance by approaching family members or a friend (66.2 per cent of men and 77.6 per cent of women). The second source of help was the police (sought by 14.6 per cent of men and 9.0 per cent of women as a first point of contact) (Table 68). Table 68: First choice of help (in percentages) Men Women Family member or friend 66.2 77.6 Health centre, hospital or other health 1.4 1.2 institution Police 14.6 9.0 Queen mother 0.9 0.7 Domestic Violence and Victim Support Unit 0.6 0.7 NGO, CSO or social worker 0.6 0.2 Lawyer or court 0.6 0.7 Traditional leaders 4.9 3.0 Religious leaders 2.3 3.5 Community group 2.6 1.2 Other 4.3 2.3 Don't know or did not answer 0.9 0.0 N 145 297 Source: Ghana Family Life and Health Survey (GFLHS) 2015. The qualitative data revealed, however, that the preference of study participants for traditional or formal legal ways of addressing domestic violence was often confounded by 214 underlying deficiencies in the justice system that, combined with structural inequalities, may have prevented victims of domestic violence from accessing it (see also Brickell, 2015a, 2015b). As stated earlier, the 2007 Domestic Violence Act encourages mediation of all types, but its reconciliatory approach brings some challenges. In addition to the challenges discussed above, poorer victims and those worried about personal social stigmatisation and reputational loss for the whole family may have preferred not to access formal justice processes. Also, the study found that respondents were more likely to withdraw a case from prosecution and seek customary mediation when the family or the community determined that the victim was: ”to blame … for being the cause of the problem, when the perpetrator shows remorse [and/or] when the family thinks about stigmatisation.” (IDI, opinion leader, Boakyekrom, Ahafo Ano district, Brong Ahafo region) Furthermore, respondents in focus group discussions said: “Family and community may discourage prosecution if it is minor case like verbal assault” or when prosecution could “lead to divorce” (FGD, married/living together women, Garu, Garu Tempane district, Upper East region) And, in many cases: “…the family would convince the victim not to report the case or follow it up for … prosecution of the perpetrator because the offender may be an opinion leader in the community or the breadwinner of the family. Also if they like the offender, then they would discourage his prosecution” and since “many cases of domestic violence that get prosecuted end up in divorce, … if the couple love each other [they] would prefer to work out their differences, the family would intervene to make sure that the union is not dissolved and so prevent divorce and family feuds; the family and some members of the community always try to discourage prosecution of such cases.” (IDI, DOVVSU, Gbawe, Ga South district, Greater Accra region) As these responses indicate, the rights of the victims are not always placed at the forefront of customary mediation processes. Customary mediation may be strongly desired and provide an important reconciliatory cost/time-effective approach in the Ghanaian social context. Yet, the rights of the victim may often become submerged by the social and economic needs of the family or community, and by prevalent gender norms. This situation may weaken victims’ rights, while encouraging perpetrators to act with impunity. 4.7.2. Public services and institutions The Government of Ghana has adopted a multi-institutional approach to support victims of domestic violence after the enactment of the 2007 Domestic Violence Act. The primary institutional response is the Domestic Violence Victim Support Unit (DOVVSU) of the Ghana Police Service. DOVVSU acts in partnership with the Department of Social Development, the International Federation of Women Lawyers (FIDA) and the Legal Aid Scheme, and collaborates with the health system to provide comprehensive support to 215 victims of domestic violence. To better understand how these services currently function, the research team conducted several observation exercises and interviews within DOVVSU units (as detailed in Appendix C), which gathered detailed information on: (i) the facilities available to and protocols followed by DOVVSU personnel; (ii) levels of training of DOVVSU personnel; and (iii) the extent of the collaboration between DOVVSU and public health services. Facilities and protocols. The 2007 Domestic Violence Act makes provisions for police assistance to victims and perpetrators of domestic violence. These include receiving complaints, offering advice about victims’ rights and about services available, helping victims to find a safe refuge when required, and providing assistance with free medical treatment and medical evidence for possible prosecution. The interviews and observations conducted during the study revealed that none of the DOVVSU units visited had all the necessary facilities to offer these support services effectively. In particular, they lacked private rooms, shelter facilities, counselling services, protection services or prosecution advice. DOVVSU personnel interviewed in the Upper West region explained that this situation is common across all ten regions: “We are handicapped so if you people could help us get fresh accommodation, it will help us very well in exercising our duties. We also need vehicles and other logistics to help us in our line of duty.” The research team’s visits to DOVVSU offices also revealed that, to date, the Ghana Police Service does not provide victims’ shelters at the national or district level. As mentioned in one of the interviews conducted with DOVVSU personnel in the Western region: “DOVVSU have a lot of challenges […]We have issues with accommodation and cannot house victims when the need arises. If DOVVSU had enough office space, it would have been convenient for some victims to be brought here.” The only CSO currently seeking to provide such a service, the Ark Foundation, receives no support from the government.32 Most DOVVSU units consisted mainly of personnel with limited office space. The few which had their own offices did not always have computers to keep their records. DOVVSU officials in all offices visited were found to be working under difficult conditions and, sometimes, paying for the medical or transportation bills of victims themselves. The needs of DOVVSU emerge in this poignant observation by the team at one of the DOVVSU units included in this study: “I saw that the police worked under difficult conditions ... they have a cage-like place near complaints which had two compartments, men who were detained were placed in the inner cell while women were placed in the outer one. […] I saw that DOVVSU needed a lot of support […] as the interviewed officer told me, […] DOVVSU did not even have a car to effect arrests. They had often had to use their little salary to foot transportation without 32 Information obtained in interviews with Mr Sandogo, Police Social Services, Director of the Ark Foundation and National Coordinator of the Domestic Violence Coalition in Ghana. 216 getting reimbursed. The winds often blew water into their offices as it rained because the louvre blades in the windows were either broken or not functioning anymore. This affected the work of the officials, as they sometimes have to move from their offices taking files along with them. Lastly, I must say this Officer was very helpful. […] The interviewed Officer was very friendly in nature and as I was interviewing him, we were constantly interrupted by people coming to call him out to assist them with their case at the unit.” Other detailed observation exercises revealed that, despite being overwhelmed with their workload and often lacking adequate facilities, the conduct of the police while dealing with cases of domestic violence was very professional, with care being taken to file documents appropriately (within infrastructure constraints) and to provide as much privacy and support as possible to those seeking their help. However, the facilities available varied considerably across regions, as follows:  Greater Accra region, Gbawe, Ga South district: There was no private room for victims to report their cases comfortably. All the offices visited, except the Commander’s office, looked cramped, with very little room for people to move. At the DOVVSU desk, there were over five personnel in a very small room. There was no fan, so the room was very hot. The room looked like an extension to the main building; from the outside, it looked like a cage. The seats in the offices were worn out, and many of them were old benches. Files were well stored away from the public.  Eastern region, Kukurantumi-Koforidua: DOVVSU officials were using four different offices. Another office, called the secretariat, was used to store all files related to domestic violence. There was a private room where victims could sit to file complaints.  Volta region, Gbi Kpoeta, Hohoe municipality: The police station had DOVVSU personnel on site, but there was no separate structure for the DOVVSU unit to use. There was only a small office that was used by the officer-in-charge. The rest of the DOVVSU officials shared the same office with the other police officers. They provided counselling in the same office as other police officers who could listen to the victims’ stories. There was very limited privacy.  Northern region, Kalpohin, Sagnerigu district: The office for the entire unit consisted of two small rooms, one of which was used by the head of the unit. The other office was shared by about ten officers. There was limited space for people to move about. There was no private room for victims, and reports had to be made in front of others. However, files were kept safely out of sight.  Upper West region, Wa municipality: Office space was very small, with files stored in a locker which looked quite old. There was no private place where victims could report their cases. Victims had to sit in an open place where anybody could hear them. The office did not have enough chairs and desks for the police personnel and victims. There were no vehicles for police personnel to use in effecting arrests or to follow up cases under investigation. The research team observed that DOVVSU officers were not in uniform, which made them more approachable to community members. They were very friendly with victims who approached their offices. 217  Upper East region, Garu Tempane district: The district did not have a DOVVSU unit or staff. Interviews were conducted with district police officers in their normal line of duty. Files were stored in a separate out-of-sight room. There was no privacy, with people being first directed to the charge office, where an officer registered their complaint. After listening, victims were referred to an investigator, who took their statement. Victims were offered a seat in front of the investigator in a private room, where he would write a detailed statement. A medical form was issued afterwards if necessary. The victim was invited to lead the police in the arrest of the perpetrator. If witnesses were available, victims were asked to provide them and lead the investigator to the scene of the crime. Victims were expected to avail themselves in court to testify.  Brong Ahafo region, Goaso, Ahafo North district: There was no building or office which catered solely for domestic violence victims. All reports were made at the charge office.  Central region, Cape Coast district: DOVVSU had its own separate office space, but it was in a poor state of repair. No private room was observed.  Ashanti region, Obuasi municipality: Files were stored in a cupboard in a restricted room at the police station. Privacy was ensured, and the procedure for filing a complaint was the same as described above for the Upper East region.  Western region, Tarkwa district: The DOVVSU unit had two rooms. The DOVVSU person was also the head of the police division and had to balance different duties. The two rooms afforded limited privacy: the first room functioned also as a waiting room and a reception area, whereas the second room was the head’s office. It was also where files were kept. Training of DOVVSU personnel. The establishment of DOVVSU units was intended to be accompanied by the training of personnel to respond to domestic violence cases. The large majority of the police personnel interviewed mentioned that they had not received specialised training to provide help to victims of domestic violence. Neither had they received training on gender-based violence or domestic violence, or specialised training in listening skills to better elicit information from those seeking help. The major formal training by the Ghana Police Service on gender-based violence took place in 2004. With the help of UNICEF, a training manual was recently prepared to make training in gender- based violence a basic part of the curriculum of the Police Training School; it should soon be more widely used. DOVVSU officers were also observed to have received limited training in information technology, and most district offices had no digital means or training to record or access local records. Collaboration between DOVVSU and public health services. There is close collaboration between DOVVSU and the district hospitals which provide treatment to victims of violence. This collaboration does not, however, include support services to victims of violence. This is because, to enable prosecution of assault cases, medical reports have to be completed by the medical doctor who treated the victim of domestic violence. But these medical services are not free of charge. Many doctors request ‘unofficial’ fees from victims 218 to cover the cost of writing the medical report and the possibility of having to go to court to explain it. This makes prosecution inaccessible to many victims, especially low-income women. One opinion leader interviewed in the Ashanti region expressed this view, which was common across all regions in Ghana: “DOVVSU is really responsible for addressing domestic violence issues, but one thing missing in their work is a centre for protection, where they will keep the victim before they carry on with investigations. This is an issue. Secondly, when it comes to issuing a medical report and they send it to the hospital, the doctors ask for a fee before they sign the report. Most of these victims are vulnerable people, and they are not able to pay such fees and this is a worry to us.” Another interview with a medical doctor who is a domestic violence activist revealed that many medical doctors were not trained on how to write medical reports for prosecution, which may sometimes give lawyers of perpetrators the upper hand in reinterpreting the reports to the benefit of their clients. Also, district hospitals do not have specialised services, such as emergency responses to victims of violence, or priority systems for victims of abuse unless they were brought into the hospital as normal emergency cases. In summary, while DOVVSU is lodged within an ideal multi-institutional framework within which to ensure effective support to victims of domestic violence, it is currently inadequately equipped to make operational the 2007 Domestic Violence Act. At the time of this study, DOVVSU faced the same institutional constraints that existed in the police service prior to legal reform (see Cusack and Manuh, 2009; Ellsberg and Heise, 2005). 4.7.3. Knowledge about support services and institutions Study respondents across all regions of Ghana showed awareness about the public services and institutions that deal with domestic violence (Table 69). DOVVSU, however, was still largely known by its previous name – the Women and Juvenile Unit (WAJU). The few districts where knowledge about the public support system was more limited (in the northern regions) were also districts where DOVVSU services were not available. In those districts, respondents typically stated they had neither heard, seen nor used any such public facilities to resolve domestic conflicts. But in districts where DOVVSU was established, respondents generally included DOVVSU in their list of known public institutions. Table 69 provides an overview of the knowledge exhibited by the GFLHS 2015 respondents about a range of available public services. The level of knowledge about public services and organisations was quite high among all respondents and across all different socio- economic groups and regions. With respect to DOVVSU, 84.9 per cent of all respondents knew about their local units. This level of knowledge was lower among women (83.7 per cent) than men (86.6 per cent), those living in rural areas (80.3 per cent) than those in urban areas (89.3 per cent), those living in the Central, Volta and Northern regions (respectively, 74.2 per cent, 53.6 per cent and 67.4 per cent, compared, for instance, to 99.0 per cent in the 219 Western region) and those who had no education (71.9 per cent, compared to 98.0 per cent for those with higher education) and were in the lowest asset quintile (70.1 per cent, compared to 95.1 per cent for those in the top asset quintile). 220 Table 69: Respondents’ knowledge about public services (in percentages) 1 2 3 4 5 6 7 8 9 10 11 12 Sex Male 99.9 98.9 86.6 53.8 78.9 94.9 82.6 94.7 98.4 99.6 98.7 97.9 Female 99.6 99.2 83.7 53.8 74.3 92.8 81.9 94.8 97.6 99.4 98.3 96.9 Age group (years) 15–19 99.2 99.5 80.5 52.7 71.9 92.7 78.2 95.2 99.0 99.7 98.2 97.5 20–24 99.1 98.7 85.8 56.5 78.1 94.5 82.0 94.6 96.8 98.8 98.0 97.1 25–29 99.9 98.8 87.7 53.9 78.0 93.5 84.3 94.8 97.5 99.2 98.2 97.1 30–39 99.6 99.1 86.5 55.6 76.7 93.3 83.0 94.4 97.9 99.6 98.4 96.7 40–49 100.0 99.2 81.9 50.6 73.4 94.0 81.0 94.5 97.9 99.8 98.8 98.0 50–60 100.0 99.2 83.6 52.4 76.4 93.7 82.0 95.7 98.9 99.7 99.1 98.2 Employment Self-employed 99.7 99.0 82.3 51.9 73.9 92.5 79.2 95.0 98.6 99.5 98.2 97.2 Employed 99.8 98.8 93.0 59.6 82.6 96.1 90.3 93.6 95.6 99.1 98.7 97.2 Not working 99.6 99.4 85.7 54.6 77.1 94.7 84.0 95.1 97.9 99.8 98.9 97.6 Marital status Never married 99.6 99.6 82.2 54.9 75.3 93.9 81.4 94.3 98.6 99.6 98.0 97.4 Married or living together 99.7 98.9 84.8 53.9 76.4 93.2 81.5 94.6 97.7 99.3 98.4 97.3 Divorced/separated/widowed 99.8 99.2 86.7 52.8 75.6 94.9 85.0 95.6 98.2 99.9 99.0 97.5 Residence Urban 99.8 99.1 89.3 56.8 79.6 95.8 87.8 94.6 96.6 99.6 98.9 97.2 Rural 99.6 99.0 80.3 50.7 72.5 91.4 76.4 95.0 99.3 99.4 97.9 97.5 221 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 Region Western 100.0 100.0 99.0 54.3 80.8 99.0 98.0 98.4 99.0 99.6 99.4 99.2 Central 99.5 98.4 74.2 25.8 54.0 96.5 84.7 96.5 98.8 99.5 99.5 97.7 Greater Accra 100.0 98.9 93.9 55.7 79.8 95.1 92.8 91.1 93.7 99.8 99.5 98.1 Volta 99.3 98.6 53.6 35.6 58.2 84.1 59.1 94.7 98.4 99.3 99.1 98.4 Eastern 100.0 100.0 96.0 72.4 92.4 99.0 98.9 100.0 100.0 100.0 100.0 99.8 Ashanti 99.9 99.5 92.6 70.1 83.3 97.1 87.9 99.1 99.5 99.8 99.5 98.6 Brong Ahafo 99.8 97.9 85.7 66.6 74.9 85.5 75.3 86.4 95.3 98.3 90.9 85.3 Northern 98.6 99.4 67.4 34.8 69.1 92.2 62.3 91.8 99.2 99.2 97.5 97.9 Upper East 100.0 97.7 84.1 21.5 72.9 85.0 51.9 91.1 99.1 99.1 99.1 99.1 Upper West 99.2 98.5 80.0 80.8 86.9 89.2 69.2 95.4 99.2 99.2 99.2 100.0 Education level None 99.3 99.1 71.9 41.6 68.9 88.4 69.6 93.0 99.0 99.4 98.5 97.9 Primary 100.0 98.9 81.7 49.3 68.5 92.9 78.4 95.3 98.5 99.3 98.3 96.8 Middle/JSS/JHS 99.6 99.3 87.2 56.0 75.5 94.6 85.3 95.8 98.3 99.7 98.4 97.2 Secondary 100.0 99.3 90.0 61.1 84.8 96.2 86.0 94.3 97.4 99.3 98.7 97.3 Technical 100.0 97.9 91.0 56.9 77.8 95.1 89.6 95.8 96.5 99.3 99.3 97.2 Higher 99.8 98.2 98.0 65.0 91.1 97.4 94.1 94.1 94.7 99.4 98.2 97.4 Asset quintile Lowest 99.3 98.7 70.1 43.2 65.9 88.2 67.6 92.9 99.0 99.5 98.0 97.7 Second 99.7 99.3 83.6 51.4 72.6 93.6 80.8 95.0 99.1 99.5 98.5 97.7 Middle 99.8 99.1 88.4 54.6 76.5 94.3 84.3 95.3 97.8 99.5 98.4 96.9 Fourth 99.8 98.8 90.4 59.1 82.1 95.6 89.8 96.4 96.6 99.4 98.8 97.6 Highest 100.0 99.3 95.1 63.4 86.7 97.4 91.1 94.3 96.6 99.5 98.4 96.3 Total 99.7 99.1 84.9 53.8 76.1 93.6 82.2 94.8 97.9 99.5 98.5 97.3 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Notes: Service providers are 1 = Health providers, 2 = Police, 3 = DOVVSU, 4 = Shelter, 5 = NGO, CSO or social worker, 6 = Lawyer or member of Court, 7 = Legal aid group, 8 = Queen mother, 9 = Traditional leaders, 10 = Religious leaders, 11 = Religious group, 12 = Community group 222 The quantitative survey showed that despite good knowledge about public services in general, and DOVVSU in particular, there was some room for improvement in terms of awareness about the existence of a law against domestic violence in Ghana. According to the GFLHS 2015, 75 per cent of respondents in urban areas and 61 per cent in rural areas knew there was a law against domestic violence. There was also a large variation across regions in terms of awareness about the Domestic Violence Act. For instance, only between 10.7 and 12.1 per cent of those in the Upper East, Upper West and Northern regions knew about it. 4.7.4. Levels of satisfaction about support services and institutions The information collected in the study has revealed that satisfaction with the public primary response institution against domestic violence, DOVVSU, was low in comparison with other means of support. However, very few respondents answered this question, as not many individuals used these services. Table 70 provides an overview of satisfaction levels with services, such as health providers, police and DOVVSU units, as well as with traditional avenues for conflict resolution, such as traditional and religious leaders and community groups. Table 70: Satisfaction levels with public services (number of respondents) Completely Somewhat Somewhat Completely N satisfied satisfied dissatisfied dissatisfied Health centre, hospital, 10 1 0 1 12 other health provider Police 60 17 11 17 105 DOVVSU 2 1 1 2 6 Traditional leaders 19 10 2 4 35 Religious leaders 20 4 3 2 29 Community group 9 2 3 2 16 Source: Ghana Family Life and Health Survey (GFLHS) 2015. The qualitative data revealed that the reasons for the lower levels of satisfaction with DOVVSU were mostly related to the direct and indirect costs of services, including legal services. What bothered people was not just the financial cost but also the cost of secondary victimisation due to lengthy procedures and additional payments: “At the police facility, they waste a lot of time and always make demand for money so I am really not satisfied with their service. I am however satisfied with help at health centres because they are efficient and sometimes even treat victims of domestic violence without referring it to the police to avoid prosecution” (IDI, opinion leader, Sagnerigu district, Northern region) 223 Families do not always have the funds to pay for the legal costs of prosecution, which puts them in a difficult situation: “The main challenge has to do with the finance. Some of the clients, especially in cases of domestic violence cannot afford the fees for their legal representation and often, we have to take up these cases pro bono.” (IDI, opinion leader, Gbolo, Sagnerigu district, Northern region) GFLHS 2015 respondents were asked whether they had to pay for services when approaching any relevant institution. The results showed that about one quarter of service users had to pay for them. Twenty per cent of respondents who used police services paid a fee, and about 6 per cent of respondents paid a bribe. This situation has prevented many from accessing the services they need, and has created considerable constraints to those who try to help and support women and men affected by domestic violence in Ghana. 224 5. Summary and discussion This study has provided an in-depth analysis of the attitudes to and incidence, determinants and consequences of domestic violence in Ghana, through the collection of a rich set of quantitative and qualitative data. This section summarises the main findings of the study in relation to each research question. 5.1. Incidence of domestic violence in Ghana  Over a quarter (27.7 per cent) of women and 20 per cent of men in Ghana experienced at least one type of domestic violence in the 12 months prior to the survey.  The most common form of domestic violence reported by women was economic violence (12.8 per cent), followed by social violence (11.6 per cent), psychological violence (9.3 per cent), physical violence (6.0 per cent) and sexual violence (2.5 per cent).  The incidence of domestic violence was generally lower among men than among women. The most common form of domestic violence experienced by men was psychological violence (7.9 per cent), followed by social violence (7.7 per cent), economic violence (7.3 per cent), physical violence (2.1 per cent) and sexual violence (1.4 per cent).  The analysis of the qualitative and quantitative data suggested that sexual violence was largely considered a private matter. The results above may, therefore, underestimate the true extent of this form of domestic violence.  The GFLHS 2015 is not comparable to prior surveys of domestic violence in Ghana because it collected data on broader and more in-depth definitions of domestic violence than other surveys. It is, however, possible to compare the incidence of physical domestic violence experienced by women and men in the 2008 DHS and in the 2015 GFLHS. The results show, when using comparable categories of domestic physical violence,33 that domestic physical violence among women decreased from 17.2 per cent in 2008 to 10.3 per cent in 2015. Domestic physical violence among men decreased from 12.7 per cent in 2008 to 11.2 per cent in 2015.  There was a dominant view that domestic violence was largely perpetrated by men against women. The idea of physical violence committed by women against men was found to be ‘funny’ and sometimes ‘unimaginable’. The qualitative and 33 Please note that the estimates below for the incidence of domestic physical violence in 2015 do not match the estimates provided in Table 26. This is because the DHS 2008 considers domestic violence to include some categories of perpetrators that are not part of the definition of domestic relation used in this study (or in the Domestic Violence Act 732). 225 quantitative data found, however, several instances of domestic violence committed by women against men. For instance, 12 per cent of physical domestic violence and 37 per cent of psychological domestic violence against male household heads was perpetrated by their spouse or (former) partner. Among the participants in the GFLHS 2015, 789 women and 563 men reported having committed at least one form of domestic violence, but these results must be interpreted with care due to potential under-reporting and reporting biases. 5.2. Attitudes towards domestic violence in Ghana  Respondents overall opposed all forms of domestic violence. However, there were important caveats as to what was perceived to constitute domestic violence. Respondents considered that some of the definitions of domestic violence used in the study (and in Act 732) were not violence but were, rather, accepted social norms, defence mechanisms or ways of resolving disputes. Only physical violence between partners was consistently perceived as violence. This is an important finding because it suggests that violence is experienced, conceived and articulated in different ways along a spectrum that locates some forms of violence (such as physical violence) as the most severe, and other forms of violence (for instance, controlling behaviour by women) as a way of managing unequal power in domestic relationships.  Psychological violence was generally perceived as a means of defence and a coping strategy – for women in particular – to ward off physical violence or to claim some autonomy or agency in contexts where they may not have the ability to exert much power beyond the use of words (insults), their body (withholding sex) or their housework (not cooking).  Physical violence was largely considered to be unacceptable across Ghana, but some acts of domestic physical violence were sometimes deemed acceptable, depending on the context. There was, in particular, acceptance of wife-beating as a punishment against disobedience and neglect of children: 11.2 per cent of men and 17.2 per cent of women mentioned that wife-beating was acceptable if the wife had disobeyed the husband, and 10.2 per cent of men and 16.5 per cent of women agreed that it was acceptable if the wife neglected the children.  Attitudes of support for wife-beating have, however, become less pervasive in Ghana since 2008: the proportion of women justifying wife-beating decreased from 36.6 per cent in 2008 to 23.1 per cent in 2015, while the proportion of men condoning wife-beating decreased from 21.8 per cent in 2008 to 13.8 per cent in 2015.  Physical violence against children was generally viewed as unacceptable, but was described in some discussions, particularly in the Northern and Eastern regions, as a form of parental education to ‘steer them in the right direction’. There has been, 226 however, a significant reduction in attitudes condoning physical violence against children: in 1998, 8 per cent of respondents thought it was never acceptable to beat a child (Coker-Appiah and Cusack, 1999), whereas in 2015, this number rose to 25.3 per cent among men and 18.8 per cent among women.  Women were on average more likely than men to find wife-beating acceptable: 23.1 per cent of women and 13.8 per cent of men condoned wife-beating. Furthermore, 65.3 per cent of women and 56.2 per cent of men fully agreed that women were to blame for rape if they wore revealing clothes. These findings are in line with previous literature and reflect three facts. First, that the above forms of domestic violence are still considered acceptable. Second, victims tend to internalise social norms once they are in abusive relationships. Third, even when not victims, women may internalise gender norms that hold them responsible for their experiences of gender-based violence, rather than placing that responsibility on men or on the broader system that reinforces gender inequalities. 5.3. Determinants of domestic violence in Ghana The study revealed that age and intergenerational factors were strong determinants of domestic violence in Ghana:  Young women and men (15–19 years old) were substantially more likely to have experienced domestic violence than other age groups. Women aged 15–19 years were between 1.3 and 4 times more likely to experience any form of domestic violence than, for instance, women aged 30–39 years. Differences in the incidence of domestic violence across age groups were similar for men, with the exception of sexual violence, which was more prevalent among men aged 20–24 years. This is in contrast with the 2008 DHS, which reported very similar incidence across age groups, but in line with the 1998 study by Coker-Appiah and Cusack (1999).  Exposure to violence in childhood was found to be strongly related to the likelihood of an individual being a victim of violence in adulthood. Women who witnessed any form of domestic violence as children were between 1.3 and 3 times more likely to report domestic violence than women who did not witness violence during their childhood. Men who witnessed social, psychological or economic violence in their homes as children were between 1.2 and 2.2 times more likely to have experienced these forms of violence in adulthood than other men. This is a common finding in domestic violence research, including the study by Tenkorang et al. (2013) on domestic violence in Ghana.  Exposure to violence in childhood was also found to be related to the likelihood of an individual being a perpetrator of violence in adulthood: about 28 per cent of women and men who admitted committing social violence witnessed social violence as children, compared to 14–15 per cent of those women and men who did not experience violence as children. Among men who witnessed sexual violence as children, almost 30 per cent committed sexual violence over the 12 months prior to the survey, compared to 2.6 per cent of men who did not experience this form of 227 domestic violence during their childhood. Across psychological, physical and economic violence, rates of perpetration are between 2 and 3 times higher for women and men who witnessed domestic violence as children than among individuals who did not experience domestic violence during their childhood.  Taken together, these findings suggest that Ghana may experience vicious cycles of domestic violence reinforced by the fact that: (i) young people are more at risk of domestic violence; (ii) experiencing domestic violence at a young age will considerably increase the odds of experiencing domestic violence later in life; and (iii) experiencing domestic violence at a young age will considerably increase the odds of the individual perpetrating violence later. Socio-economic factors were reported to be among the most important determinants of domestic violence in Ghana, albeit in complex ways, as follows:  Low asset levels were found to be strong determinants of psychological and economic violence among women, but high asset levels were associated with higher levels of social and sexual violence among men.  Tensions and fights over money or property were found to be leading determinants of all forms of domestic violence across all regions.  Women with no education were between 0.5 and 2.5 times less likely to experience domestic violence than women with some level of education. More educated men were around 3 times more likely to experience economic and sexual violence than less educated men. This result may be due to the fact that women and men with more education may also be more likely to report domestic violence, as emphasised in previous studies conducted internationally and in Ghana.  Factors such as poverty, unemployment and economic shocks and associated tensions were identified as important determinants of domestic violence. However, employment status was not found to be a particularly significant determinant of domestic violence, with some exceptions: women who were not working were 43 per cent more likely to have experienced economic violence than self-employed women, while men who were employed or not working were, respectively, 2 times and 2.5 times more likely to have experienced economic violence than self- employed men.  Sexual violence was found to be more pronounced among individuals with higher levels of education and socio-economic status, particularly men. The study did not find a clear relationship between marital status and domestic violence. However, the following findings are noteworthy:  Divorced, separated or widowed women were twice as likely to have experienced physical violence than women who were never married.  Married women or divorced, separated or widowed women were, respectively, 1.8 times and 2 times more likely to have experienced economic violence than non- married women.  Married or divorced, separated or widowed men were, respectively, 6 times and almost 8 times more likely to have experienced sexual violence than non-married 228 men, and almost 3 times and 1.7 times more likely to have experienced social violence than men who were never married.  Women in polygamous relationships were almost twice as likely to experience psychological violence than women who were never married. These women were also 3.5 times more likely to have experienced physical violence and 4.2 times more likely to have experienced economic violence. Men in a polygamous relationship were also more likely than men in monogamous relationships to experience sexual or social violence.  Domestic violence sometimes followed women’s refusal to have sex with their husbands, but this form of marital dispute was often reported to be used as a way to retain some degree of power and autonomy in the face of other forms of violence, especially economic violence. There was no clear statistical relationship between geographical factors (urban/rural and regional) and the incidence of domestic violence once the statistical analysis controlled for other factors. This was probably because domestic violence was not necessarily related to geographical characteristics per se but, rather, to the incidence of lower levels of development in certain regions. There was an association between the incidence of domestic violence and perceptions of gender roles, patriarchal norms and masculinity. The results were particularly striking in the qualitative data, which showed domestic violence to be closely determined by strong gender norms that perpetuate gender inequalities. The survey also showed the following:  Women who exhibited higher levels of acceptance of wife-beating were 14 per cent more likely to have experienced economic violence than other women. This is, however, not a causal effect and may well suggest that victims of violence internalise and accept gender norms that perpetuate or justify domestic violence.  Women who expressed a higher acceptance of women’s sexual autonomy were 16 per cent more likely to have experienced psychological or social violence, 17 per cent more likely to have experienced physical violence, and 11 per cent more likely to have experienced economic violence than women with a lower acceptance of women’s sexual autonomy. This finding suggests that women who are more likely to challenge accepted gender norms may be at risk of domestic violence because they may be punished for not complying with established norms of behaviour.  Men who expressed stronger patriarchal gender norms were 26 per cent less likely to report social violence than men who did not accept such norms.  Men who displayed higher levels of tolerance towards wife-beating were 21 per cent, 47 per cent and 30 per cent more likely to have experienced psychological, physical or economic violence, respectively, than men who reported lower levels of tolerance towards wife-beating. This result may potentially reflect retaliation by women who were beaten by their husbands.  Men who exercised more decision-making power within their households were 34 per cent less likely to experience social violence than men who exercised lower levels of decision-making within the household. 229 The study found that exposure to criminal forms of violence in the community (theft or robbery, vandalism, arson, kidnapping or abduction, extortion, bullying, political violence, highway attacks, harassment, rape or defilement) was significantly associated with higher incidence of domestic violence among men and women.  The odds of victimisation among women increased by 77 per cent for psychological violence, 49 per cent for physical violence and 57 per cent for economic violence when the respondent lived in a community that experienced at least one form of violence. The effect was stronger for sexual violence and social violence, for which the odds of having experienced this violence were 3.6 times and 1.9 times higher, respectively, for women exposed to violence in their communities than for women in communities that did not experience other forms of violence.  Men who reported some violence in their communities were between 2 and 3 times more likely than men living in more peaceful communities to report having experienced any form of domestic violence.  These findings emphasise the importance of taking into consideration factors beyond the individual and the household when studying the determinants of domestic violence. Overall, the determinants of domestic violence varied considerably among types of violence, emphasising the need for domestic violence legislation in Ghana to continue to distinguish between different forms of violence (as in Act 732). Only young age, exposure to domestic violence as a child and high levels of violence in the community were common determinants of domestic violence across all types. 5.4. Consequences of domestic violence in Ghana  The GFLHS 2015 revealed adverse consequences of domestic violence on physical health, including injuries and illness: 43.8 per cent of women who had been physically assaulted declared having been ill in the 30 days prior to the survey – a rate that was a third higher than among women who had not been physically assaulted (31.2 per cent) – and 42.3 per cent of women who had experienced psychological violence reported ill health, compared to 30.9 per cent of women who had not experienced this form of domestic violence. The effects of domestic violence on physical health were observed almost exclusively among women.  The results showed a very strong correlation between exposure to domestic violence and mental health, particularly among women: women who reported serious mental illness were 26.7, 24.2, 21.9, 19.3 and 26.3 per cent more likely to have experienced, respectively, psychological, physical, sexual, social or economic violence than women who did not report serious mental illness. This finding is in line with previous studies in Ghana (Amoakohene, 2004; Adu-Gyamfi, 2014). Only physical violence had an effect on men’s mental health status. 230  Domestic violence was found to have adverse qualitative consequences on daily life in terms of the victim’s ability to work, go to school or do domestic work, their ability to concentrate on daily activities, levels of confidence and feelings of living in fear, although few respondents answered this question. The effects were larger for women than for men, and for physical violence than for other types of violence.  Domestic violence was found to have adverse qualitative effects on child development, educational opportunities among children and on youth deviant behaviour, as a result of broken marriages and the resulting economic burden, and the direct exposure of children to violence within the household – which may result in those children being more likely to experience or perpetrate domestic violence later in life. 5.5. Use of and access to victims’ support services and institutions Just over one in every three women and men who experienced domestic violence approached a person or formal or informal organisation for help or support. The majority of respondents who sought help first approached family members or a friend (66.2 per cent of men and 77.6 per cent of women), whereas 14.6 per cent of men and 9.0 per cent of women first contacted the police for help. These findings suggest that individuals who experience domestic violence only resort to external help in extreme circumstances, and that domestic violence may still be considered largely a private family matter. Reasons for seeking help included not being able to endure the abuse further (53.3 per cent of men and 58.3 per cent of women), encouragement by friends and family (28.5 per cent of men and 24.5 per cent of women), and the extent of injuries sustained (8.0 per cent of men and 7.5 per cent of women). Reasons for not seeking help included predominantly the fact that individuals considered the violence as normal or not serious (22.7 per cent of men and 17 per cent of women), and not knowing who to ask (10.7 per cent of men and 16.4 per cent of women). Distrust of, and poor experiences with, state authorities and public services contributed to not seeking help. Factors for these low levels of trust included the costs of the services, lengthy or ineffective delays in the handling of cases, and corruption. In addition, the Ghana Police Service had only limited logistical support to encourage the proper functioning of DOVVSU:  In nine out of ten districts sampled, DOVVSU units or available DOVVSU staff did not have appropriate separate offices or offer victims privacy when reporting domestic violence. The only exception was Obuasi in the Ashanti region.  Police personnel had not received adequate specialised training to provide specific help to the victims of violence.  Observations revealed that police personnel were motivated and professional despite workloads and lack of resources. 231 In addition to underlying deficiencies in the justice and victims’ support systems, preferences for traditional over formal legal ways of addressing domestic violence were also explained by structural inequalities. Notably, poorer victims and those worried about personal social stigmatisation and reputational loss for the whole family preferred not to access formal justice processes and resorted to informal processes of mediation. At times, this was imposed by the social and economic needs of the family or community, and by prevalent gender norms. There is room for improvement in terms of raising awareness about the existence of the Domestic Violence Act and DOVVSU units among Ghana’s general population: 75 per cent of respondents in urban areas and 61 per cent of respondents in rural areas knew there is a law against domestic violence in Ghana. There was also a large knowledge variation across regions, with 10.7 and 12.1 per cent of respondents in the Upper and Northern regions, respectively, knowing about the law. These were also the regions where the fewest respondents knew whether there was a DOVVSU unit at their nearest police station. Satisfaction with the support provided by DOVVSU was low in comparison with other means of support, but few people answered this question in the survey. Reasons for the lower levels of satisfaction with the police and DOVVSU mostly related to the direct and indirect costs of services, including legal services. 232 6. Way forward and policy recommendations The key purpose of this study was to collect and analyse new evidence about the incidence levels, determinants and consequences of and attitudes towards domestic violence in Ghana in 2015 that can be used by the Government of Ghana, CSOs and development partners to assess current patterns of domestic violence across the country, and to advance legal, policy and programmatic interventions. To this purpose, this section reflects on how existing policies and interventions, including the National Domestic Violence Policy and Plan of Action to Implement the Domestic Violence Act, 2007 (Act 732) (2009–2019) could be reviewed to better support victims of domestic violence and prevent domestic violence in Ghana. The study found that domestic violence directly affects 27.7 per cent of women and 20 per cent of men in Ghana. In 2015, these women and men experienced at least one form of violence within domestic relations – either social, physical, sexual, psychological or economic violence. The causes underlying these levels of domestic violence are complex. The study found that no single factor or process explained domestic violence victimisation or perpetration – these varied considerably across the different types of domestic violence and depended on (i) personal and socio-economic characteristics of victims or perpetrators; (ii) specific relationship dynamics within families; (iii) norms and structures that operate at the level of the household, the community and the wider society; and (iv) national- and global-level factors that shape prevailing gender and social norms, and access to resources and opportunities by different individuals and social groups. To address these complex, intersecting factors, policy responses must operate at several levels and in several sectors. This section discusses which potential interventions may work – and how existing interventions and programmes may be strengthened – to reduce the incidence and impact of domestic violence in Ghana. The section is organised around the main sections of the analysis, and discusses potential recommendations with respect to the incidence of domestic violence (subsection 6.1), attitudes towards domestic violence (subsection 6.2), determinants of domestic violence (subsection 6.3), consequences of domestic violence (subsection 6.4) and use of and access to support services and institutions (subsection 6.5). As with other studies, this study uncovered intriguing findings that could not easily be explained within the scope and time frame of the project, and generated several additional questions. This section concludes, therefore, with recommendations for future research work on domestic violence in Ghana. These recommendations focus on four important areas that were outside the scope of this study: (i) the incidence, causes and consequences of non-domestic forms of violence, which were found to affect considerable numbers of women and men in Ghana; (ii) the analysis of causal effects with respect to the causes and determinants of domestic violence; (iii) attitudes towards and the incidence, causes and consequences of sexual violence, particularly among men; and (iv) attitudes towards and 233 the incidence, causes and consequences of domestic violence among children, which was uncovered as an area of concern for future research and policy work. 6.1. Recommendations with respect to the incidence of domestic violence in Ghana 1. Amendment of the Domestic Violence Act, 2007 (Act 732) to include social abuse as a specific dimension of domestic violence in the law. The study has shown that social violence is the second most predominant form of domestic violence against men and women. It is therefore recommended that the Domestic Violence Act, 2007 (Act 732) is amended to specifically include this category of abuse within the categories of prohibited forms of domestic violence in Section 1, in addition to covering physical abuse, sexual abuse, economic abuse and emotional abuse, as well as harassment. Currently under Section 1(c) of the Domestic Violence Act it is indicated that domestic violence includes “harassment including sexual harassment and intimidation by inducing fear in another person”. In Section 42 of the Domestic Violence Act ‘harassment’ is defined as including acts of intimidation, threatening acts and some aspects of controlling behaviour. However, not all the forms of social violence found in the course of the study are set out or obvious in this definition. It is, therefore, important to specifically include social violence in the definition of violence in Section 1. The working definition of social violence for the study was “acts of controlling behaviour experienced by men and women, such as preventing someone from seeing friends or family of birth; stopping someone from leaving the house; insisting to know where someone is at all times; stalking; spreading false information, videos or photos without permission; or forcing women to have an abortion”. This definition could be adopted in any further amendments of the law. 2. Extend gender-sensitive legislation and awareness and advocacy programmes to bring attention to non-physical forms of domestic violence. The study has shown that people’s lived experiences of domestic violence are sometimes at odds with the way in which domestic violence is conceptualised by policymakers and researchers. This may undermine effective, tailored policy responses at a practical level. For example, many acts defined as psychological or economic violence in the Domestic Violence Act were not perceived as violence by the respondents. These perceptions may result in a mismatch between legislation and practice, whereby victims of these forms of violence do not feel they ought to seek support, leading to fewer meaningful policy responses to less overt and less physical forms of violence. Addressing these issues will involve taking a more comprehensive approach to gender-responsive legislation in Ghana through gender- sensitive legislative frameworks that may enable state institutions to develop responsive policies and transformative protocols to challenge all forms of (physical and non-physical) domestic violence. These measures should be combined with awareness-raising campaigns targeting the public, media, police, health, legislators and other decision- makers and stakeholders about the importance of non-physical forms of domestic violence. 234 3. Improve support available to victims of psychological domestic violence. The study found that psychological violence was the most common type of violence experienced by men (affecting 7.9 per cent of men) and the third most common form of domestic violence experienced by women in Ghana (affecting 9.3 per cent of women). Psychological violence was also found to be associated with a high risk of women suffering from serious mental illness. At the same time, psychological violence was generally perceived as a means of defence and a coping strategy – for women in particular – to ward off physical violence or to claim some autonomy. These findings required a multi-component approach including awareness-raising campaigns about the importance and prevalence of psychological violence, further recognition in support services about this form of violence, strengthening the skills of mental health professionals to identify and offer protection to those at risk of this form of abuse (such as psychotherapy and counselling services), and further cooperation between mental health professionals, NGOs and police units involved in domestic violence prevention. 4. Further attention is needed to address the widespread prevalence of economic violence. Economic violence was defined as the denial of household expenses money (chop money) even if enough financial means are available; unsolicited taking of money; control of belongings and spending decisions; damage to or destruction of someone’s property; denial of the right to work; forcing someone to work against their will; or denial of food and other basic needs. Economic violence was the most prevalent form of domestic violence against women (affecting 12.8 per cent of women) and the third most common form of domestic violence against men (affecting 7.3 per cent of men). Economic violence was particularly prevalent among married women, and less wealthy men and women living in rural areas, and was found to have severe consequences for the physical and mental health status of women. These findings call for a stronger coordination of the laws and policies relating to abuse of individual property rights within domestic relations, and to economic maintenance provision for married, separated or divorced individuals. There are provisions on property rights of spouses in Article 22 of the 1992 Constitution and the Matrimonial Causes Act, 1971 (Act 367), and, apart from the Domestic Violence Act, there are provisions on economic violence in the Criminal Offences Act, 1960 (Act 29). There will also be a need to better coordinate domestic violence support services with ongoing interventions aimed at improving the economic conditions of women (such as microfinance and cash transfer programmes). Better coordination may ensure a reduction of the economic dependence of women on men – a factor found to be central to explaining the prevalence and consequences of domestic economic violence. 5. Extend education, awareness and advocacy programmes to bring attention to domestic violence against men. The study showed that, although women are more at risk of experiencing domestic violence, a substantial number of men are also at risk of violence within domestic relations. Addressing this issue will require bringing attention to forms of domestic violence experienced by men, and targeting awareness-raising campaigns towards the value of understanding how men also struggle with harmful masculinities that might be imposed on them by other men and women. There is also a need to shift 235 social norms about the acceptance of different forms of violence among women – and men. Further training of police and other support services may also be required, as men find it even more difficult than women to report cases of domestic violence (when the response may be disbelief, or ‘giggling’ as some of the qualitative transcripts show). 6. Further attention is needed with respect to child protection. Some forms of physical violence against children are sometimes perceived as a form of parenting. Non-biological children can be at particular risk of domestic violence, particularly when the mother is in a vulnerable economic position. This calls for a policy decision to ensure protection for children, especially vulnerable non-biological children. It may also call for health and education workers who engage with children most frequently to receive specialised awareness training on indicators of domestic violence and the ways available to ensure that children are protected. Interventions aimed at schools, teachers, children and parents have reduced violence against children in other countries (see Fulu et al., 2014). Examples of such interventions include work in schools and with teachers to raise their awareness about violence-affected children, and improve their skills to behave in non-violent ways; work with schools and children through curriculum-based awareness-raising interventions; and work with parents to build skills that involve non-violent means of parenting. This is, however, an area not explored in detail in the study (the study did not involve children as participants) and requires further research. 6.2. Recommendations with respect to attitudes towards domestic violence in Ghana 1. Need for new sensitisation programmes about harmful social norms that drive domestic violence. The study found that attitudes against any form of domestic violence were strong and widespread, but many of the definitions of domestic violence used in the study (and in Act 732) were seen by respondents not as violence but, rather, as accepted social norms, defence mechanisms or ways of resolving disputes. Only physical violence between partners was consistently perceived as violence. Psychological violence was generally perceived as a means of defence and a coping strategy – for women in particular. Other acts defined as domestic violence in the 2007 Domestic Violence Act – such as controlling behaviour, verbal abuse or men making sexual remarks to women – were sometimes deemed acceptable. In particular, wife-beating was accepted as a valid punishment against disobedience and neglect of children, and rape was considered by the majority of respondents to be the fault of the woman if she wore revealing clothes. There was also limited discussion about sexual violence in all focus group discussions and individual interviews, reflecting the fact that many acts defined as domestic sexual violence may either be viewed as private or not perceived as violence but, rather, as ‘normal’ behaviour. These findings suggest that advocacy and awareness-raising campaigns (perhaps through the use of the media) still have a strong role to play in working to change these widely accepted norms and behaviours that justify the violence associated with them. 236 2. Need for new sensitisation programmes to reduce acceptance of domestic violence, particularly among women and girls. The study found that women were on average more likely than men to find some forms of domestic violence more acceptable. These included wife-beating, appropriateness of sexual remarks against women, and rape under some circumstances (if women wore revealing clothes). These findings reflect two important facts that shape the prevalence of domestic violence in Ghana. First, women who are victims of domestic violence may have internalised the acceptability of domestic violence once they are in abusive domestic relations. Second, even when not victims, women may internalise gender norms that hold women responsible for their experiences of gender- based violence, rather than placing that responsibility on men or on the broader system that reinforces gender inequalities. The prevalence of harmful gender norms – such as discouraging women from engaging in paid labour, expectations that men are the breadwinner, forcing women to obey men or blaming and shaming women for violence inflicted on them – was found throughout the study as central to explaining the prevalence, causes and consequences of domestic violence in Ghana. For instance, the regression analyses conducted in the study showed that women who expressed a higher acceptance of women’s sexual autonomy were at greater risk of experiencing psychological, social, physical or economic violence. The research team interpreted this finding as indicating that women who challenge established gender norms may be at risk of domestic violence because they may be punished for their non-compliance with established norms of behaviour. These findings call for multi-component interventions and strong cooperation between state agencies and civil society to change the prevalence of harmful gender norms that perpetuate gender inequalities and condone several forms of domestic violence. Examples of interventions include (i) economic interventions to reduce the economic dependence of women on men (such as microfinance, cash-for-work, or cash transfer programmes); (ii) interventions that work with men and boys to reduce the prevalence of harmful masculinities and harness them to bring about positive change; (iii) interventions that work with girls and women to improve their confidence and sense of worth (perhaps also including psychotherapy); (iv) interventions at the level of family relationships (such as counselling); and (v) group education campaigns and mobilisation activities at the level of communities (involving men and women, boys and girls) to raise awareness about the prevalence of these gender norms and find collective ways to reduce or abandon such norms altogether. 3. Further awareness-raising is needed about physical violence against men and children. The qualitative data found a close relationship between participants’ knowledge of legal rights and acts, such as the Domestic Violence Act, and their perception that physical forms of domestic violence, especially against women, were unacceptable. Physical violence was generally interpreted as committed by men against women. Discussions about perceptions of physical violence committed by women against men revealed that for many people this was ‘funny’ and sometimes ‘unimaginable’. Physical violence against children, while often perceived as unacceptable by both men and women, was described in some focus group discussions, particularly in the Northern and Eastern regions, as a form of education. This calls for more education and awareness-raising campaigns about the fact that physical 237 violence against anyone, including children, is unacceptable. Interventions discussed above with respect to the prevalence of sexual violence among men and violence against children would be useful to increase awareness about the prevalence of attitudes that condone physical violence against men and children. The Northern and Eastern regions in particular should be targeted for further interventions in this latter area. 4. Need to target further sensitisation programmes against domestic violence in less economically developed areas. The study has shown that attitudes against domestic violence were consistently less pronounced in rural and in the less economically developed northern areas, with individuals in the lowest asset quintiles and with no education consistently being more likely to condone several forms of domestic violence than other socio-economic groups. This finding indicates that sensitisation programmes should target messages in less developed areas and among populations with lower economic status to enable a change in attitudes and norms that reinforce domestic violence. The findings also call for increased awareness-raising in rural communities to ensure an elimination of the acceptance of abusive behaviour in inter-personal relations and within the domestic setting. This is particularly important given the intergenerational dynamics of domestic violence found in this study, and the value of stopping the vicious cycle that adversely affects people who experience domestic violence at a young age. 6.3. Recommendations with respect to the determinants of domestic violence in Ghana 1. Need to target interventions towards adolescents and young adults. The findings indicate strong effects of age and intergenerational factors on the incidence of domestic violence in Ghana. Young women and men (15–19 years old) were substantially more at risk of domestic violence than other age groups. Exposure to violence in childhood is also strongly related to the likelihood of being a victim and perpetrator of violence in adulthood. These findings suggest that Ghana may experience vicious cycles of domestic violence reinforced by the fact that young people are more exposed to domestic violence and that experiencing domestic violence at young ages considerably increases the odds of further exposure to and perpetration of violence. This calls for targeted interventions, including sensitisation and therapy for individuals who were exposed to violence at a young age, to stop the cycle of violence, and coordination with the interventions included above in Section 6.1 with respect to the incidence of violence against children. 2. Further attention is needed on polygamous relationships. Even though the study found no clear relationship between marital status and domestic violence, it found that women in a polygamous relationship were 1.9, 3.5 and 4.3 times more likely to experience psychological, physical or economic violence, respectively, than never-married women. Men in polygamous relations were also the most likely to experience sexual or social domestic violence. This finding emphasises the need to strengthen legislative systems against polygamy, which violates Article 17 of Ghana’s 1992 Constitution, or to ensure that 238 individuals in polygamous relations are protected under legislation against domestic violence. 3. Reinforcement of laws and practices about economic maintenance. The study found that increases in domestic violence were associated with (women’s) refusal to have sex. This refusal was often used as retaliation against other forms of violence, especially economic violence. In addition, shirking of economic responsibilities, especially by men not providing chop money, not paying children’s school fees or not paying other bills, was seen as an important determinant of domestic violence. In addition to provisions against economic violence in the Domestic Violence Act, the Children’s Act, 1998 (Act 560) has provisions that can be used to hold parents responsible to provide for their children. It is, therefore, important to enforce these laws alongside each other. A recommendation is also made to support the proposal by the MoGCSP to set up a Child Support Unit to hold non- custodial parents responsible for contributing to the maintenance of their children, as set out in the Justice for Children Policy. On the issue of the insistence on the provision of daily chop money by men for the upkeep of homes, there is a need to work on changing gender norms, as discussed above, that place women as economically dependent on men, and expect men to be the sole provider for the family. There may also be the need to explore whether the expectation for men to provide chop money is simply a means by which men are made to contribute to housekeeping costs. Encouraging community dialogue (between men and women) that promotes women’s economic participation, and that values women’s unpaid labour in the household, might also promote a greater awareness of the important roles that women can play in the household, alongside men. 4. Reinforcement of property laws. Fights over money, assets or property were found to be prominent determinants of domestic violence across all regions. Economic tensions were mentioned to lead to multiple forms of domestic violence among adults – including between partners, siblings and in-laws. There should be a concerted effort to ensure that laws in place that protect the property rights of spouses on the death of a spouse – such as the Intestate Succession Act, 1985 (PNDCL 111) – are disseminated and enforced. With regard to the property rights of spouses, the Property Rights of Spouses Bill, 2013 that seeks to actualise Article 22 of Ghana’s 1992 Constitution, which is before Parliament, should be passed into law as a matter of urgency to protect the property rights of married people at the dissolution of their marriage or the separation of cohabiting people. The police should be made aware of these laws to ensure that they are enforced alongside the Domestic Violence Act. 5. Need to continue sensitisation programmes about alcoholism. Alcoholism was frequently mentioned as one particularly problematic form of harmful behaviour associated with domestic violence. Some domestic conflicts were shown to be triggered following excessive alcohol intake. There is a need for policymakers to continue to educate people about the harms caused by alcoholism and provide support for the treatment of alcoholism. There is also a need to advocate that the use of alcohol is not an acceptable 239 excuse or justification for acts of domestic violence and will not stop prosecution for violations committed. 6. Reinforce legislation and sensitisation interventions against early child marriage. Early child marriages can be considered a form of domestic violence perpetrated by families against girls with the consent of both parents. The qualitative data showed evidence of this practice in some northern regions. Even though the Children’s Act prohibits forced and early marriage, it is a practice that goes on in some parts of Ghana. A recommendation is made that the MoGCSP continues undertaking research into the issue and formulates appropriate policies to address it, including prohibiting it as well as including it in the Domestic Violence Act. 7. Need to look more closely at the association between domestic violence and community violence. The results of the study showed a strong association between criminal violence in the community and incidence of domestic violence. Men who reported violence in their communities were two to three times more likely than others to report any form of domestic violence. The odds of women experiencing any form of domestic violence in communities that experienced other forms of criminal violence increased by 77 per cent for psychological violence, 49 per cent for physical violence and 57 per cent for economic violence. The effect was even stronger for sexual violence and social violence, for which the odds of domestic violence were 3.6 times and 1.9 times higher, respectively, for women living in communities that experienced other forms of violence, in comparison to women living in other communities. The qualitative data emphasised in addition that in the northern regions of Ghana chieftaincy disputes can erupt into community-wide violence. This finding calls for a national policy to address community violence generally, and violence associated with chieftaincy and other traditional forms of leadership in specific places, since this may result in increased incidence of domestic violence. This requires coordination between different parts of the Ghanaian police dealing with different forms of violence across communities. 6.4. Recommendations with respect to the consequences of domestic violence in Ghana 1. Strengthen the involvement of medical practitioners in identifying victims of domestic violence. The study showed that domestic violence had several adverse consequences for the physical health of women. These included injuries (particularly because of physical violence), increased blood pressure and general illnesses. This finding suggests the need for further cooperation and integration of police and medical practitioners in the identification and reporting of victims of domestic violence, since the Domestic Violence Act allows medical practitioners to lodge complaints of domestic violence on behalf of victims. There is also the need for sensitisation programmes to be undertaken with medical practitioners. 240 2. Strengthen mental health policy in relation to domestic violence. The results showed a very strong correlation between exposure to domestic violence and mental health, particularly for women, and between domestic violence and the inability to perform daily tasks, feeling depressed and being unable to cope on a daily basis. This finding suggests that stronger linkages should be forged between domestic violence and mental health policies. It also calls for stronger and more effective inter-agency collaboration to address the incidence and harmful consequences of domestic violence. This may include targeted interventions such as sensitisation and therapy for individuals who experienced domestic violence, including those who were exposed to violence at a young age. It should also include supporting the use of psychologists under the Domestic Violence Act and making the cost of accessing their services affordable to victims and perpetrators. 3. Strengthen child protection policy in relation to domestic violence. The study found that domestic violence adversely affected child development, educational opportunities among children and youth behaviour. Domestic violence was also found to have negative impacts on the education and economic potential of children. This finding suggests that the MoGCSP must intensify its work around child protection. It suggests also the need for stronger and more effective inter-agency linkages and strategies to address child protection more effectively across its various dimensions. There is a particular need to incorporate information about the adverse consequences of domestic violence in education, awareness- raising and advocacy programmes, and in work with schools and parents, as effects of domestic violence go beyond the victim (for instance, on children and on family stability) and are not widely acknowledged. 6.5. Recommendations with respect to the use of and access to victims’ support services and institutions in Ghana 1. Reinforce sensitisation programmes about the need to report all forms of domestic violence. The findings indicated that only about one third of women and men who experienced at least one form of domestic violence in Ghana in 2015 tried to approach a person or organisation for help or support. Reasons for seeking help include not being able to endure the abuse further, encouragement by friends and family, and the extent of injuries sustained. Individuals exposed to domestic violence seemed to only resort to external help under more extreme circumstances, and the majority of respondents who sought help approached family members or a friend first. Reasons for not seeking help included considering the violence ‘normal’, not knowing who to ask, lack of trust, and feelings of embarrassment and shame. These findings suggest that domestic violence is still considered largely a private matter. There is a need for more education and sensitisation to make people aware that domestic violence is an offence, and not a private matter, and that this applies to all forms of violence under Act 732. People should be encouraged to report the occurrence of all forms of domestic violence and should receive adequate support when turning to formal institutions. 241 2. Improve awareness of the 2007 Domestic Violence Act and available services, particularly in rural and northern areas. Awareness about domestic violence has improved since previous studies in Ghana. However, 25 per cent of respondents in urban areas and 39 per cent of respondents in rural areas were not aware of the law against domestic violence in Ghana. Television and radio were found to be good means of transmission of information in urban areas (48.3 per cent of urban people were made aware of the Act via television, and 67 per cent via radio). These means were less effective in rural areas (19.3 per cent of rural respondents knew about the Act via television, and 54.6 per cent via radio). Building further collaboration between DOVVSU personnel and specialised CSOs, public units, including the Information Services Department and the National Commission for Civic Education, media and resource persons in local communities may strengthen institutional responses to domestic violence. There also seems to be a large variation across regions in terms of awareness about the law. For instance, 66 per cent of respondents overall knew that there is a law, in contrast to 10.7–12.1 per cent of respondents in the Upper East, Upper West and Northern regions. In these regions, very few respondents were also aware of the presence of a DOVVSU unit. Further efforts are needed to improve awareness of existing legislation and support channels in these areas. 3. Increase the number, spread, awareness and effectiveness of DOVVSU units. The study has shown that there is a need for more service units. Only a quarter of survey respondents in rural areas and 36 per cent in urban areas mentioned that the nearest police station had a DOVVSU unit. More DOVVSU units across the country would also reduce travel burdens for victims of domestic violence. It is, however, important to note that part of the reason for the low usage and knowledge of DOVSSU had also to do with people perceiving traditional mediation as more acceptable, cheaper and less likely to result in social shaming. The NPPOA has made provisions for coordination by the DV Secretariat to ensure that various actors, including CSOs, traditional authorities and others involved in addressing domestic violence issues, play a role at the community level and report to the regional and national levels, and there is the need for the Domestic Violence Management Board to ensure that this coordination system is actualised. It is important that formal state institutions collaborate with more informal community actors in settling disputes, documenting settlements and referring matters where necessary – and also in ensuring that different actors do not further reinforce existing social norms and behaviours that restrict action against and reporting of domestic violence. There is, therefore, an urgent need to ensure that the implementing regulations (legislative instruments) that will roll out the Domestic Violence Act are enacted as soon as possible to implement the Act further, including how to apply for various protection orders, qualifications required for providing psychological services, minimum standards for setting up shelters and developing training manuals, and how to access the Domestic Violence Victim Support Fund. 4. Increase the capacity of service providers to ensure privacy in units and provide additional support services, such as counselling and shelters. People who experience domestic violence may have immediate needs beyond medical care, such as access to therapeutic services, accommodation, legal aid and economic opportunities. Psychological 242 assistance is also often needed by victims of domestic violence who struggle to address feelings of shame and guilt. The Domestic Violence Act makes provision for these, and the regulations of the Act are further expected to elaborate on how to access these services. It is important that the regulations make clear the need for helpline support and information, shelter/refuge and safe accommodation, short- and long-term psychological counselling, legal advice, advocacy and outreach services. 5. Build the response capacity of specialised personnel. There is a need to update and improve the training of police officers to support women and men, boys and girls, as recommended in the NPPOA. The study found that the staff involved in domestic violence support services were highly motivated but lacking in wider institutional support. Additional training and support should enable them to better support the needs of victims of domestic violence. Consistent training should be encouraged and made part of the short- and long-term budget of the MoGCSP. Training should also include awareness about current research findings on the prevalence and gendered dimension of violence globally, regionally and locally in Ghana. This will encourage the design of appropriate tools to help specialised support units detect domestic violence at the first contact with women and men who ask for help, as either victims or perpetrators. It would also help support units to approach domestic violence respectfully and sensitively. 6. Improve trust in public services that support victims of domestic violence. Distrust of health care, police and legal services poses a challenge to support services across all regions of the country. While state institutions may be providing resources to effectively implement the Domestic Violence Act, they need to engage more directly with how resources are allocated at a local level and improve the public’s trust in the way in which domestic violence cases are handled and processed. At a practical level, a free telephone helpline for people to call in and report cases might be useful. The government could also consider setting up incentive programmes to encourage police and health care workers to report cases of domestic violence and follow appropriate procedures. The enactment of the regulations of the Domestic Violence Act will also strengthen the work of public services that support victims of domestic violence. 7. Improve inter-agency collaboration. The study observed that district hospitals do not have specialised services, such as emergency response to individuals who experience domestic violence, or priority responses for victims of abuse – unless victims are brought in as normal emergency cases. The MoGCSP should facilitate the development of inter- agency collaboration between service providers to address this gap so that there is better collaboration between medical facilities, the police, the judiciary and other institutions mentioned in the NPPOA. The MoGCSP should also collaborate with the Ministry of Health to ensure that special spaces are created in district and regional hospitals for victims of domestic violence. 243 6.6. Recommendations for future research on domestic violence in Ghana 1. Explore further the incidence, determinants and consequences of non-domestic forms of domestic violence in Ghana. The main aim of this study was to document and analyse the underlying attitudes towards and incidence, norms, determinants and consequences of domestic violence in Ghana – i.e. violence that takes place within domestic relations as defined by the 2007 Domestic Violence Act. The GFLHS 2015 collected, however, a substantial amount of information on social, physical, sexual, psychological and economic violence that takes place in non-domestic settings, such as the workplace and the community. The study found that a large percentage of women and men experienced non- domestic forms of violence. This is an interesting finding that deserves further exploration for two main reasons. First, the large incidence of forms of gender-based violence outside domestic relations suggests that harmful gender norms that condone the use of domestic violence may also be highly prevalent in the wider society. Therefore, policies that address violence within domestic relations may miss out vulnerable men and women who continue to experience forms of violence outside domestic relations. Second, the persistence of gender-based violence outside domestic settings suggests the widespread acceptance of gender norms that may contribute to the entrenchment and acceptance of forms of domestic violence as ‘normal’. The findings of this study suggest an urgent need to explore further the incidence, determinants and consequences of gender-based violence outside domestic settings in Ghana – and the data collected in this study could provide the baseline for such future research. 2. Further research on establishing causality in the determinants and consequences of domestic violence in Ghana. The multivariate regressions discussed in Section 4 of this report allowed this study to provide important information about the determinants and consequences of domestic violence in Ghana that took into consideration potential correlations between individual variables. This is important because policy action based on simple cross-tabulations may be misleading, as simple correlations may miss important interactions between the different complex factors associated with the incidence and consequences of domestic violence among some groups of people or some regions. However, the multivariate regressions analysed in this study are only a small step towards identifying precisely the determinants and consequences of domestic violence. The multivariate regressions produced in the study provided statistical relationships that take into account complex interactions between variables but do not allow the identification of precise causal relations. Establishing the direction of causality among the various factors that shape domestic violence and among the consequences of domestic violence requires the use of sophisticated econometric methods that were outside the scope of this study due to budget and time restrictions. One example regards findings about the role of education as a determinant of domestic violence. The study found that the risks of social or sexual violence were higher for men with higher education levels. This interesting result could be due to a variety of causes. For instance, it is possible that the risk of domestic violence increases for men with higher 244 education levels because they may have more confidence to report domestic violence. In this case, domestic violence is higher among this group not due to education per se but due to other social factors that jointly explain education levels and reporting domestic violence (for instance, more self-confidence). The analysis provided in this report is not sufficient to establish the causal reasons that explain the overall finding. Another example is the relationship reported in the study between violence at the community level and incidence of domestic violence. The analysis conducted in the study revealed a strong and statistically significant relationship between these forms of violence but was not able to identify the direction of causality. It is possible that communities that experience high levels of community violence may also experience high levels of domestic violence because violence becomes an accepted norm of behaviour. However, it is also possible that experiencing domestic violence may predispose individuals to other forms of violence. These two examples illustrate two statistical challenges – called, respectively, omitted variable biases and reverse causality – that can only be solved with more sophisticated quasi-experimental techniques. Efforts should be made to advance causal research on the determinants and consequences of domestic violence, as this would uncover more precise entry points for interventions aimed at preventing and remedying domestic violence and their rigorous evaluation. The datasets collected in this study can be used for this purpose in future research. 3. Further research on the incidence, causes and consequences of and attitudes towards sexual violence, particularly against men. This study found that domestic sexual violence exhibited patterns that were different from all other forms of domestic violence. Notably, it was more prominent among employed, better-off, less economically dependent individuals living in urban areas, particularly men. Domestic sexual violence against men – as well as overall sexual violence over their lifetime and in the 12 months prior to the GFLHS 2015 – was more common among men in the top asset quintile and with higher education than among men with lower levels of assets and education. The association for women was less clear, but there was evidence that better educated women are more likely to have experienced overall sexual violence (within and outside domestic relations) over their lifetime and in the 12 months prior to the survey than women with lower levels of education. Furthermore, while most acts of sexual violence against women were predominantly perpetrated within domestic relations, most acts of sexual violence against men were predominantly perpetrated outside domestic relations. It was difficult to provide a clear explanation for these results without the use of more sophisticated econometric techniques. The research team postulated that one likely explanation for these results may be that these individuals are more likely to have reported these forms of violence, which remain hidden across most of Ghana. The results discussed in subsection 4.2.3 on attitudes towards domestic sexual violence provided some evidence that may support this explanation. First, the study found that there was very limited discussion about sexual violence in the qualitative data. Second, most survey respondents showed high levels of acceptance of 245 behaviours that the study defined as sexual violence. Taken together, these findings suggest that sexual violence may be understood largely as something that takes place outside domestic relations. Perceptions of what may constitute sexual violence within domestic relations seemed very limited – or hidden. However, much more research is needed to better understand: (i) the true incidence of domestic sexual violence; and (ii) the reasons why sexual violence, both domestic and outside domestic relations, was more prominent against wealthier, more educated men. The research team was also not able to relate the incidence of this violence to the place where it may take place – for instance, their workplace, places where they socialise or within the community. There is a need for future research that makes use of the GFLHS 2015 data to analyse in more depth, using more sophisticated statistical techniques, the reasons underlying these findings, and also future studies that analyse in more detail the incidence, causes and consequences of domestic violence – but also the places where it takes place and the behaviours that may shape its incidence among particular population groups. 4. Further research on the exposure of children to domestic violence. This study focused on the analysis of the incidence, determinants and consequences of domestic violence in Ghana among people aged between 15 and 60 years old. However, the findings reported in the study – particularly in the qualitative data – highlight the exposure of younger children to domestic violence. This exposure takes the form of direct victimisation, when children are beaten or abused in other ways as part of parental disciplining, or indirect exposure to violence committed against others within their homes. The quantitative data has shown, in addition, that the adults surveyed in the GFLHS 2015 are more likely to experience domestic violence if they were exposed directly or indirectly to domestic violence when they were children. These findings suggest that some families and communities in Ghana may experience the transmission of norms and attitudes across time and generations that reinforce the persistence of domestic violence. Addressing this issue will require more knowledge about the incidence, determinants and consequences of domestic violence among children under the age of 15. This research was outside the scope of this study. However, the findings indicate an urgent need for further research in this area. 246 References Abane, H. (2003). “For Better, For Worse: Social Dimensions of Marital Conflict in Ghana:The Case of Cape Coast”, Gender and Behaviour 1: 34–54. Abramsky, T., Watts, C.H., Garcia-Moreno, C., Devries, K., Kiss, L., Ellsberg, M., Jansen,H.A. and Heise, L. (2011). “What Factors Are Associated with Recent Intimate Partner Violence? Findings from the WHO Multi-country Study on Women's Health and Domestic Violence”, BMC Public Health 11: 109. Addai, I. and Adjei, J. (2014). “Predictors of Self-Appraised Health Status in Sub-Saharan Africa: The Case of Ghana”, Applied Research Quality Life 9: 233–53. Adinkrah, M. (2014). “Homicide–Suicide in Ghana Perpetrators, Victims, and Incidence Characteristics”, International Journal of Offender Therapy and Comparative Criminology 58 (3): 364–87. Adjetey, F.N.-A. (1994). “Reclaiming the African Woman’s Individuality: The Struggle Between Women’s Reproductive Autonomy and African Society and Culture”, Am. UL Rev., 44: 1351. Adomako Ampofo, A.A. (2008). “Collective Activism: The Domestic Violence Bill becoming Law in Ghana”, African and Asian Studies 7: 395-421. Adomako Ampofo, A.A. (1993). “Controlling and Punishing Women: Violence against Ghanaian Women”, Review of African Political Economy 20 (56): 102–11. Adu-Gyamfi, E. (2014). “Challenges Undermining Domestic Violence Victims’ Access to Justice in Mampong Municipality of Ghana”, Journal of Law, Policy and Globalization 27 (0): 75–90. Africa, A. (2010). “Standpoint:“Murderous Women”? Rethinking Gender and Theories of Violence”, Feminist Africa 14 Rethinking Gender and Violence. Aizer, A. (2011). “Poverty, Violence, and Health: The Impact of Domestic Violence during Pregnancy on Newborn Health”, Journal of Human Resources 46 (3): 518–538. Aizer, A. (2010). “The Gender Wage Gap and Domestic Violence”, American Economic Review 1847–1859. Akpalu, E.A., Ofei-Aboagye, E.O. and Derbyshire, H. (2000). “Gender Mainstreaming: Emerging Lessons from Ghana”. Alhassan, E. and Odame, F.S. (2015). “Gender Inequality in Basic Education in the Northern Region of Ghana: Household and Contextual Factors in Perspectives”, Ghana Journal of Development Studies 12: 125–141. 247 Amoah, J. (2007). “The World on Her Shoulders: The Rights of the Girl-Child in the Context of Culture and Identity”, Essex Human Rights Review 4 (2): 1–23. Amoakohene, M.I. (2004). “Violence Against Women in Ghana: A Look at Women’s Perceptions and Review of Policy and Social Responses”, Social Science & Medicine 59 (11): 2373–2385. Armstrong, A.K. (1998). Culture and Choice: Lessons from Survivors of Gender Violence in Zimbabwe. Chapel Hill, NC: AKA. Artz, L. (2014). “Fear or Failure: Why Victims of Domestic Violence Retract from the Criminal Justice Process”, South African Crime Quarterly 37: 3–10. Atinmo, M. (2001). “Sociocultural Implications of Wife Beating Among the Yoruba in Ibadan City, Nigeria”. In: Oyekanmi, Felicia (ed.), Men, Women and Violence. Dakar: Council for the Development of Social Science Research in Africa. Banda, F. (1995). “Provision of Maintenance for Women and Children in Zimbabwe”, Cardozo Women's LJ 2 (71). Barker, G. and Ricardo, C. (2005). Young Men and the Construction of Masculinity in Sub- Saharan Africa: Implications for HIV/AIDS, Conflict, and Violence. Washington, DC: World Bank. Bell, C.C. and Mattis, J. (2000). “The Importance of Cultural Competence in Ministering to African American Victims of Domestic Violence”, Violence Against Women 6: 515– 532. Belknap, J. (2014). The Invisible Woman: Gender, Crime, and Justice. Boston, MA: Cengage Learning. Bloch, F. and Rao, V. 2002. “Terror as a Bargaining Instrument: A Case Study of Dowry Violence in Rural India”, American Economic Review 92: 1029–1043. Bograd, M. (1999). “Strengthening Domestic Violence Theories: Intersections of Race, Class, Sexual orientation, and Gender”, Journal of Marital and Family Therapy 25: 275– 289. Bourdieu, P. (2001). “Structure, Habitus, Practice”. In: Erickson, P. and Murphy, I. (eds), Readings for a History of Anthropological Theory. Peterborough, ON: Broadview Press. Bobonis, G.J., Gonzalez-Brenes, M. and Castro, R. (2013). “Public Transfers and Domestic Violence: The Roles of Private Information and Spousal Control”, American Economic Journal: Economic Policy 5 (1): 179–205. Bowman, C.G. (2003). “Domestic Violence: Does the African Context Demand a Different Approach?”, International Journal of Law and Psychiatry 26 (5): 473–491. 248 Bowman, C.G. (2002). “Theories of Domestic Violence in the African Context”, Journal of Gender, Social Policy & the Law 11 (2). Bowman, C.G. and Schneider, E.M. (1998). “Feminist Legal Theory, Feminist Lawmaking, and the Legal Profession”, Fordham L. Rev. 67: 249. Brickell, K. (2015a). “Clouding the Judgment of Domestic Violence Law: Victim Blaming by Institutional Stakeholders in Cambodia”, Journal of Interpersonal Violence, June 2015: 1–21. Brickell, K. (2015b). “Towards Intimate Geographies of Peace? Local Reconciliation of Domestic Violence in Cambodia”, Transactions of the Institute of British Geographers 40 (3): 321–33. Bronfenbrenner, U. (1979). The Ecology of Human Development: Experiments by Nature and Design. Cambridge, MA: Harvard University Press. Burgess, R. and Campbell, C. (2015). “Creating Social Policy to Support Women's Agency in Coercive Settings: A Case Study from Uganda”, Global Public Health 1–17. Butegwa, F. (1994). “Using the African Charter on Human Rights and People’s Rights to Secure Women's Access to Land in Africa”. In: R.J. Cook (ed.), Human Rights of Women: National and International Perspectives. Philadelphia, PA: University of Pennsylvania Press: 495. Buzawa, E.S. and Buzawa, C.G. (2003). Domestic Violence: The Criminal Justice Response. New York: Sage. Campbell, C. (1997). “Migrancy, Masculine Identities and AIDS: The Psychosocial Context of HIV Transmission on the South African Gold Mines”, Social Science & Medicine 45: 273–281. Carlson, B.E. (1984). “Causes and Maintenance of Domestic Violence: An Ecological Analysis”, Social Service Review 58: 569–587. Carrell, S.E. and Hoekstra, M.L. (2010). “Externalities in the Classroom: How Children Exposed to Domestic Violence Affect Everyone's Kids”, American Economic Journal: Applied Economics 2 (1): 211–228. Cerulli, C., Kothari, C., Dichter, M., Marcus, S., Kim, T.K., Wiley, J. and Rhodes, K.V. (2015). “Help-Seeking Patterns among Women Experiencing Intimate Partner Violence: Do They Forgo the Criminal Justice System if their Adjudication Wishes Are Not Met?”, Violence and Victims 30: 16–31. Clark, C.J., Silverman, J.G., Shahrouri, M., Everson-Rose, S. and Groce, N. (2010). “The Role of the Extended Family on Women's Risk of Intimate Partner Violence in Jordan”, Social Science and Medicine 70: 144–151. 249 Coker-Appiah, D. and Cusack, K. (1999). Breaking the Silence and Challenging the Myths of Violence Against Women and Children in Ghana: Report of a National Study on Violence. Accra-North, Ghana: Gender Studies & Human Rights Documentation Centre. Cook, R.J. (2011). Human Rights of Women: National and International Perspectives. Philadelphia, PA: University of Pennsylvania Press. Cornwall, A., Harrison, E. and Whitehead, A. (2007). “Gender Myths and Feminist Fables: The Struggle for Interpretive Power in Gender and Development”, Development and Change 38: 1–20. Cornwall, A. and White, S. (2000). “Men, Masculinities and Development: Politics, Policies and Practices”, IDS Bulletin 31, 1–6. Coughtry, S.E. (2011). “Patriarchy and the Trap of Masculinity: A Post-Colonial Analysis of Violence against Sexual Minorities in Uganda”. MSW Thesis. Northampton, MA: Smith College for Social Work. Cusack, K. and Manuh, T. (2009) (eds.) The Architecture for Violence against Women in Ghana. Accra-North, Ghana: Gender Studies & Human Rights Documentation Centre. http://gendercentreghana.org/?p=632. Daniel, W.W. (1999). Biostatistics: a foundation for analysis in the health sciences. Hoboken, NJ: Wiley. Deaton, A. (1997). The Analysis of Household Surveys: A Microeconometric Approach to Development Policy. Baltimore, MD: Johns Hopkins University Press. Dobash, R.E. and Dobash, R. (1979). Violence against Wives. New York: Free Press. Dobash, R.E. and Dobash, R. (1977). “Love, Honour and Obey: Institutional Ideologies and the Struggle for Battered Women”, Crime, Law and Social Change 1 (4): 403–415. Doughty, K.C. (2014). “‘Our Goal Is Not to Punish but to Reconcile’: Mediation in Postgenocide Rwanda”, American Anthropologist 116: 780–794. Duvvury, N., Carney, P. and Minh, N. (2012). Estimating the Cost of Domestic Violence in Viet Nam. New York: UN Women. Dyson, T. and Moore, M. 1983. “Kinship Structure, Female Autonomy and Demographic Behavior in India”, Population and Development Review 9: 35–60. Edström, J., Das, A. and Dolan, C. (2014a). “Introduction: Undressing Patriarchy and Masculinities to Re-politicise Gender”, IDS Bulletin 45: 1–10. Edström, J., Izugbara, C., Nesbitt-Ahmed, Z., Otieno, P.E., Granvik, M. and Matindi, S. (2014b). “Men in Collective Action on SGBV in Kenya: A Case Study”, IDS Evidence Report 70. Brighton, UK: IDS. 250 Ellsberg, M. and Heise, L. (2005). Researching Violence Against Women: A Practical Guide for Researchers and Activists. Geneva: World Health Organization, PATH. Ellsberg, M., Jansen, H.A., Heise, L., Watts, C.H. and Garcia-Moreno, C. (2008). “Intimate Partner Violence and Women's Physical and Mental Health in the WHO Multi- country Study on Women’s Health and Domestic Violence: An Observational Study”, Lancet 371: 1165–1172. Emmel, N. (2008). Participatory Mapping: An innovative sociological method (Toolkit No. 3). Southampton, UK: ESRC National Centre for Research Methods. Eswaran, M. and Malthotra, N. (2011). “Domestic Violence and Women's Autonomy in Developing Countries: Theory and Evidence”, Canadian Journal of Economics 44 (4): 1222–1263. Eves, R. (2010). “Masculinity Matters: Men, Gender-Based Violence and the AIDS Epidemic in Papua New Guinea”, Civic Insecurity: Law, Order and HIV in Papua New Guinea, 1. Ewelukwa, U. (2002). “Post-Colonialism, Gender, Customary Injustice: Widows in African Societies”, Human Rights Quarterly 24: 424–486. Freedman, J. and Jacobson, S. (2012). “Masculinities, Gender, and Violence”, Engaging Men in the Fight Against Gender Violence: Case Studies from Africa, 1. Freeman, M.D. (1977). “Le Vice Anglais? Wife-Battering in English and American Law”, Family Law Quarterly: 199–251. Friedman, L. and Couper, S. (1987). The Cost of Domestic Violence: A Preliminary Investigation of the Financial Cost of Domestic Violence. New York, Victims Services Agency. Fulu, E., Kerr-Wilson, A., Lang, J., Gibbs, A., Jacobson, J., Jewkes, R., Mazars, C., Shauerhammer, V. and Watts, C. 2014. “What Works to Prevent Violence Against Women and Girls: Evidence Review of Interventions to Prevent Violence Against Women and Men”. Gallegos, J.V. and Gutierrez, A. (2011). “The Effect of Civil Conflict on Domestic Violence: The Case of Peru”, unpublished manuscript. Garcia-Moreno, C., Jansen, H.A., Ellsberg, M., Heise, L. and Watts, C.H. (2006). “Prevalence of Intimate Partner Violence: Findings from the WHO Multi-country Study on Women's Health and Domestic Violence”, Lancet 368: 1260–1269. García-Moreno, C., Jansen, H.A. Ellsberg, M., Heise, L. and Watts, C.H. (2005). WHO Multi- country Study on Women’s Health and Domestic Violence Against Women: Initial Results on Prevalence, Health Outcomes and Women’s Responses. Geneva: World Health Organization. 251 García-Moreno, C., World Health Organization, London School of Hygiene and Tropical Medicine and South African Medical Research Council (2013). Global and Regional Estimates of Violence Against Women: Prevalence and Health Effects of Intimate Partner Violence and Non-Partner Sexual Violence. Geneva: World Health Organization. Ghana Statistical Service (GSS) (2013). 2010 Population and Housing Census: National Analytical Report. Accra: Ghana Statistical Service. Ghana Statistical Service, Ghana Health Service and ICF Macro (2009). Ghana Demographic and Health Survey 2008 (DHS Final Reports No. FR221). Accra: Ghana Statistical Service and Ghana Health Service, and Calverton, MD: ICF Macro. Geisler, G. (1995). “Troubled Sisterhood: Women and Politics in Southern Africa: Case Studies from Zambia, Zimbabwe and Botswana”, African Affairs 545–578. Glaser, B.G. and Strauss, A. (1998). “Grounded Theory”, Strategien qualitativer Forschung. Bern: Huber. Goode, W.J. (1971). “Force and Violence in the Family”, Journal of Marriage and the Family 33: 624–636. Gosselin, D.K. (2000). Heavy hands: An introduction to the crimes of domestic violence. Upper Saddle River, NJ: Prentice Hall. Gupte, J., Justino, P. and Tranchant, J.-P. (2014). “Households amid Urban Riots: The Economic Consequences of Civil Violence in India”, Journal of Conflict Resolution 58 (8): 1445–1473. Haraway, D. (1988). “Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective”, Feminist Studies 575–599. Harrison, A., O’Sullivan, L., Hoffman, S., Dolezal, C. and Morrell, R. (2006). “Gender Role and Relationship Norms among Young Adults in South Africa: Measuring the Context of Masculinity and HIV Risk”, Journal of Urban Health 83: 709–722. Heath, R. (2014). “Women’s Access to Labor Market Oportunities, Control of Household Resources, and Domestic Violence: Evidence from Bangladesh”, World Development 57 (C): 32–46. Heise, L.L. (1998). “Violence against Women An Integrated, Ecological Framework”, Violence Against Women 4(3): 262–290. Heise, L.L. and Garcia-Moreno, C. (2002). “Violence by Intimate Partners”. In: E.G. King, L.L. Dahlberg, J.A. Mercy, A.B. Zwi and R. Lozano (eds.), World Report on Violence and Health. Geneva: World Health Organization. Hidrobo, M. and Fernald, L. (2013). “Cash Transfers and Domestic Violence”, Journal of Health Economics 32: 304–319. 252 Hindin, M.J., Kishor, S. and Ansara, D.L. (2008). “Intimate Partner Violence Among Couples in 10 DHS Countries: Predictors And Health Outcomes”’, DHS Analytical Studies No. 18. Calverton, MD: Macro International Inc. Hodgson, D.L. (2002). “Women's Rights as Human Rights: Women in Law and Development in Africa (WILDAF)”, Africa Today 49: 3–26. Hoyle, C. (1998). Negotiating Domestic Violence: Police, Criminal Justice, and Victims. Oxford, UK: Oxford University Press. Hunnicutt, G. (2009). “Varieties of Patriarchy and Violence against Women: 'Resurrecting Patriarchy' as a Theoretical Tool”, Violence against Women 15, (5): 553–573. IDS, Promundo and Sonke Gender Justice (2015). “Engendering Men: A Collaborative Review of Evidence on Men and Boys in Social Change and Gender Equality”, EMERGE Evidence Review Report. Brighton, UK: IDS. Jewkes, R. (2002). “Intimate Partner Violence: Causes and Prevention”, Lancet 359 (9315): 1423–1429. Jewkes, R., Levin, J. and Penn-Kekana, L. (2002). “Risk Factors for Domestic Violence: Findings from a South African Cross-Sectional Study”, Social Science & Medicine 55 (9): 1603–1617. Jones, A. (2006). “Straight as a Rule Heteronormativity, Gendercide, and the Noncombatant Male”, Men and Masculinities 8: 451–469. Jones, R.B. (2011). “Development of Feminist Postcolonial Theory”, Postcolonial Representations of Women. Dordrecht: Springer Netherlands. Justino, P., Leone, M. and Salardi, P. (2015). “Does War Empower Women? Evidence from Timor Leste”, IDS Evidence Report 121. Brighton, UK: IDS. Justino, P., Cardona, I., Mitchell, B. and Muller, C. (2012). “The Role of Women in Post- Conflict Economic Recovery”. New York: UN Women. Kelly, K.A. (2003). Domestic violence and the politics of privacy. Ithaca, NY: Cornell University Press. Kimuna, S.R. and Djamba, Y.K. (2008). “Gender Based Violence: Correlates of Physical and Sexual Wife Abuse in Kenya”, Journal of Family Violence 23: 333–342. Kishor, S. and Bradley, S.E.K. (2012). “Women’s and Men’s Experience of Spousal Violence in Two African Countries: Does Gender Matter?”, DHS Analytical Studies No. 27. Calverton, MD: ICF International. Koenen, K., Moffitt, T., Caspi, A., Taylor, A. and Purcell, S. (2003). “Domestic Violence is Associated with Environmental Suppression of IQ in Young Children”, Development and Psychopathology 15 (2): 297–311. 253 Koenig, M.A., Lutalo, T., Zhao, F., Nalugoda, F., Wabwire-Mangen, F., Kiwanuka, N., Wagman, J., Serwadda, D., Wawer, M. and Gray, R. (2003). “Domestic Violence in Rural Uganda: Evidence from a Community-Based Study”, Bulletin of the World Health Organization 81: 53–60. La Mattina, G. (2013). Armed Conflict and Domestic Violence: Evidence from Rwanda. Technical Report. Boston, MA: Boston University. Little, J. and Jones, O. (2000). “Masculinity, Gender, and Rural Policy”, Rural Sociology 65: 621. Lloyd, S. (1997). “The Effects of Domestic Violence on Women's Employment”, Law & Politics 19 (2): 139–167. Lugo, B.L. (2015). An Exploratory Study on Mental Health Effects of Therapist Minimization of Domestic Violence Victims’ Experiences. San Diego, CA: Alliant International University. Maguigan, H. (2002). “Wading into Professor Schneider’s Murky Middle Ground between Acceptance and Rejection of Criminal Justice Responses to Domestic Violence”, Am. UJ Gender Soc. Pol’y & L. 11: 427. Mann, J.R. and Takyi, B.K. (2009). “Autonomy, Dependence or Culture: Examining the Impact of Resources and Socio-Cultural Processes on Attitudes Towards Intimate Partner Violence in Ghana, Africa”, Journal of Family Violence 24 (5): 323–335. Mann, R.M. (2000). Who Owns Domestic Abuse?: The Local Politics of a Social Problem. Toronto, ON: University of Toronto Press. Manuh,T. (2007). “The Passage of Domestic Violence Legislation in Ghana”, http://r4d.dfid.gov.uk/PDF/Outputs/WomenEmp/GhanaDV.pdf. Martin, D. (1976). Battered Wives. San Francisco, CA: Glide Publications. McFadden, P. (2000). “The State of Feminism in Africa Today”, News from the Nordic Africa Institute 2. Menon, N. and Johnson, M.P. 2007. “Patriarchy and Paternalism in Intimate Partner Violence: A Study of Domestic Violence in Rural India”. In: K.K. Misra and J.H. Lowry (eds), Recent Studies on Indian Women: Empirical Work of Social Scientists. Jaipur: Rawat Publications. Mills, E., De Paoli, M.M. and Grønningsaeter, A. (2009). “Love in the Time of AIDS: The Relational Gender Dynamics of Prevention, Testing and Treatment”, CSSR Working Paper Series 252. Cape Towm, South Africa: Centre for Social Science Research. 254 Mills, E., Nesbitt-Ahmed, Z., Diggins, J. and Mackieu, T.D. (2015a). ‘They Call Me Warrior’: The Legacy of Conflict and the Struggle to End Sexual and Gender-based Violence in Sierra Leone. Brighton, UK: IDS. Mills, E., Saharokh, T., Wheeler, J., Black, G., Cornelius, R. and Van Den Heever, L. (2015b). Turning the Tide: The Role of Collective Action for Addressing Structural and Gender-based Violence in South Africa. Brighton, UK: IDS. Minayo, M.C.D.S. (1994). “Inequality, Violence, and Ecology in Brazil”, Cadernos de Saúde Pública 10: 241–250. Mitchell, L. (2011). Service Users’ Perceptions of the Domestic Violence and Victims’ Support Unit, Ghana Police Service. Accra: Voluntary Service Overseas (VSO) and Domestic Violence
Secretariat. Moosa, Z. (2012). Violence Against Women and Girls in the Post-2015 Framework: Why & How. London: ActionAid. Murphy, J.C. (2002). “Engaging with the State: The Growing Reliance on Lawyers and Judges to Protect Battered Women”, Am. UJ Gender Soc. Pol’y & L. 11: 499. NETRIGHT (2012). “Ghana NGOs Alternative Report for CSW 57,” Accra-North: NETRIGHT, Association of African Women in Research & Development (AAWORD) and UN Women Ghana Country Programme Office. Njovana, E. and Watts, C. (1996). “Gender Violence in Zimbabwe: A Need for Collaborative Action”, Reproductive Health Matters 4: 46–55. Noe, D. and Rieckmann, J. (2013). “Violent Behaviour: The Effect of Civil Conflict on Domestic Violence in Colombia”, Poverty, Equity and Growth Discussion Paper. Göttingen, Germany: Courant Research Centre. Ofei-Aboagye, R.O. (1994a). “Domestic Violence in Ghana: An Initial Step”, Colum. J. Gender & L., 4: 1. Ofei-Aboagye, R.O. (1994b). “Initial Questions”, The Public Nature of Private Violence: The Discovery of Domestic Abuse. New York: Routledge: 260. Ofei-Aboagye, R.O. (1994c). “Altering the Strands of the Fabric: A Preliminary Look at Domestic Violence in Ghana”, Signs 19 (4): 924–938. Panda, P. and Agarwal, B. (2005). “Marital Violence, Human Development and Women's Property Status in India”, World Development 33: 823–850. Perilla, J.L., Bakeman, R. and Norris, F.H. (1994). “Culture and Domestic Violence: The Ecology of Abused Latinas”, Violence and Victims 9: 325–339. Pollack, R.A. (2004). “An Intergenerational Model of Domestic Violence”, Journal of Population Economics 17: 311–329. 255 Pool, M.S., Otupiri, E., Owusu-Dabo, E., de Jonge, A. and Agyemang, C. (2014). “Physical Violence during Pregnancy and Pregnancy Outcomes in Ghana”, BMC Pregnancy and Childbirth 14 (1): 71. Ribero, R. and Sanchez, F. (2005). “Determinants, Effects and Costs of Domestic Violence”, Documento CEDE 2005-38. Bogota, Colombia: Universidad de los Andes. Richards, P. (2005). “To Fight or to Farm? Agrarian Dimensions of the Mano River Conflicts (Liberia and Sierra Leone)”, African Affairs 104: 571–590. Richards, P. (1996). Fighting for the Rain Forest: War, Youth and Resources in Sierra Leone. Oxford: James Currey. Rowe, M. (2015). “Community Impact Statements in Domestic Violence Criminal Cases: A World First”, Legaldate 27: 11. Schneider, E.M. (2000). Battered Women and Feminist Lawmaking. New Haven, CT: Yale University Press. Schneider, E.M. (2008). “Domestic Violence Law Reform in the Twenty-First Century: Looking Back and Looking Forward”, Fam. LQ 42: 353. Shanley, M.L. (1993). Feminism, Marriage, and the Law in Victorian England. Princeton, NJ: Princeton University Press. Shepard, M. and Pence, E. (1988). “The Effect of Battering in the Employment Status of Women”, Affilia 3 (2): 55–61. Sherman, L.W., Schmidt, J.D. and Rogan, D.P. (1992). Policing Domestic Violence: Experiments and Dilemmas. New York: Free Press. Sipsma, H., Ofori-Atta, A., Canavan, M., Osei-Akoto, I., Udry, C. and Bradley, E. (2013). “Poor Mental Health in Ghana: Who Is at Risk?”, BMC Public Health 13 (1): 288. Sokoloff, N.J. and Dupont, I. (2005). “Domestic Violence at the Intersections of Race, Class, and Gender Challenges and Contributions to Understanding Violence Against Marginalized Women in Diverse Communities”, Violence against Women 11: 38–64. Stockl, H., Devries, K., Rotstein, A., Abrahams, N., Campbell, J., Watts, C. and Garcia- Moreno, C. (2013). “The Prevalence of Intimate Partner Homicide: A Systematic Review”, Lancet 382. Sternberg, K.J., Lamb, M.E., Greenbaum, C., Cicchetti, D., Dawud, S., Cortes, R., Krispin, O. and Lorey, F. (1993). “Effects of Domestic Violence on Children's Behavior Problems and Depression”, Developmental Psychology 29 (1): 44–52. Sulak, T.N., Saxon, T.F. and Fearon, D. (2014). “Applying the Theory of Reasoned Action to Domestic Violence Reporting Behavior: The Role of Sex and Victimization”, Journal of Family Violence 29: 165–173. 256 Takyi, B.K. and Mann, J. (2006). “Intimate Partner Violence in Ghana, Africa: The Perspectives of Men regarding Wife Beating”, International Journal of Sociology of the Family 32 (1): 61–78. Tamale, S.R. (1993). “Law Reform and Women's Rights in Uganda”, East African Journal of Peace and Human Rights 1 (2): 164–94. Tanchen, H.V., Witte, A.D. and Long, S.K. (1991). “Domestic Violence: A Nonrandom Affair”, International Economic Review 32: 491–511. Tenkorang, E.Y., Owusu, A.Y., Yeboah, E.H. and Bannerman, R. (2013). “Factors Influencing Domestic and Marital Violence against Women in Ghana”, Journal of Family Violence 28 (8): 771–81. Thaler, K. (2012). “Norms About Intimate Partner Violence among Urban South Africans: A Quantitative and Qualitative Vignette Analysis”, CSSR Working Paper No. 302. Cape Towm, South Africa: Centre for Social Science Research, Social Surveys Unit. Tsikata, D. and Seini, W. (2004). “Identities, Inequalities and Conflicts in Ghana”, CRISE Working Paper 5. Oxford: Centre for Research on Inequality, Human Security and Ethnicity. Tsikata, P.Y. (2014). “The Historical and Contemporary Representation of Africa in Global Media Flows: Can the Continent Speak Back for Itself on Its Own Terms?”, Communication 40: 34-48. Turshen, M. and Turshen, M. (2000). “Even Strong Backs Need Support: Introductory Notes”, African Women’s Health 8. UNITAR (2011). “River of Life Method”, www.click4it.org. USAID (2006). Addressing Gender-based Violence through USAID’s Health Programs: A Guide for Health Sector Program Officers. Washington, DC: USAID. Walker, L.E. (1999). “Psychology and Domestic Violence Around the World”, American Psychologist 54: 21. Watts, C., Osam, S. and Win, E. (1995). “The Private is Public: A Study of Violence against Women in Southern Africa”. Harare: Zimbabwe: Women in Law and Development in Africa. Wolfe, D.A., Crooks, C.V., Lee, V., McIntyre-Smith, A., Jaffe, P.G. (2003), “The Effects of Children's Exposure to Domestic Violence: A Meta-Analysis and Critique”, Clinical Child and family Psychology Review 6 (3): 171–187. WHO (2001). “Putting Women First: Ethical and Safety Recommendations for Research on Domestic Violence against Women”. Geneva: World Health Organization, http://www.who.int/gender/documents/violence/who_fch_gwh_01.1/en/. 257 Wilson, M. and Daly, M. (1996). “Male Sexual Proprietariness and Violence against Wives”, Current Directions in Psychological Science 5: 2–7. Wilson, M. and Daly, M. (1993). “An Evolutionary Psychological Perspective on Male Sexual Proprietariness and Violence against Wives”, Violence and Victims 8: 271–294. Yllo, K.A. and Strauss, M.A. (1990). “Patriarchy and Violence against Wives: The Impact of Structural and Normative Factors”. In: M.A. Strauss and R.J. Gelles (eds), Physical Violence in American Families. New Brunswick, NJ: Transaction Publishers. Yodanis, C.L. (2004). “Gender Inequality, Violence against Women, and Fear A Cross- National Test of the Feminist Theory of Violence Against Women”, Journal of Interpersonal Violence 19: 655–675. 258 Appendices 259 Appendix A: Incidence of domestic violence worldwide Prevalence of physical Prevalence of sexual violence violence against women (%) against women (%) All By intimate All Intimate perpetrators partner perpetrators partner Region Country or area Year Life Last Life Last Life Last Life Last time 12 time 12 time 12 time 12 mths mths mths mths Africa Burkina Faso 2010 19.8 9.4 11.1 9.2 .. .. 1.5 1.2 Africa Cabo Verde 2005 21.5 20.3 15.7 .. .. .. 3.6 .. Africa Cameroon 2011 54.6 27.4 44.8 29.0 29.0 9.8 20.3 11.2 Africa Central African 2006 35.3 28.7 25.4 22.6 .. .. 11.6 8.3 Republic Africa Chad 2010 .. .. .. 18.4 .. .. .. 12.0 Africa Comoros 2012 14.0 5.8 5.6 4.2 6.0 1.3 1.8 1.3 Africa Côte d'Ivoire 2011-12 35.6 19.9 24.6 22.2 4.5 .. 5.3 4.6 Africa Democratic 2007 63.7 49.0 56.9 .. 16.0 4.2 35.3 .. Republic of the Congo Africa Egypt 2005 47.4 15.7 33.2 18.2 .. .. 6.6 3.9 Africa Equatorial Guinea 2011 62.8 55.8 54.4 40.2 31.9 8.2 17.4 12.5 Africa Ethiopia - 2002 .. .. 48.7 29.0 .. .. 58.6 44.4 Province Africa Gabon 2012 52.4 22.3 46.2 28.3 20.8 8.3 17.0 11.8 Africa Ghana 2008 36.6 17.2 20.6 18.0 18.8 .. 8.2 5.2 Africa Kenya 2008-09 38.5 24.0 37.0 31.3 20.6 .. 17.2 13.6 Africa Liberia 2007 44.0 28.9 35.0 33.0 17.6 .. 10.8 9.6 Africa Malawi 2010 28.2 14.2 21.7 14.7 25.3 .. 18.9 13.4 Africa Mali 2012-13 38.3 24.7 29.8 20.7 12.6 10.5 13.9 12.1 Africa Morocco 2009-10 35.3 15.2 .. 6.4 22.6 8.7 .. 6.6 Africa Mozambique 2011 33.4 25.0 31.5 25.9 12.3 6.9 7.9 6.9 Africa Namibia – City 2002 .. .. 30.6 15.9 .. .. 16.5 9.1 Africa Nigeria 2013 27.8 11.2 14.4 9.3 7.4 3.3 4.8 3.7 Africa Rwanda 2010 41.2 .. 55.6 32.8 22.3 .. 17.5 13.3 Africa Sao Tome and 2008-09 33.4 21.1 26.5 .. 13.1 .. 8.3 .. Principe Africa Sierra Leone 2013 55.5 27.0 44.2 27.2 10.5 4.9 7.3 5.1 Africa South Africa 1998 .. .. 12.5 6.3 .. .. 4.4 .. Africa Tunisia 2010 31.7 7.3 20.3 7.2 15.7 7.4 14.2 9.0 Africa Uganda 2011 56.1 26.9 42.7 24.9 27.8 16.2 27.3 20.9 Africa United Republic 2010 38.7 33.0 39.2 33.4 20.3 .. 17.2 13.7 of Tanzania Africa Zambia 2007 46.8 32.5 46.5 39.6 20.2 .. 16.7 16.0 Africa Zimbabwe 2010-11 29.9 18.4 28.8 20.7 27.2 .. 26.0 13.3 Asia Armenia 2008 .. .. 8.9 .. .. .. 3.3 .. Asia Azerbaijan 2006 13.3 7.7 12.8 9.7 3.7 .. 2.9 2.0 Asia Bangladesh 2011 .. .. 64.6 46.4 .. .. 36.5 24.1 260 Asia Cambodia 2005 22.3 10.3 12.8 7.9 .. .. 2.7 1.7 Asia China 1999-00 .. .. 15.4 .. .. .. .. .. Asia China, Hong 2005 12.0 2.0 6.0 1.0 14.0 3.0 5.0 1.0 Kong Asia Cyprus 2014 20.0 5.0 14.0 3.0 5.0 1.0 4.0 1.0 Asia Georgia 2010 .. .. 4.5 1.4 .. .. 1.7 0.5 Asia India 2005-06 33.5 18.9 35.1 21.4 8.5 .. 10.0 7.2 Asia Japan 2010 .. .. 25.9 .. .. .. 14.1 .. Asia Jordan 2012 34.3 12.6 21.1 11.2 9.2 .. 8.6 6.0 Asia Kyrgyzstan 2012 23.1 13.3 25.1 16.9 3.4 2.1 4.0 2.8 Asia Maldives 2006 .. .. 17.9 5.7 .. .. 6.7 2.0 Asia Mongolia 2008 .. .. .. .. 4.0 .. .. .. Asia Nepal 2011 21.5 9.3 23.1 10.4 12.3 6.4 14.3 7.7 Asia Pakistan 2012-13 32.2 19.2 26.8 18.0 .. .. .. .. Asia Philippines 2013 19.6 5.6 12.7 5.3 6.3 2.7 5.3 3.2 Asia Republic of Korea 2013 .. .. .. 7.2 19.5 2.7 .. 5.4 Asia Singapore 2009 6.8 1.0 5.7 0.9 4.2 0.3 1.2 0.1 Asia State of Palestine 2011 .. .. 30.6 23.5 .. .. 14.6 11.8 Asia Tajikistan 2012 18.8 13.0 19.5 14.5 3.7 2.5 4.4 3.3 Asia Thailand - 2002 .. .. 33.8 13.4 .. .. 28.9 15.6 Province Asia Thailand - City 2005 .. .. 22.9 7.9 .. .. 29.9 17.1 Asia Timor-Leste 2009-10 38.1 29.2 33.5 30.7 3.4 .. 2.9 2.0 Asia Turkey 2014 .. .. 36.0 8.0 .. .. 12.0 5.0 Asia Viet Nam 2010 35.2 .. 31.5 6.4 10.8 .. 9.9 4.2 Europe Albania 2013 .. .. 23.7 14.7 .. .. 7.9 5 Europe Austria 2014 17.0 4.0 12.0 2.0 9.0 2.0 6.0 1.0 Europe Belgium 2014 33.0 10.0 22.0 5.0 13.0 2.0 9.0 1.0 Europe Bulgaria 2014 27.0 7.0 22.0 6.0 12.0 3.0 9.0 3.0 Europe Croatia 2014 19.0 5.0 12.0 3.0 5.0 1.0 3.0 0.0 Europe Czech Republic 2014 30.0 7.0 19.0 4.0 9.0 2.0 7.0 1.0 Europe Denmark 2014 48.0 10.0 29.0 3.0 19.0 2.0 11.0 1.0 Europe Estonia 2014 30.0 4.0 19.0 2.0 13.0 2.0 7.0 1.0 Europe Finland 2014 43.0 9.0 27.0 4.0 17.0 3.0 11.0 1.0 Europe France 2014 42.0 11.0 25.0 5.0 15.0 2.0 9.0 1.0 Europe Germany 2014 33.0 7.0 20.0 3.0 12.0 1.0 8.0 1.0 Europe Greece 2014 24.0 7.0 18.0 5.0 6.0 2.0 5.0 2.0 Europe Hungary 2014 25.0 8.0 19.0 5.0 9.0 3.0 7.0 2.0 Europe Iceland 2008 29.8 2.1 19.7 1.2 24.2 1.6 6.1 0.1 Europe Ireland 2014 24.0 7.0 14.0 3.0 8.0 2.0 6.0 1.0 Europe Italy 2014 25.0 6.0 17.0 5.0 9.0 4.0 7.0 4.0 Europe Latvia 2014 36.0 6.0 31.0 5.0 14.0 2.0 9.0 1.0 Europe Lithuania 2014 30.0 5.0 24.0 4.0 7.0 0.0 4.0 0.0 Europe Luxembourg 2014 36.0 6.0 21.0 3.0 15.0 2.0 9.0 1.0 Europe Malta 2014 19.0 4.0 13.0 3.0 9.0 2.0 6.0 1.0 Europe Netherlands 2014 41.0 9.0 22.0 4.0 18.0 3.0 11.0 2.0 261 Europe Norway 2008 .. .. 13.9 .. .. .. 9.4 1.0 Europe Poland 2014 18.0 4.0 12.0 2.0 5.0 1.0 4.0 1.0 Europe Portugal 2014 23.0 5.0 18.0 4.0 4.0 1.0 3.0 1.0 Europe Republic of 2010 .. .. 39.7 8.9 .. .. 18.6 4.1 Moldova Europe Romania 2014 29.0 7.0 23.0 6.0 6.0 2.0 5.0 2.0 Europe Serbia - City 2003 .. .. 22.8 3.2 .. .. 6.3 1.1 Europe Slovakia 2014 33.0 9.0 22.0 6.0 10.0 3.0 8.0 2.0 Europe Slovenia 2014 21.0 3.0 12.0 2.0 7.0 1.0 4.0 0.0 Europe Spain 2014 20.0 3.0 12.0 1.0 6.0 1.0 4.0 1.0 Europe Sweden 2014 41.0 8.0 24.0 3.0 18.0 3.0 10.0 2.0 Europe Switzerland 2003 27.0 1.0 9.0 1.0 25.0 1.0 3.0 .. Europe Ukraine 2007 16.9 8.8 12.7 10.4 5.4 .. 3.3 2.2 Europe United Kingdom 2014 42.0 8.0 28.0 4.0 14.0 2.0 10.0 1.0 LA & Carib. Bolivia 2008 .. .. 23.1 .. .. .. 6.4 .. LA & Carib. Brazil - City 2000-03 .. .. 27.2 8.3 .. .. 10.1 2.8 LA & Carib. Brazil - Province 2000-03 .. .. 33.8 12.9 .. .. 14.3 5.6 LA & Carib. Colombia 2010 .. .. 37.4 .. .. .. 9.7 .. LA & Carib. Costa Rica 2003 47.0 11.0 33.0 7.0 41.0 7.0 15.0 3.0 LA & Carib. Dominican 2007 19.7 9.7 16.1 10.9 10.1 .. 5.5 3.6 Republic LA & Carib. Ecuador 2011 38.0 .. 35.0 .. 25.7 .. 14.5 .. LA & Carib. El Salvador 2008 .. .. 24.2 6.8 13.4 .. 11.5 3.3 LA & Carib. Guatemala 2008-09 .. .. 24.5 7.8 5.8 .. 12.3 4.8 LA & Carib. Haiti 2012 28.2 10.4 15.6 10.3 13.1 .. 11.1 8.6 LA & Carib. Honduras 2011-12 27.0 .. .. 10.0 4.0 .. .. 3.2 LA & Carib. Jamaica 2008-09 .. .. 17.0 6.5 12.0 .. 7.6 2.8 LA & Carib. Mexico 2011 15.2 6.4 12.3 5.4 38.9 20.8 5.9 2.5 LA & Carib. Nicaragua 2006-07 19.2 .. 27.0 8.0 11.4 .. 13.1 4.4 LA & Carib. Paraguay 2008 .. .. 17.9 6.7 .. .. 5.0 1.7 LA & Carib. Peru 2013 .. .. 35.7 11.5 .. .. 8.4 3.0 LA & Carib. Puerto Rico 1995-96 .. .. .. .. 9.1 .. 5.7 .. Northern Canada 2009 .. 3.4 .. 1.3 .. 2.0 .. .. America Northern United States of 2011 .. .. 31.5 4.0 19.3 1.6 8.8 0.8 America America Oceania Australia 2012 34.0 4.6 15.6 .. 19.0 1.2 4.9 .. Oceania Fiji 2010-11 68.5 .. 61.1 19.4 35.6 .. 33.9 14.2 Oceania Kiribati 2008 .. .. 60.0 32.4 .. .. 46.4 33.7 Oceania New Zealand - 2002 .. .. 30.2 5.3 .. .. 14.1 2.1 City Oceania Samoa 2000 .. .. 40.5 17.9 .. .. 19.5 11.5 Oceania Solomon Islands 2009 .. .. 45.5 .. .. .. 54.7 .. Oceania Tonga 2009 76.8 .. 33.4 12.5 17.4 .. 16.5 11.0 Oceania Vanuatu 2009 .. .. 51.0 33.0 .. .. 44.0 33.0 262 Appendix B: Timeline of major policies related to domestic and gender-based violence International Regional National •Universal Declaration of Human Rights 1948 •International Covenant on Civil and Political Rights 1966 •International Covenant on Social, Economic and Cultural Rights •African Charter on Human and Peoples' Rights 1981 •Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW) 1982 •Republic of Ghana Constitution - does not include domestic violence or violence against women 1992 but does include equality for women •Declaration on the Elimination of Violence Against Women - comprehensive set of standards in 1993 international law for the protection of women against sexual and gender-based violence •Republic of Ghana Criminal Code Article 69A - makes female genital mutilation a second-degree 1994 felony •African Platform for Action on Women - Dakar Declaration - "violation of women's rights is 1994 violation of human rights" … appointment of a Special Rapporteur on Violence Against Women •Protocol to the African Charter on Human and People's Rights on the Rights of Women in Africa 1995 - includes the need for States to adopt legislation aimed to eradicate violence against women •Beijing Declaration and Platform for Action - Conference Report from Fourth World Conference on Women - Countries (including Ghana) that have signed up to the Beijing Declaration are 1995 requried to give five-yearly reports detailing the action they have taken to improve the situation for women and girls •African Commission on Human and People's Rights recruits a Special Rapporteur on the Rights of Women with a mandate to servce as a focal point for promotion and protection of the rights 1999 of women •UN Resolution 55/68 - Elimination fo all forms of violence, including crimes against women 2000 263 •UN Resolution 58/185 - In-depth study on all forms of violence against women 2003 •Ghanian Poverty Reduction Strategy (GPRS) - includes the necessity to formulate and enact 2003 Domestic Violence Act. Put together in consultation with the International Monetary Fund •Ghanaian Strategic Implementation Plan - includes 'mainstreaming of gender issues', decentralisation of 'women's machinery' and monitoring of policies affecting women and 2003 children •UN Resolution 18/147 Elimination of Domestic Violence Against Women - specifically 2004 addressing domestic violence as a separate issue for the first time •African Union - Solemn Declaration on Gender Equality and Women's Empowerment - no 2004 specific mention of domestic violence •UN Resolution 61/143 - intensification of efforts to eliminate all forms of violence against 2006 women •African Commission on Human and People's Rights - Resolution 111 - On the right to remedy 2007 and reparation for women and girl victims of sexual violence •Republic of Ghana Domestic Violence Act 2007 •UN Resolution 7/24 - Elimination of Violence Against Women 2008 •UN Resolution 11/2 - Accelerating efforts to eliminate all forms of violence against women 2009 •UN Resolution 12/17 - Elimination of discrimination against women •GHANAP 1325 - the Ghanaian Action Plan for the implementation of UN Resolution 1325 (women, peace, security). Although not explicitly about domestic violence, the mechanisms 2010 that have been establised to protect women from domestic violence will be adopted to protect women in situations of conflict and gender-based violence •Ghanaian Sector Medium-Term Development Plan (SMTDP)- includes the need to education women regarding their human rights and access to the law - for example, the Domestic Violence 2010 Act •Ghanaian Shared Growth and Development Agenda (GSGDA) - the Domestic Violence Act is seen 2010 as the solution to women's inequality 264 Appendix C: Summary of qualitative sample, methods and instruments Method of data Sampling Sample characteristics Instruments collection FOCUS GROUP One urban and one 1 hour /8–12 people Semi-structured DISCUSSIONS rural site from one 80 FGDs: 8 per region (4 in questionnaire district in each rural, 4 in urban sites) region Group 1: Married/living together men (aged 18–50) Group 2: Married /living together women (aged 18–50) Group 3: Unmarried/youth (aged 18–25) Group 4: Mixed (aged 50+) OBSERVATIONS Primary public Interviews with 20 DOVVSU Observations AND IN-DEPTH institutions of officers (2 from each chosen using a checklist INTERVIEWS victim support: district office in each region) with additional police and health in descriptions one urban and one Observations conducted in rural site from one each of the 10 district Semi-structured district in each DOVVSUs visited where staff questionnaire region were interviewed 28 health staff interviewed, including doctors, nurses and hospital administrators; 10 district hospitals were observed IN-DEPTH Purposive sample Total of 200 opinion leaders:34 Observations INTERVIEWS in all regions 10 legal practitioners (one per using a checklist with OPINION equally divided region); 20 activists (2 per with additional LEADERS, across urban and region); 20 media persons (2 descriptions ACTIVISTS and rural areas per region); 102 community LEGAL EXPERTS leaders as described in the Semi-structured text questionnaire 34 Legal practitioners, media and activists were mainly located at urban sites. 265 Appendix D: Distribution of enumeration areas across regions and survey weights Table D1: Distribution of enumeration areas 2010 Propor- Adjusted Propor- 2010 Adjusted population tionate number of tionate population number of in household allocation HHs allocation Region in household HHs distribution of selected of distribution selected per cent selected per EA selected per EA EAs EAs Western 2,307,395 9.6 32 480 32 480 Central 2,113,766 8.9 29 435 29 435 Greater Accra 3,888,512 16.3 53 795 53 795 Volta 2,086,567 8.6 29 435 29 435 Eastern 2,574,549 10.7 35 525 35 525 Ashanti 4,671,982 19.4 64 960 64 960 Brong Ahafo 2,265,458 9.4 31 465 31 465 Northern 2,445,061 10.1 34 510 34 510 Upper East 1,034,704 4.2 14 210 14 210 Upper West 688,333 2.8 9 135 9 135 Total 24,076,327 100.0 331 4,950 331 4,950 Source: Ghana Family Life and Health Survey (GFLHS) 2015. Weighting of data The survey data were weighted at the data management stage, taking into account the probability of selecting a primary sampling unit from a region, the probability of selecting a household from the primary sampling unit, and the probability of selecting a qualified individual for interviewing. The weight was standardised to be applicable to the data. This was because the sampling design was not self-weighting: households were selected with unequal probabilities at the second stage of sampling. As a result, a sampling weight accounted for differences in the selection probability and in the number of completed interviews among primary sampling units. The required information for computing the weights at the level of the primary sampling unit and household selection stage are explained below. Computation of weights The weight is the product of the first- and second-stage probabilities of selection. The first- stage probability of selection was calculated based on the number of clusters selected in the stratum (ah), multiplied by household size (Mhi) (among those aged 15–60 years) and divided by stratum size (∑Mhi). The second-stage probability of selection was computed as the number of households selected in the clusters i (b) divided by the number of households listed in cluster i (Mhi*) (i.e. the probability of selecting 15 households aged 15– 266 60 years out of the listed households aged 15–60 years). The reciprocal of the overall probability of selection is the weighting factor, also known as the design weight. Let Mhi = Number of 2010 Population Census households in the ith selected enumeration area in the hth stratum or region Mhi* = Number of households listed in the ith selected enumeration area in the hth stratum (urban/rural in the region) ΣMhi = Total number of households in the ith stratum (i.e. number of households in either an urban or rural areas in a region) ahi = Number of sample enumeration areas allocated to the hth stratum (urban/rural in the region) e.g. a11 = 14 for urban area in the Western region and a12 = 18 for a rural area in the Western region b = 15 (number of selected households per enumeration area in each stratum) Then, the first- and second-stage probabilities of selection are: a M b P1hi  hi hi and P2hi  M *hi M hi Where, P1 hi is the probability of selecting the ith enumeration area in the hth stratum, and P2 hi is the probability of selecting a household in the ith enumeration area of the hth stratum. The overall probability of selection of a household in the ith selected enumeration area of the hth stratum is the design weight given by: Fhi  P 1hi *P2hi ahib M hi Fhi  * M *hi M hi The weighting factor (or expansion factor), Whi, for a household in the ith selected enumeration area in the hth stratum is the reciprocal (inverse) of the overall probability of selecting that household. That is, 1 Whi  Fhi 267 M M *  hi * hi ahib M hi The number of households successfully interviewed in each enumeration area was used in the computation. Therefore, the final weight for the sample households in the jth cluster and in the ith sample PSU in stratum h is given by: ' ' b W hi  *W b '' hi Where: b '  the number of interviews plus the number of no interviews in the sample cluster b ''  the total number of interviewed sample households selected in the jth sample PSU within the ith sample stratum h. 268 Appendix E: Regression tables Table E1: Determinants of women having experienced domestic violence in the 12 months prior to survey: odds ratios estimated using logistic regressions Psychological Physical Sexual Social Economic violence violence violence violence violence Age category (r: 15–19) 20–24 0.54** 0.50** 0.44* 0.52*** 0.77 (0.13) (0.14) (0.22) (0.12) (0.22) 25–29 0.42*** 0.34*** 0.37** 0.44*** 1.12 (0.11) (0.095) (0.19) (0.10) (0.31) 30–39 0.35*** 0.31*** 0.17*** 0.23*** 0.69 (0.087) (0.095) (0.088) (0.058) (0.19) 40–49 0.21*** 0.12*** 0.033*** 0.15*** 0.40*** (0.068) (0.045) (0.028) (0.045) (0.13) 50–60 0.16*** 0.0086*** 0.020*** 0.11*** 0.27*** (0.054) (0.0090) (0.021) (0.036) (0.088) Work situation (r: self-employed) Employed 0.80 0.76 0.95 1.35 0.73 (0.20) (0.25) (0.39) (0.27) (0.18) Not working 1.20 1.10 0.81 1.23 1.43** (0.23) (0.25) (0.25) (0.19) (0.22) Asset Index 0.95** 0.96 1.01 0.98 0.95** (0.022) (0.025) (0.011) (0.011) (0.020) Marital status (r: never been in a relationship) In a monogamous 0.75 1.55 1863037.8*** 0.89 1.81* relationship (0.20) (0.52) (728997.8) (0.22) (0.55) In a polygamous 1.91** 3.53*** 3613973.4*** 1.47 4.27*** relationship (0.54) (1.34) (1855414.2) (0.44) (1.47) Divorced, 1.17 2.14** 1844043.1*** 1.06 2.02** separated or (0.34) (0.80) (805891.7) (0.30) (0.67) widowed 269 Psycho logical Phys ical Sex ual Social v iolence Econ omic violence violence violence violence Education (r: none) Primary school 1.12 1.82** 1.23 1.82** 1.50** (0.26) (0.51) (0.71) (0.43) (0.29) Middle school 1.07 1.35 1.70 1.89*** 1.38 (0.23) (0.37) (0.82) (0.39) (0.28) Secondary school 1.45 1.44 2.55* 1.71** 1.21 (0.38) (0.52) (1.38) (0.41) (0.32) Technical school 1.67 1.21 0.75 1.45 1.24 (0.75) (0.79) (0.82) (0.77) (0.55) Higher education 1.40 1.17 0.93 1.12 0.98 (0.59) (0.63) (0.66) (0.39) (0.39) Rural area 1.31 0.96 1.00 0.97 1.39** (0.22) (0.18) (0.30) (0.13) (0.22) Patriarchal Gender 1.07 0.98 0.89 0.99 0.93 Norms Index (0.081) (0.092) (0.11) (0.068) (0.059) Tolerance to Wife- 1.01 1.13 1.01 1.14** 1.09 beating Index (0.069) (0.088) (0.15) (0.072) (0.067) Acceptance of 1.16** 1.17* 1.13 1.16*** 1.11* Women’s Sexual (0.078) (0.10) (0.14) (0.066) (0.071) Autonomy Index Decision-making 0.97 0.93 1.14 0.91 0.95 Power Index (0.064) (0.079) (0.14) (0.056) (0.056) within the household Respondent 1.24 1.35 1.22 0.97 1.39* consumes alcohol (0.26) (0.31) (0.44) (0.19) (0.26) Exposure to 1.77*** 1.49** 3.64*** 1.88*** 1.57*** violence in the (0.24) (0.26) (1.00) (0.25) (0.22) community Witnessed 2.21*** psychological (0.46) violence at home as a child 270 Psycho logical Phys ical Sex ual Social v iolence Econ omic violence violence violence violence Witnessed 1.73* physical violence (0.49) at home as a child Witnessed sexual 1.18 violence at home (0.66) as a child Witnessed social 1.97*** violence at home (0.36) as a child Witnessed 2.93*** economic violence (0.60) at home as a child Regions (r: Western) Central 1.63* 1.32 4.77*** 0.97 0.61** (0.47) (0.60) (2.27) (0.26) (0.15) Greater Accra 1.50 1.11 3.08** 1.28 0.51*** (0.41) (0.44) (1.70) (0.32) (0.13) Volta 0.96 0.96 2.00 0.99 0.22*** (0.25) (0.39) (1.13) (0.26) (0.063) Eastern 2.95*** 2.03* 1.72 1.66* 0.89 (0.83) (0.76) (0.92) (0.44) (0.20) Ashanti 1.66* 1.68 1.56 1.20 0.70* (0.49) (0.65) (0.94) (0.29) (0.15) Brong Ahafo 1.21 1.19 0.55 0.70 0.70 (0.34) (0.48) (0.37) (0.21) (0.16) Northern 0.70 0.68 0.65 1.13 0.26*** (0.25) (0.31) (0.53) (0.33) (0.079) Upper East 0.11** 0.14* 0.97 1.10 0.22*** (0.11) (0.15) (1.09) (0.41) (0.087) Upper West 0.71 1.07 1.64 1.07 0.47* (0.41) (0.51) (1.76) (0.35) (0.18) Observations 2,979 2,879 2,979 2,979 2,979 F-test for regional 0.14 0.67 0.31 0.75 0.54 effects (p-value) Odds ratio; Standard errors in parentheses are clustered within enumeration areas (EA). * p < 0.1; ** p < 0.05; *** p < 0.01 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 271 Table E2: Determinants of men having experienced domestic violence in the 12 months prior to the survey: odds ratios estimated using logistic regressions Psychological Physical Sexual Social Economic violence violence violence violence Violence Age categories (r: 15–19) 20–24 0.73 0.37** 1.73 0.61 0.86 (0.24) (0.16) (1.69) (0.20) (0.24) 25–29 0.64 0.26** 1.49 0.44** 0.70 (0.25) (0.14) (1.55) (0.18) (0.26) 30–39 0.37** 0.051*** 1.88 0.24*** 0.24*** (0.16) (0.046) (1.87) (0.10) (0.11) 40–49 0.48* 0.059*** 0.52 0.20*** 0.52 (0.20) (0.050) (0.63) (0.088) (0.23) 50–60 0.30** 0.12*** 0.17*** 0.27** (0.16) (0.092) (0.086) (0.15) Work situation (r: self-employed) Employed 0.91 0.97 2.69* 1.01 2.03*** (0.22) (0.46) (1.59) (0.24) (0.52) Not working 1.32 0.97 1.68 0.79 2.25** (0.40) (0.41) (1.22) (0.23) (0.71) Asset Index 0.99 0.94 1.03* 1.02** 1.00 (0.0089) (0.055) (0.016) (0.0089) (0.019) Marital status (r: never been in a relationship) In a monogamous 1.09 0.75 6.02* 2.97*** 1.03 relationship (0.29) (0.32) (6.18) (0.85) (0.30) In a polygamous 1.55 0.91 17.4** 5.23*** 0.52 relationship (0.66) (0.67) (19.9) (2.17) (0.33) Divorced, 1.19 0.41 7.91* 1.77* 0.99 separated or (0.37) (0.27) (8.51) (0.61) (0.34) widowed 272 Psycho logical Phys ical Sex ual Soc ial Econ omic violence violence violence violence violence Education (r: none. For sexual violence, r: Higher) Primary school 1.87* 1.97 2064392.7*** 1.94 2.87* (0.68) (1.66) (4702352.7) (0.97) (1.62) Middle school 1.32 2.18 9448049.2*** 1.73 2.96** (0.46) (1.63) (26877735.0) (0.74) (1.43) Secondary 1.91* 4.00* 7752342.1*** 1.54 2.97** (0.71) (3.12) (16778822.1) (0.71) (1.46) Technical 1.20 3899351.4*** 2.16 3.70* (0.73) (9933915.0) (1.35) (2.48) Higher 1.64 1.81 6917039.8*** 2.23 2.56* (0.69) (1.91) (3946834.7) (1.09) (1.35) Rural area 1.28 2.18** 1.18 1.27 1.32 (0.27) (0.75) (0.57) (0.30) (0.31) Patriarchal Gender 0.92 1.06 0.85 0.84* 1.01 Norms Index (0.088) (0.22) (0.18) (0.083) (0.098) Tolerance to Wife- 1.21** 1.47*** 1.18 1.04 1.30** beating Index (0.12) (0.19) (0.28) (0.11) (0.14) Acceptance of 0.91 0.99 1.27 0.96 1.09 Women’s Sexual (0.082) (0.16) (0.34) (0.095) (0.11) Autonomy Index Decision-making 1.06 1.08 0.85 0.65*** 0.88 Power Index (0.13) (0.22) (0.25) (0.071) (0.10) within household Respondent 1.66*** 1.86 2.04* 1.28 1.19 consumes alcohol (0.32) (0.72) (0.77) (0.26) (0.30) Exposure to 1.96*** 2.26** 2.83*** 3.40*** 2.62*** violence in the (0.38) (0.77) (1.07) (0.72) (0.51) community 273 Psycho logical Phys ical Sex ual Soc ial Econ omic violence violence violence violence Violence Witnessed 1.80** psychological (0.51) violence at home as a child Witnessed 1.28 physical violence (0.58) at home as a child Witnessed sexual 2.00 violence at home (1.52) as a child Witnessed social 2.29*** violence at home (0.65) as a child Witnessed 2.28*** economic violence (0.70) at home as a child Regions (r: Western) Central 1.44 2.17 1.21 0.79 1.68 (0.53) (1.60) (0.82) (0.28) (0.64) Greater Accra 1.25 2.39 0.72 1.00 0.80 (0.46) (1.92) (0.49) (0.34) (0.27) Volta 0.93 0.79 0.60 0.35** 0.20*** (0.32) (0.59) (0.42) (0.15) (0.11) Eastern 2.43** 2.65 0.44 0.89 1.25 (0.86) (1.96) (0.36) (0.33) (0.47) Ashanti 1.13 1.51 0.25 0.86 0.89 (0.43) (1.13) (0.21) (0.30) (0.32) Brong Ahafo 0.35** 1.00 0.27*** 0.86 (0.18) (0.88) (0.12) (0.37) Northern 0.87 0.24 0.95 0.33** 0.73 (0.35) (0.29) (0.79) (0.15) (0.36) Upper East 0.35* 1.41 0.16** 0.27* (0.22) (1.22) (0.12) (0.20) Upper West 1.48 0.46 0.25 (1.54) (0.26) (0.28) Observations 1,923 1,904 1,425 1,974 1,974 F-test for regional 0.33 0.70 0.78 0.35 0.09 effects (p-value) Odds ratio; Standard errors in parentheses are clustered within enumeration areas (EA). * p < 0.1; ** p < 0.05; *** p < 0.01 Source: Ghana Family Life and Health Survey (GFLHS) 2015. 274 Questionnaire annex 275 Q1: Household questionnaire IDENTIFICATION ADDRESS _______________________________________________________ LOCALITY NAME ________________________________________________ NAME OF FIRST PRINCIPAL RESPONDENT ________________________ CLUSTER NUMBER_______________________________________________ STRUCTURE NUMBER____________________________________________ HOUSEHOLD NUMBER___________________________________________ REGION_________________________________________________________ DISTRICT________________________________________________________ URBAN =1 | RURAL = 2 _________________________________________________________________ PERSON INTERVIEWED: Man = 1 | Woman = 2 _________________________________________________________________ INTERVIEWER VISITS 1 2 3 FINAL VISIT DATE ___________ ____________ ____________ DAY MONTH INTERVIEWER’S YEAR NAME ___________ ____________ ____________ INT.NUMBER RESULT SUPERVISOR’S NAME ___________ ____________ ____________ RESULT ___________ ___________ ___________ NEXT VISIT: ___________ ____________ TOTAL DATE ___________ ____________ NUMBER TIME OF VISITS RESULT CODES: 1 COMPLETED TOTAL PERSONS 2 NO HOUSEHOLD MEMBER AT HOME OR NO ELIGIBLE IN HOUSEHOLD RESPONDENT AT HOME AT TIME OF VISIT 3 ENTIRE HOUSEHOLD ABSENT FOR EXTENDED TOTAL ELIGIBLE PERIOD OF TIME WOMEN 4 POSTPONED 5 REFUSED TOTAL ELIGIBLE 6 DWELLING VACANT OR ADDRESS NOT A DWELLING MEN 7 DWELLING DESTROYED 8 DWELLING NOT FOUND 9 OTHER ___________________________ LANGUAGE OF LANGUAGE OF RESPONDENT INTERVIEW 276 LANGUAGE CODES: ENGLISH = 1 | AKAN = 2 | GA = 3 | EWE = 4 | NZEMA = 5 | DAGBANI = 6 | OTHER = 7 Informed consent form Hello, I am [name], and I am working for Ghana Statistical Service. We are conducting a national survey about family life and women’s/men’s health. Your household has been randomly chosen to participate in this survey; and we would like to ask you and another household member some questions. I want to assure you that all of your answers will be kept strictly secret. I will not keep a record of your name or address. I am using a tablet computer to make the interview easier and quicker, but I will not record anything or take any pictures. You have the right to stop the interview at any time, or to skip any questions that you don’t want to answer. There are no right or wrong answers. We would very much appreciate your household’s participation in this survey. The interview takes approximately 45 minutes to complete. Do you have any questions for me? YES………………………….1 May we have your permission to ask you questions, and are you willing to participate? YES, BUT NOT RIGHT NOW………………………..2 UNPROMPTED NO…………………………..0 If no, May I ask why you are not willing to participate? 1. I don’t have the time 2. My spouse/partner/guardian DO NOT ASK ANY FURTHER QUESTIONS TO does not let me THE RESPONDENT AND THANK HIM/HER FOR 3. I do not want to talk about my HIS/HER TIME. health and life experiences 4. I do not want to talk to you 5. Other If 2, when would be more suitable for you? _________________ Date DO NOT ASK ANY FURTHER QUESTIONS TO THE RESPONDENT AND THANK HIM/HER FOR _________________ Time HIS/HER TIME. 277 TO BE COMPLETED BY INTERVIEWER I certify that I have read the above consent procedure to the participant. SIGNED: ___________________________________________________ 278 READ: First of all, I would like to make a comprehensive list of all members of this household. I would therefore like to ask you to give me the names of all individuals who normally live and eat their meals together here. Please do not include people who have a household elsewhere and come here only for some hours during the day or to visit temporarily, e.g. servants and visiting family members or friends. Let us start with the head of the household. MAKE SURE THAT RESPONDENT IS THE HEAD OF THE HOUSEHOLD! Special Cases: Domestic Servants if they sleep 5 nights a week or more in the household. Visitors if they have slept in the household for the past 4 weeks. Please circle ELIGIBLE individual respondents: ANY WOMAN/MAN 15–60 YEARS OF AGE, LIVING IN THE HOUSEHOLD Q1: Q2: Q3: Q4: Q5: Q6: Q7: Q8: Person Name How old is What is What is What is What is Is [NAME] Can ID [name]? [NAME]’s [name]’s [NAME]’s [NAME]’s currently [NAME] AGE IN sex? relationship ethnicity? religion? away from read or COMPLETED to the head this write in YEARS of the 98=Don’t household? any Circle those who MALE = 1 household? know Yes = 1 language? are eligible for the FEMALE No = 0 Yes = 1 individual-level = 2 SEE CODES THIS COULD No = 0 survey BELOW BE FOR ASK IF WORK, AGE IS 11 EDUCATION OR OR OLDER SPENDING TIME WITH FAMILY 1 2 3 4 5 6 7 8 279 9 10 11 12 13 14 15 16 17 18 What is the ID of the respondent? ______________ CODES: Relationships to head Q4 CODES: Religion Q6 CODES: Ethnic group Q5 1 = Head 9 = Brother or sister 1 = No religion 1 = Akan 2 = Spouse 10 = Niece/nephew by blood 2 = Catholic 2 = Ga/Dangme 3 = Partner 11 = Niece/nephew by marriage 3 = 3 = Ewe 4 = Son or daughter 12 = Other relative Anglican/Methodist/Presbyterian 4 = Guan 5 = Son-in-law or daughter-in-law 13 = Adopted/foster/step child 4 = Pentecostal/Charismatic 5 = Mole-Dagbani 6 = Grandchild 14 = Not related 5 = Other Christian 6 = Grussi 7 = Parent 98 = Don’t know 6 = Islam 7 = Gruma 8 = Parent-in-law 7 = Traditional/Spiritualist 8 = Mande 8 = Buddhist 9 = Other 9 = Other CODES: Education level Q10 CODES: Work situation Q12 6 = Student/pupil 1 = Pre-school 1 = Self-employed 7 = Apprentice 2 = Primary 2 = Employee 8 = Retired/Too old to work 3 = Middle/JSS/JHS 3 = Casual work 9 = Disabled/sick 4 = Secondary/SSS/SHS 4 = Not working and not looking 10 = Too young to work 5 = TECH/VOC 5 = Not working and looking for work 98 = Don’t know 6 = Higher 98 = Don’t know 280 CODES: Occupation Q13 CODES: Marital Status Q14 1 = Managers 7 = Craft and related trade 1 = Never married 2 = Professionals workers 2 = Living together but not married 3 = Technicians and associate 8 = Plant and machine 3 = Married professionals operators and assemblers 4 = Divorced 4 = Clerical support workers 9 = Elementary occupations 5 = Separated 5 = Service and sales workers 10 = Other occupations 6 = Widowed 6 = Skilled agricultural, forestry and fishery workers 281 Q9: Q10: Q11: Q12: Q13: Q14: Q15: Person Has [NAME] ever What is the How many What describes FOR What is How many years has ID attended school highest level years of [NAME]’s INTERVIEWER: [NAME]’s [NAME] lived in this or is [NAME] of school schooling has current main What describes marital locality? attending school [NAME] has [NAME] work situation [NAME]’s main status? now? attained? completed at best? current occupation FULL NUMBER OF 1= Never Q11 this level? SEE CODES best? SEE CODES YEARS 2 = Now SEE CODES ABOVE ABOVE 3 = Past ABOVE ASK IF AGE SEE CODES ASK IF AGE HAS ALWAYS LIVED ASK IF AGE IS 3 IS 3 OR ASK IF AGE IS 5 ABOVE IS 12 OR HERE = 991 OR OLDER OLDER OR OLDER OLDER LESS THAN ONE YEAR ASK IF AGE IS 5 = 992 OR OLDER 998 = Don’t know DO NOT READ OUT LOUT THE CODES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 282 15 16 17 18 LIST THE IDs OF ALL THE PERSONS IN THE HOUSEHOLD WHO ARE AT LEAST 13 YEARS OLD EMPLOYMENT Q1: Q2: Q3: Q4: Q5: Q6: Q7: Q8: Person Has [NAME] Over the last Did [NAME] Please describe For interviewer: On Was [name] How was ID done any 12 months, work during the what work What work average, paid in [NAME] paid work over how many last 7 days? [NAME] has done activities was how many cash, in- for all the the last 12 months was 1 = Yes over the last 7 [NAME] involved hours per kind (self- work months? [NAME] 0 = No days. in during the last day was consumptio activities involved in 7 days? [NAME] n) or as a during the 1 = Yes any economic EXCLUDING engaged contributin last 7 days? 0 = No  activities? UNPAID in all work g family (multiple Q13 HOUSEWORK activities member for response) EXCLUDING OR CHORES SEE during the any of these UNPAID OCCUPATION last 7 activities? 1 = In kind HOUSEWORK IF NO >>> CODES days? 2 = Cash OR CHORES NEXT PERSON 1 = Yes 3 = IF LAST DO NOT READ 0 = No Contributing PERSON AND OUT LOUT THE If No >> family NO>>Q13 CODES Q11 member IF 2 >> Q10 If 3 >> Q11 Main Second Main Second activity ary activity ary activity activity 283 1 2 3 4 5 6 7 8 9 10 11 12 Q9: Q10: Q11: Q12: Q13: Q14: Person What did How much did [NAME] Who was [NAME] Did [NAME] have Was [NAME] On average, how ID [NAME] receive earn for all work working for during the a work contract involved in house many hours per in-kind for all activities? last 7 days? Prompted for the main work/chores (e.g. day did [NAME] work activities Per: activity? cleaning, laundry, work on house during the last 1 = Day 1 = Employed by family taking care of work/chores month? 2 = Week member 1 = Formal children, cooking) during the last 7 3 = Month 2 = Private employer contract during the last 7 days? 1 = Food 3 = Public sector 2 = Verbal contract days? 2 = IF PAID IN-KIND, WHAT 4 = Self-employed 0 = No contract THEN MOVE TO Accommodation IS THE VALUE? within the family 1 = Yes NEXT or lodging 5 = Self-employed 0 = No INDIVIDUAL 3 = Education outside the family (schooling or 6 = Friend IF NO>>>NEXT vocational) 7 = Other INDIVIDUAL 4 = Clothes 5 = Medicine 2 ANSWERS 6 = Other ALLOWED; 284 Main Secondary Main Secondary occupation occupation occupation occupation Cash Kind Cash Kind 1 2 3 4 5 6 Q16: Over the last 12 months, has anyone in your household lost their job? 1 = Yes 0 = No >>>then next module Q17: If yes, then who? Use household IDs Thank you for this information. Now, I would like to know a little bit more about what your household consumes for food. Name of the item a) Did your household consume b) On how many days over [ITEM] in the last 7 days? the last 7 days did your Yes…1 household consume this No….0 >>>> Next item [ITEM]? Q1: Cereals Q2: Beans/pulses/lentils Q3: Edible oil Q4: Leafy vegetables Q5: Any other vegetables Q6: Tubers (yam, cocoyam, cassava etc.) Q7: Plantain Q8: Dairy product, milk Q9: Eggs Q10: Meat Q11: Fruits 285 Name of the item a) Did your household consume b) On how many days over [ITEM] in the last 7 days? the last 7 days did your Yes…1 household consume this No….0 >>>> Next item [ITEM]? Q12: Fish Q13: Spices Q14: Other food (e.g. sugar, honey, tea leaves, coffee) Q15: During the last lean season, how often have you or a member of your household 1 = more than once a week gone to bed while still hungry? 2 = once a week 3 = a few times per month 4 = once a month 5 = rarely 6 = never Q16: During the last harvest season, how often have you or a member of your 1 = more than once a week household gone to bed while still hungry? 2 = once a week 3 = a few times per month 4 = once a month 5 = rarely 6 = never Now I would like to ask you some questions related to income shocks that your household might have experienced in the last 12 months. Over the last 12 months, have you or anyone in your household been affected by any 1 = Yes 0 = No of the following things? Q17: Flooding/too much rain Q18: Drought Q19: Landslides Q20: Pest or diseases (crops or livestock) Q21: Fire (burning down property) Q22: Insufficient water supply for farming or gardening 286 Q23: How would you or other members of the 1 = Taking a formal loan (from e.g. bank) household get 1000 Cedis if you really needed to. 2 = Taking an informal loan (from e.g. moneylenders, farmers) For example, because of a medical emergency, 3 = Asking relatives/friends in the same village/town 4 = Asking relatives/friends in different location MULTIPLE ANSWERS ALLOWED 5 = Use savings 6 = Use microfinance 7 = Sell durable goods/equipment 8 = Sell land/house 9 = Sell livestock 10 = Take extra work 11 = Other 287 Now I would like to know about goods that belong to you or any other household member. Please also consider goods that are part of your business. How many of the following things does your household possess A. OWN B. Owner(s): Use hhd IDs and who (mainly) owns and uses the respective item(s)? Yes = 1, No = 0 MULTIPLE IDs ALLOWED Q1: Bicycle Q2: Motorcycle or motor scooter Q3: Tricycle Q4: Animal-drawn cart Q5: Car Q6: Truck Q7: Boat with a motor Q8: Boat without a motor Q9: Tractor Q10: A) How much agricultural land does your household possess? B) Unit: If 0, skip to Q13 1= Hectares 2 = Acres 3 = Poles 4 = Plots C) Who owns the agricultural land? Household IDs D) Who usually make decisions regarding the use of agricultural land Household IDs THIS INCLUDES DECISIONS ON WHAT TO GROW AND SALE AND RENT Q11: Is part of this land cultivated by a tenant? Yes=1, No=0 >>> Q13 Q12: F or the plots which you did not crop yourself, what net rent did Total amount in Cedis you receive from the tenant during the last year? IF RENT IS ____________________ PAID IN KIND, ASK THE RESPONDENT TO ESTIMATE THE 98 = Don’t know MONETARY VALUE Q13: H ow much agricultural land does your 1= Hectares household cultivate? 2 = Acres INCLUDE CULTIVATED LAND THAT IS 3 = Poles RENTED IN AND EXCLUDE LAND THAT IS 4 = Plots RENTED OUT. If 0 >> NEXT MODULE 288 Q14: H ow much of this cultivated land is irrigated? 1= Hectares 3 = Poles If 0 >>Q16 2 = Acres 4 = Plots Q15: W hat is the principal source of 1 = Dam 5 = River/lagoon/lake water for this agricultural land? 2 = Hole with a pump 6 = Storage tank 3 = Well 7 = Rainfall 4 = Natural spring 8 = Other Q16: IF HOUSEHOLD OWNS AND CULTIVATE LAND, What were Total amount in Cedis your net earnings of the last 12 months from the land you own ______________________ and cultivate? IF HARVEST IS NOT SOLD TO THE MARKET, 98 = Don’t know ASK THE RESPONDENT TO ESTIMATE THE MONETARY VALUE Q17: D oes your household cultivate land owned by someone outside 1=Yes your household? 0=No >>> NEXT MODULE Q18: W hat was the net earnings of the last 12 months for this rented Total amount in Cedis land? IF HARVEST IS NOT SOLD TO THE MARKET, ASK THE ________________________ RESPONDENT TO ESTIMATE THE MONETARY VALUE 98 = Don’t know For agricultural purposes: IF HOUSEHOLD DOES NOT OWN OR CULTIVATE LAND, SKIP TO Q32 How many of the following things does your household possess and who A. How many? B. Owner(s): Use hhd IDs (mainly) owns the respective item(s)? MULTIPLE IDs ALLOWED Q19: Plough Q20: Cart Q21: Thresher Q22: Trolley Q23: Water pump Q24: Irrigation pump Q25: Grain storage bin Q26: Sprayers Q27: Cattle Q28: Milk cow and bulls Q29: Horses, donkeys, mules 289 Q30: Goats Q31: Sheep Q32: Pigs Q33: Rabbits Q34: Grasscutter Q35: Poultry Q36: Who in your household makes decisions regarding the use of the following animals: a) Cattle b) Milk cows and bulls c) Horses, donkeys and mules d) Goats e) Sheep f) Pigs g) Rabbits h) Grasscutter i) Poultry Q37: Does anyone in your household have a bank account? Who? Yes = 1, No = 0 Q38: Does anyone in your household use a microfinance scheme? Who? Yes = 1, No = 0 Q39: Do you or any member of your household receive remittances from people Yes = 1 living outside your household? No = 0 >>>Q41 Q40: What is the value of these remittances? ____________________ 1 = Day 3 = Month 2 = Week 4 = Year Q41: Are you or any member of your household a participant in or beneficiary of a Yes = 1 cash transfer programme by the government? (LIST) THIS DOES NOT No = 0 >>Q43 INCLUDE CASH TRANSFERS FROM FRIENDS OR FAMILY Q42: What is the value of the transfer? _______________________ THIS DOES NOT INCLUDE CASH TRANSFERS FROM FRIENDS OR FAMILY 1 = Day 3 = Month 2 = Week 4 = Year 290 I would like to ask you about spending decisions in your household. How are spending decisions made in your household? Who usually… Decision-maker (Use Household IDs) 0 = None of us 199 = All of us 98 = Don’t know 99 = Not applicable Q43: Decides how the earnings of the head of the household are used? Q44: Decides how the money from the remittances is used? Q45: Decides how the money from the cash transfer programme is used? Who decides how to spend money on the following items? Q46: Food Q47: Education for the children Q48: Healthcare for the children Q49: Healthcare for yourself Q50: Clothing Q51: Purchase for daily household needs Q52: Major household purchases Q53: Do you own or rent the dwelling that you 1= Owner occupied 4 = Perching >>>>>Q55 live in? 2 = Renting >>>>>Q55 5 = Squatting >>>>>Q55 3 = Rent free >>>>>Q55 Q54: W ho owns it? Roster IDs Q55: I wHould like to ask you a few questions 1 = Piped into dwelling/indoor 9 = Rainwater ab0out your housing condition. 2 = Piped to yard/plot 10 = Tanker truck 1 3 = Public tap/standpipe 11 = Cart with small tank What is the main source of drinking water 4 = Tube well or borehole 12 = Surface water (river/dam/lake/pond for members of your household? 5 = Protected well /stream/canal/irrigation channel) 6 = Unprotected well 13 = Bottled water 7 = Protected spring 14 = Sachet water 8 = Unprotected spring 15 = Other Q56: W hich of the following does your 1 = A private toilet 3 = A public toilet household use the most: 2 = A shared toilet with other families 4 = None of these 291 Q57: WHhat main kind of toilet 1= Flush to piped sewer system 7 = Pit latrine with slab fac0ility do members of your 2 = Flush to septic tank 8 = Pit latrine without slab/open pit ho2u sehold usually use? 3 = Flush to pit latrine 9 = Bucket/pan 4 = Flush to somewhere else 10 = Composting toilet 5 = Flush, don’t know where 11 = No facility/bush/field/sea 6 = Ventilated improved pit latrine Q58: WHhich material is mainly used 1 = No roof 8 = Calamine/cement fibre fo0r the roof of your house? 2 = Thatch/palm leaf/sod 9 = Ceramic tiles/brick tiles 3 3 = Palm/bamboo 10 = Burnt bricks 4 = Wood planks 11 = Cement 5 = Cardboard 12 = Roofing shingles 6 = Metal 13 = Asbestos/slate 7 = Wood 14 = Other Q59: W hich material is mainly used 1 = No walls 8 = Cardboard for the exterior walls of your 2 = Cane/palm/trunks 9 = Reused wood house? 3 = Dirt/mud 10 = Cement or cement blocks 4 = Bamboo with mud 11 = Stone with lime/cement 5 = Stone with mud 12 = Bricks 6 = Uncovered adobe 13 = Covered adobe 7 = Plywood 14 = Wood planks/shingles Q60: W hich material is mainly used 1 = Earth/sand 6 = Vinyl or asphalt strips for the floor of your house? 2 = Dung 7 = Ceramic tiles/terrazzo 3 = Wood planks 8 = Cement 4 = Palm/bamboo 9 = Woollen carpet/synthetic carpet 5 = Parquet or polished wood 10 = Linoleum/rubber carpet Q61: Ho w many rooms do you have in the dwelling that you live in? Q62: Ho w many rooms do you use for sleeping in the dwelling that you live in? 292 Does your household have: 1 = Yes 1 = Yes 0 = No 0 = No Q1: Ra dio Q2: TV Q3: M obile phone Q4: Be d Q5: Ot her furniture Q6: C ooking utensils Q7: W ashing machine Q8: Fa n Q9: Ai r conditioner Q10: Re frigerator Q11: Iro n Q12: Se wing machine Q13: Ke rosene stove Q14: G as/LPG/biogas stove Q15: El ectric stove Q16: O pen stove/open fire/coal pot Q17: Di sh/cable TV Q18: W all clock Q19: La ndline telephone Q20: D igital photo-camera Q21: Ta blet Q22: C omputer Q23: M osquito nets Now I would like to ask you about your source(s) of electricity. On a typical day, how many hours per day do you use this source of energy for lighting? Q24: G rid electricity (household is a utility consumer) Q25: G enerator (from neighbour or relative) Q26: G enerator (personal) Q27: M ini-hydro Q28: S olar home system (e.g. solar panel) Q29: S torage (car) battery Q30: C andles Q31: K erosene (wick and hurricane lamps) Q32: L PG Finally, I would like to ask you some questions about your community. Q33: O n a typical day, in your community, how many hours of electricity is available? Do you know if any of the following are problems in your Yes = 1 No = 0 community or neighbourhood? Q34: C hieftaincy disputes Q35: S ales of drugs (illicit or hard drugs) Q36: D rug abuse (illicit or hard drugs) Q37: A lcohol abuse Q38: D omestic Violence Q39: PUo or/unreliable water provision Q40: P oor/unreliable electricity provision Q41: P oor/unreliable waste removal Q42: P oor roads Q43: P oor health service provision Q44: I s there a shortage of water in this community 1 = Yes during the dry season? 0 = No >>> Q52 2 = There is always dry season >>>Q49 293 1 = More 2 = 3 = A 4 = 5 = 6 = than Once few Once a Rarely Never once a a times a month (once in week week month a while) Q45: H ow often is the shortage of water in this community during the dry season? Q46: H ow often do disputes over water occur between members of this community during the dry season? Q47: H ow often do disputes over water occur between members of this community and a different community during the dry season? Q48: H ow often is the shortage of water in this community? Q49: H ow often do disputes over water occur between members of this community? Q50: H ow often do disputes over water occur between members of this community and a different community? Q51: I s there a shortage of agricultural land in this 1 = Yes, 0 = No >>> Q56, community? 99 = N.A. >>> Q56 Q52: H ow often do disputes over agricultural land occur between members of this community? Q53: H ow often do arguments over grazing land occur in this community? If not applicable, because there is no grazing land, enter 99 Q54: H ow often do arguments over grazing land occur between members of different communities? If not applicable, because there is no grazing land, enter 99 Q55: A re mobile phone networks available 1 = All 4 = Few in all areas of this community? 2 = Most 5 = None 3 = Some 294 Q56: W hich method of transport would you 1 = Foot 7 = Public transport (e.g. primarily use to go to the district capital? 2 = Car bus or tro-tro) 3 = Bicycle 8 = Truck CIRCLE THE APPROPRIATE AND 4 = Motorcycle 9 = Animal drawn cart MAIN MEAN OF TRANSPORT 5 = Tricycle 10 = Tractor 6 = Boat/canoe 98 = Don’t know Q57: H ow long does it take to go to the district capital using this method of transport in minutes? 1 = Minutes 2 = Hours Thank you very much for your time. I would like to interview another member of your household to go into more detail about issues related to health and society. Could I take your phone number for verification purposes? __________________________ Situation of respondent: 1 Alone 2 Children were present some of the time (6–14 years old) 3 Children were present all of the time (6–14 years old) 4 Adults were present some of the time (above 14 years old) 5 Adults were present all of the time (above 14 years old) 66 Other 295 Look at the last digit of the household questionnaire serial number on the cover page. This is the row number you should go to. Check the total number of eligible females or males in the household schedule. How many are eligible? ___________ This is the column
you should go to. The cell where the row and the column meet is the number of the selected woman or man for the domestic violence module in the household schedule. For example, the household was selected to interview a woman and there are three eligible women age 15–60 (line IDs 02, 04 and 05). If the household questionnaire serial number is ‘216’, the last digit is ‘6’; therefore, go to row ‘6’. There are three eligible women age 15–60 in the household, therefore go to column ‘3’. Follow the row and column and find the number where the row and column meet (‘2’) and circle the box. Now go to the household schedule and find the second woman who is eligible for the woman's interview (line number ‘04’ in our example). Write her line ID and name (if given) below. TABLE FOR SELECTION OF RESPONDENT LAST DIGIT OF TOTAL NUMBER OF ELIGIBLE WOMEN 15–60 / THE HOUSEHOLD MEN 15–60 IN THE HOUSEHOLD QUESTIONNAIRE 1 2 3 4 5 6 7 8 SERIAL NUMBER 0 1 2 2 4 3 6 5 4 1 1 1 3 1 4 1 6 5 2 1 2 1 2 5 2 7 6 3 1 1 2 3 1 3 1 7 4 1 2 3 4 2 4 2 8 5 1 1 1 1 3 5 3 1 6 1 2 2 2 4 6 4 2 7 1 1 3 3 5 1 5 3 8 1 2 1 4 1 2 6 4 9 1 1 2 1 2 3 7 5 Line ID of selected woman/man: _________ Name of selected woman/man: ___________________________________________________ ASK IF YOU CAN TALK WITH THE SELECTED WOMAN/MAN for the second part of the questionnaire. IF SHE/HE IS NOT AT HOME, AGREE ON A DATE FOR A RETURN VISIT. 296 Q2: Individual questionnaire IDENTIFICATION ADDRESS _________________________________________________________ LOCALITY NAME _________________________________________________ NAME OF FIRST PRINCIPAL RESPONDENT _________________________ CLUSTER NUMBER _______________________________________________ STRUCTURE NUMBER ____________________________________________ HOUSEHOLD NUMBER ___________________________________________ REGION _________________________________________________________ DISTRICT ________________________________________________________ URBAN =1 | RURAL = 2 _________________________________________________________________ PERSON INTERVIEWED: Man = 1 | Woman = 2 _________________________________________________________________ INTERVIEWER VISITS 1 2 3 FINAL VISIT DATE ___________ ____________ ____________ DAY MONTH INTERVIEWER’S YEAR NAME ___________ ____________ ____________ INT. NUMBER RESULT SUPERVISOR’S NAME ___________ ____________ ____________ RESULT ___________ ___________ ___________ NEXT VISIT: ___________ ____________ TOTAL DATE ___________ ____________ NUMBER TIME OF VISITS RESULT CODES: 1 COMPLETED 2 THE FIRST RESPONDENT DID NOT AGREE TO THE INTERVIEW OF THE SECOND RESPONDENT 3 REFUSED BY THE SECOND RESPONDENT 4 SECOND RESPONDENT NOT AVAILABLE 5 COULD NOT PROCEED BECAUSE OF LACK OF PRIVACY 6 OTHER ___________________________ LANGUAGE OF LANGUAGE OF RESPONDENT INTERVIEW LANGUAGE CODES: ENGLISH = 1 | AKAN = 2 | GA = 3 | EWE = 4 | NZEMA = 5 | DAGBANI = 6 | OTHER = 7 297 Was the respondent alone during the following domestic violence 1 = Yes 0 = No modules? Social violence or controlling behaviour Sexual violence Physical violence Psychological violence Economic violence 298 Hello, I am [name], and I am working for Ghana Statistical Service. We are conducting a national survey about women’s/men’s health and life experiences. Your household has been randomly chosen to participate in this survey; and we would like to ask you some questions. I want to assure you that all of your answers will be kept strictly secret. I will not keep a record of your name or address. You have the right to stop the interview at any time, or to skip any questions that you don’t want to answer. There are no right or wrong answers. Some of the topics may be difficult to discuss, but many women/men have found it useful to have the opportunity to talk. We would very much appreciate your participation in this survey. Do you have any questions? (The interview takes approximately 1 hour and 30 minutes to complete.) YES………………………….1 May I have your permission to ask you questions, and NOT NOW………………….2 are you willing to participate? NO…………………………..0 If no, May I ask why you are not willing to participate? 1. I don’t have any time 2. My spouse/partner/guardian DO NOT ASK ANY FURTHER QUESTIONS TO does not let me THE RESPONDENT AND THANK HIM/HER FOR 3. I do not want to talk about my HIS/HER TIME. health and life experiences 4. I do not want to talk to you 5. Other If it is not a good time to talk, when would be more suitable for you? _________________ Date DO NOT ASK ANY FURTHER QUESTIONS TO THE RESPONDENT AND THANK HIM/HER FOR _________________ Time HIS/HER TIME. Q1: Fo r interviewers: What is sex of 1 = Male, 2 = Female respondent? TO BE COMPLETED BY INTERVIEWER I certify that I have read the above consent procedure to the participant. SIGNED: _______________________________________________________ 299 First, I would be interested to know a bit more about how you feel living in your community. Q2: H ow safe do you feel in the area you 1 = Very safe 3 = Somewhat unsafe live in? 2 = Somewhat safe 4 = Very unsafe Q3: W hy don’t you feel 1. Fear of being robbed 4. Fear of verbal harassment safe in the area you 2. Fear of 5. Fear of physical live in? burglary/vandalism while harassment/attacked being away 6. Fear of being insulted/ridiculed 3. Fear of sexual 7. Fear of something else harassment/rape Q4: H ow safe do you During the day During the night think women and 1 = Very 3 = Somewhat 1 = Very safe 3 = Somewhat girls in this area feel safe unsafe 2 = unsafe when going out 2 = 4 = Very Somewhat 4 = Very unsafe alone? Somewhat unsafe safe safe Q5: H ow safe do you During the day During the night think men and boys in 1 = Very 3 = Somewhat 1 = Very safe 3 = Somewhat this area feel when safe unsafe 2 = unsafe going out alone? 2 = 4 = Very Somewhat 4 = Very unsafe Somewhat unsafe safe safe Q6: Q7: Q8: Have you Have you ever How were you affected? ever personally 1 = Damage to 5 = Loss of witnessed been affected personal property income an act by or 2 = Loss of 6 = Damage related to experienced personal property to communal [problem] acts related to 3 = Physical harm property over the last [problem] over 4 = Psychological 7 = Fear of 12 months? the last 12 harm, stress or leaving the 1 = Yes, months? worry house 0 = No 1 = Yes 8 = Other 0 = No >>> Next item A) Theft, robbery, burglary B) Vandalism C) Arson D) Kidnapping/abduction E) Extortion F) Local bullies G) Political/communal violence H) Highway attacks I ) Sexual harassment G) Rape K) Defilement L) Seduction 300 Q9: If a group of people in your neighbourhood decided to organise 1 = Yes, happily a neighbourhood project, such as a sanitation or clean-up project, 2 = Yes, I will would you like to contribute (time or money) to this project? have to 0 = No Q10: H ow many marriage ceremonies (weddings, engagements) have you been personally invited to in the last 12 months? ________________ Q11: H ow many funerals have you attended in the last 12 months? ________________ Q12: A re you a member of any of the following groups, organisations and associations? Prompt 1 = Area residents’ association 8 = Sports group or club, 14 = Women’s group 2 = Cooperative youth group 15 = Religious group 3 = Professional association 9 = NGO or civic group 16 = Family committee 4 = Trade union 10 = Wives’ association or 17 = Ethnic or tribal 5 = Credit/finance group club association 6 = Neighbourhood council 11 = Fan/social club 18 = Student (including landlords’ association) 12 = Cultural association (e.g. association (Old Boy’s 7 = School committee or parent’s theatre group, reading club) group (PTA) group) 19 = Other 13 = Political group or association Q13: If a member of your household were to get into a disagreement with someone from this community, who would they approach to resolve it? 1 = No one, they will solve it 5 = Other government 9 = Religious leaders amongst themselves officials 10 = Other members of the 2 = Their friends 6 = Community community 3 = Family members leaders 11 = Other 4 = Police 7 = Chief/Queen 12 = Don’t know mother 8 = The district assembly 301 Q1: H ow safe do you feel in your 1 = Very safe home? 2 = Somewhat safe 3 = Somewhat unsafe 4 = Very unsafe The next questions are about things a) Kept you b) Stopped you c) Insisted on d) Stalked you, e) Spread false f) Forced you to that happen to many men and from seeing from leaving knowing where including by information have an women. Please can you tell me if your friends your house? you are at all calling, about you or abortion anyone has ever (even when you or family of times including messaging, distributed were a child) done or tried to do the birth by controlling, watching or photos or following things to you: checking your following you to videos of you Clarify: Anyone includes your partner, phones, texts or a point where without your members of your household, family, emails you felt permission work colleagues etc. uncomfortable Q2: 1 = Yes 2 = Someone tried >>> NEXT ITEM 0 = No >>> NEXT ITEM Q3: H as this happened in the last 12 months? 1 = Yes, 0 = No >>> NEXT ITEM Q4: I n the last 12 months, was the person who did this to you from your current household? 1 = Yes all of them, 2 = Some of them, 0 = No >>Q6 Q5: I n the last 12 months, who did this to you from your current household? Use IDs from the household roster Then skip to Q8 302 Q6: I n the last 12 months, who did this to you? Use codes below Q7: W hat was the sex of this person? 1 = Male, 2 = Female Q8: O ver the last 12 months, how often has this happened? Use codes below Q9: W hen you were a child, did you ever observe this kind of behaviour between 1 = Yes, parents your parents or other members of your household? 2 = Yes, other members 0 = No Q10: T hink of 10 of your closest neighbours, how many of them do you know, or think, experienced this kind of behaviour over the last 12 months? _______________ out of 10 98 = Don’t know CODES FOR Q6 and 16: WHO DID THIS TO YOU? AND WHO DID YOU DO THIS TO? CODES FOR Q8 and 13: HOW OFTEN HAS THIS HAPPENED? 1 = Partner 11 = Other relatives who do not live in this household 1 = Once only 6 = Once a week 2 = Former partner but in the community 2 = Once in a while 7 = 2–3 times a week 3 = In-laws/partner’s family 12 = Other relatives who live outside this community 3 = Less than once per 8 = Almost every day 4 = Parents 13 = Friend month 9 = Every day 5 = Sons, daughters 14 = Teacher 4 = Once per month 6 = Brothers, sisters 15 = Employer 5 = 2–3 times per month 7 = Uncles, aunts 16 = Classmates 8 = Grandparents 17 = Work colleagues 9 = Cousins, nieces, nephews, other 18 = Police relatives 19 = Other people in the community 10 = Priest or religious leader 20 = Other people outside the community 303 Please can you tell me if a) Kept b) Stopped c) Insisted on d) Stalked e) Spread false f) Forced someone you have ever done or tried someone from someone from knowing where someone information about to have an abortion to do the following things seeing his/her leaving their someone is at all including by someone and/or to someone else: friends or home times including by calling, texting, distributed Clarify: Anyone includes your family of birth controlling or watching or photos or videos partner, members of your checking their following to a of someone family, work colleagues etc. phones, texts or point where you without their emails made them feel permission uncomfortable Q11: 1 = Yes 2 = I tried >>> NEXT ITEM 0 = No >>> NEXT ITEM Q12: H as this happened in the last 12 months? 1 = Yes, 0 = No >>> NEXT ITEM Q13: O ver the last 12 months, how often has this happened? Use codes above Q14: O ver the last 12 months, was the person you did this to from your household? 1 =Yes, 0 = No >>>>Q16 304 Q15: O ver the last 12 months, who did you do this to from your household? Use Household IDs Q16: O ver the last 12 months who did you do this to? Use codes above Q17: W hat was the sex of this person? 1 = Male, 2 = Female 305 I would like to know more about how people sometimes relate to each other and your thoughts about it. We have prepared 4 short stories and would like to know your reaction. Please keep in mind that there are no right or wrong answers. S1: After a marital dispute the wife refused to talk to her husband, and would refuse to make his dinner or any other household chores. The husband hit his wife, but without leaving a mark. Q1: Ho w acceptable do you think the 1 = Completely 4 = Completely behaviour of the husband is? unacceptable acceptable 2 = Somewhat 98 = Don’t know/Can’t unacceptable say 3 = Somewhat acceptable Q2: Do you think it is the woman’s fault? 1 = Yes, mostly 3 = Not at all 2 = Partially S1’: After a marital dispute the wife refused to talk to her husband, and would refuse to make his dinner or any other household chores. The husband hit his wife, but without leaving a mark. The local priest/Imam said that it was the woman’s fault because she didn’t fulfil her duties. Q1: Ho w acceptable do you think the 1 = Completely 4 = Completely behaviour of the husband is? unacceptable acceptable 2 = Somewhat 98 = Don’t know/Can’t unacceptable say 3 = Somewhat acceptable Q2: Do you think it is the woman’s fault? 1 = Yes, mostly 3 = Not at all 2 = Partially Q3: Ho w likely do you think it is that 1 = Very likely 4 = Very unlikely your local priest/Imam would react 2 = Somewhat likely 98 = Don’t know/Can’t like this? 3 = Somewhat say unlikely S2: A newly married man in your community, who is usually a very calm and generous man, has been facing some financial problems. One day his boss informs him that he has lost his job. He comes home after this, has an argument with his wife, and beats her up to the point that she passes out/becomes unconscious. This is the first time an incident like this happens. The wife later reports the incident to the police, and the man is arrested. Q4: Do you agree that the woman did the 1 = Fully agree 4 = Fully disagree right thing by going to the police? 2 = Somewhat agree 98 = Don’t know/Can’t 3 = Somewhat say disagree Q5: If 3–4, why do 1 = Because it was the first time it 4 = Because it will break the you think the happened family woman should 2 = Because such matters should 5 = Because it will create not have gone to remain in the family financial problems the police? 3 = Because this is a normal 6 = Other behaviour Q6: Do you agree that the police did the 1 = Fully agree 4 = Fully disagree right thing by arresting the man? 2 = Somewhat agree 98 = Don’t know/Can’t 3 = Somewhat disagree say 306 Q7: If 3–4, why do you 1 = Because it was the first time it 5 = Because it will create think the police happened financial problems should not have 2 = Because such matters should 6 = Because the punishment is arrested the man? remain in the family too harsh 3 = Because this is a normal 7 = Other behaviour 4 = Because it will break the family Q8: WQhat do you think 1 = Financial compensation to the 6 = Apology th8e most adequate wife or her family 7 = Fine pu: nishment for the 2 = Discussions between the wife 8 = Husband should be bonded husband would be? and husband’ s families to a note of good behaviour 3 = Sentencing to jail 9 = It is enough to just arrest 4 = Official warning by the him police/justice 10 = No punishment needed 5 = Police should be physically rough with him S2’: A woman repeatedly makes demeaning comments to her husband about him not making enough money, and not being able to provide for a nice home and gifts for her. She constantly compares him to his brother, who is a successful businessman. One evening, after the woman made such comments and the couple get into an argument, the man beats her up to the point that she passes out/becomes unconscious. This is the first time an incident like this has happened. The wife later reports the incident to the police, and the man is arrested. Q1: Ho w common do you think it is that a 1 = Very common 3 = Somewhat woman treats her husband like this? 2 = Somewhat uncommon common 4 = Very uncommon Q2: Do you agree that the woman did the 1=Fully agree 4=Fully disagree right thing by going to the police? 2= Somewhat agree 98 = Don’t know/Can’t 3=Somewhat disagree say Q3: If 3-4, why do you 1 = Because it was the first time it 5 = Because it will create think the woman happened financial problems should not have 2 = Because such matters should 6 = Because the woman is the gone to the police? remain in the family one to blame 3 = Because this is a normal 7 = Other behaviour 4 = Because it will break the family Q4: Do you agree that the police did the 1 = Fully agree 4 = Fully disagree right thing by arresting the man? 2 = Somewhat agree 98 = Don’t know/Can’t 3 = Somewhat say disagree Q5: If 3–4, why do you 1 = Because it was the first 5 = Because it will create think the police should time it happened financial problems not have arrested the 2 = Because such matters 6 = Because the punishment is man? should remain in the family too harsh 3 = Because this is a normal 7 = Because the man is not to behaviour blame 4 = Because it will break the 8 = Other family 307 Q6: W hat do you think the 1 = Financial compensation to 6 = Apology most adequate the wife or her family 7 = Fine punishment for the 2 = Discussions between the 8 = Husband should be bonded husband would be? wife and husband’ s families to a note of good behaviour 3 = Sentencing to jail 9 = It is enough to just arrest him 4 = Official warning by the 10 = No punishment needed police/justice 5 = Police should be physically rough with him S3: Suppose you hear a dispute between your closest neighbour and her young son. You then witness the son leave the house crying and with several injuries on his face. You know that there are two other young children in the house, and you can still hear your neighbour screaming. Q7: W hat do you 1 = Nothing, I don’t think it is a 6 = Go directly to your neighbour’s think you should problem house and try to stop the dispute do? 2 = Nothing, it is not my and calm down your neighbour business 7 = Approach a relative of the 3 = Approach the police and ask neighbour and ask them to deal them to deal with it with it MULTIPLE 4 = Approach religious 8 = Approach other members of the ANSWERS authorities and ask them to deal community for help. ALLOWED with it 9 = I would go and check what 5 = Approach community happened leaders and ask them to deal with it S4: Suppose a man in your community makes lewd comments to his niece and sometimes touches her inappropriately. This man is a community leader, and talks about the girl inappropriately in front of other people, making comments about her revealing clothes and body. The girl’s mother asks her to wear more modest clothes whenever the uncle visits but does nothing further. Q8: W hat do you 1 = She should dress more 6 = She should tell religious leaders think the girl modestly 7 = She should tell the police or should do? 2 = She should stand up for other authorities herself/ask the uncle to stop 8 = The girl should take this as a 3 = She should threaten the uncle compliment MULTIPLE that she will tell the police or other 9 = This is what men do; she should ANSWERS authorities avoid being with him ALLOWED 4 = She should tell her father 5 = She should tell other family members Q9: W hat do you 1 = The mother is right to ask her 6 = She should tell the police or think the daughter to dress modestly other authorities mother should 2 = She should ask the uncle to stop 7 = She should advise her daughter do? 3 = She should threaten the uncle to go to the police/authorities that she will tell the police or other 8 = She should tell the girl to take authorities this as a compliment 4 = She should tell other family 9 = This is what men do; she should members advise/help her daughter to avoid 5 = She should tell religious leaders being with him 308 S5: Suppose that someone that you know, who has been dating his girlfriend for two years now, starts to become very paranoid that she is being unfaithful to him. This started when the girlfriend mentioned that a young man has joined the office where she works. Ever since then, he checks the messages on her phone, follows her when she is with her friends and shows up at her work place unannounced. He also demands to know where she is at all times, and they get into arguments whenever she comes home from work late. Q10: W hat do you think of the 1 = He has the right to do this man’s reaction? 2 = I sympathise with him, but he should not do this 3 = He should never do this S5’: Suppose that someone that you know, who has been dating her boyfriend for two years now, starts to become very paranoid that he is being unfaithful to her. This started when the boyfriend mentioned that a young woman has joined the office where he works. Ever since then, she checks the messages on his phone, follows him when he is with his friends and shows up at his work place unannounced. She also demands to know where he is at all times, and they get into arguments whenever he comes home from work late. Q11: W hat do you think of the 1 = She has the right to do this woman’s reaction? 2 = I sympathise with her, but she should not do this 3 = She should never do this READ: I am now going to read you a list of statements, and I would like you to tell me whether you generally agree or disagree with the statement. There are no right or wrong answers. 1 = Fully 2 = 3 = 4 = Fully 98 = agree Somewhat Somewhat disagree Don’t agree disagree know/ Can’t say Q12: A good wife obeys her husband even if she disagrees with him Q13: F amily problems should only be discussed with people in the family Q14: I n order to bring up, raise or educate a child properly, the child sometimes needs to be physically punished Q15: IIf a woman wears revealing clothes, it is okay for men to make some sexual comments as she walks in her neighbourhood Q16: I f a woman wears revealing clothes, it is her fault if she is raped Q17: I t is allowed for men to stare at women they don’t know 309 Q18: I n your opinion, does a man have a good reason to hit his wife if: Yes No Don’t know a) She does not complete her household work to his 1 0 98 satisfaction b) She disobeys him 1 0 98 c) She refuses to have sex with him 1 0 98 d) She asks him whether he has other girlfriends 1 0 98 e) He suspects that she is unfaithful 1 0 98 f) He finds out that she has been unfaithful 1 0 98 g) She contracts a sexually transmitted disease 1 0 98 h) She goes out without telling him 1 0 98 i) She neglects the children 1 0 98 j) She argues with him 1 0 98 k) She burns the food 1 0 98 Q19: I n your opinion, can a married woman refuse to have sex with her husband if: Yes No Don’t know a) She doesn't want to 1 0 98 b) He is drunk 1 0 98 c) She is sick 1 0 98 d) He mistreats her 1 0 98 e) She is menstruating 1 0 98 f) He does not want to use a condom/contraception 1 0 98 g) She finds out that he has girlfriends 1 0 98 I want to ask a few questions about you and your current or most recent partner. By partner we mean someone that you have a relationship with, for example a husband or wife, girlfriend or boyfriend, lover or other romantic relationship. Q1: H ave you ever lived with a partner? 1 = Yes 0 = No >>>> Q3 Q2: H ow many partners have you ever lived with? INCLUDE PARTNERS THE RESPONDENT IS _________________ times MARRIED TO AND THOSE YOU ARE CURRENTLY LIVE WITH Q3: H ave you ever been married? 1 = Yes 0 = No >>>> Q5 THIS INCLUDES CURRENT MARRIAGE Q4: H ow many times have you ever been married? _________________ times Q5: A re you currently in a relationship? 1 = Yes 0 = No >>>>> skip to Q52 Q6: A re you in a relationship with more than one 1 = Yes person? 0 = No >>>>> skip to Q10 Q7: If yes, how many persons are you in a relationship with? __________________ number THIS INCLUDES ALL RELATIONSHIPS, EVEN THE PRIMARY RELATIONSHIP Q8: A re you married to any of these other persons? 1 = Yes 0 = No 310 Q9: If yes, how many of these persons are you married to? THIS INCLUDES ALL MARRIAGES, EVEN THE PRIMARY MARRIAGE For the remainder of these questions I would like to ask you about your primary/main partner. This could be the person that you live with. Q10: A re you in a relationship with a man or a woman? 1 = Man 2 = Woman Q11: H ow old were you when you entered this relationship? _________________ years Q12: W as this an approximation 1 = Yes 0 = No Q13: H ow did you get to know your 1 = Arranged meeting current partner? 2 = We are family (e.g. cousin) 3 = Friends/classmates/colleagues/neighbours/customer 4 = At party, festival or other social event 5 = Other Q14: W ho influenced the decision of 1 = Myself choosing your partner? 2 = Parents MULTIPLE ANSWERS 3 = Extended family ALLOWED 4 = Church/Mosque/cult/sect 5 = Other If respondent is man then skip to Q17 Q15: Fo r women: Does your partner have any wives in 1 = Yes addition to your relationship? 0 = No >>>>> Q17 THIS INCLUDES THE RESPONDENT IF S/HE IS 98 = Don’t know >>>>> Q17 MARRIED Q16: Fo r women: How many wives does your partner ______________ have? 98 = Don’t know Q17: D oes your partner have any partners/lovers in 1= Yes addition to yourself? 0 = No >>>>> Q19 98 = Don’t know >>>>> Q19 Q18: H ow many partners/lovers does your partner have? ______________ THIS INCLUDES THE RESPONDENT IF S/HE IS 98 = Don’t know IN A RELATIONSHIP Q19: In general, do you and your current partner discuss Yes No Can’t say or the following topics together: don’t want to say a) Things that have happened to your partner 1 0 3 during the day b) Things that happen to you during the day 1 0 3 c) Your worries or feelings 1 0 3 d) your partner’s worries or feelings 1 0 3 Q20: In your relationship with 1 = Less than once per month 5 = 2–3 times a week your current partner, how 2 = Once per month 6 = Almost every day often would you say that 3 = 2–3 times per month 7 = Every day you quarrelled? 4 = Once a week 311 1 = Yes, all 2 = Yes, 3 = the time sometimes No, never Q21: In your relationship with your current recent partner, has your partner ever gotten angry if you speak with another man/woman? Q22: In your relationship with your current partner, have you ever gotten angry if your partner speaks with another man/woman? Q23: In your relationship with your current partner, does 1 = Very often 3 = Rarely your partner ever suspect that you are unfaithful, and 2 = A few times 4 = Never not trust you? Q24: In your relationship with your current partner, do you 1 = Very often 3 = Rarely ever suspect that your partner is unfaithful, and not 2 = A few times 4 = Never trust him/her? Q25: Ar e you currently living together? 1 = Yes >>>>> Q30 0 = No Q26: W hy not? 1 = I am not the first wife 6 = In this part of Ghana, we don’t have to 2 = S/he is travelling for live with our partner when we are married work or education 7 = S/he is migrating for work or education 3 = Too early in the 8 = I have migrated here for work or relationship education 4 = Waiting to be married 9 = We have separated 5 = Need a bigger house 10 = Other If not, can I Q1: Ho w old is ask you some your partner? questions Q2: W hat is the 1 = Pre-school 5 = TECH/VOC about your highest level of 2 = Primary 6 = Higher partner? education that 3 = Middle/JSS/JHS 98 = Don’t know your partner 4 = Secondary/SSS/SHS has received? Q3: W hat would 1 = Self-employed 7 = Apprentice best describe 2 = Employee 8 = Retired/too old to your partner’s 3 = Casual work work current work 4 = Not working and not 9 = Disabled/sick situation? looking 10 = Too young to work 5 = Not working and looking 98 = Don’t know for work 6 = Student Q4: Ar e you currently married? 1 = Yes 0 = No >>>>>>> Q50 Q5: Ho w many years have you been married? Q6: W hich type of marriage have you 1 = Customary marriage had? 2 = Mohammedan/Islamic 3 = Ordinance 312 I would like to ask you a few questions about what would happen if you were to divorce. I know this might not be happening to you now, but your opinion will help us better understand family life in Ghana. Q7: W ho would be likely 1 = Myself 6 = Girls will go to the mother and boys to the to get custody of your 2 = My spouse father children (0–4 years) 3 = Other family 7 = Divided depending on other assets in the case of a members 8 = Depends on who is at fault or who initiated divorce? 4 = Children the divorce choose 66 = Other 5 = Divided in 98 = Don’t know half between husband and wife Q8: W ho would be likely 1 = Myself 6 = Girls will go to the mother and boys to the to get custody of your 2 = My spouse father children (5–14 years) 3 = Other family 7 = Divided depending on other assets in the case of a members 8 = Depends on who is at fault or who initiated divorce? 4 = Children the divorce choose 66 = Other 5 = Divided in 98 = Don’t know half between husband and wife Q9: W ould you expect the 1 = Yes 4 = Depends on the financial situation main breadwinner to 2 = Yes, but only if they 0 = No provide financial have custody of the 66 = Other support for the children 98 = Don’t know children in case of a 3 = Depends on who is at divorce? fault or who initiated the divorce Q10: W ould you expect the 1 = Yes 4 = Depends on the financial situation main breadwinner to 2 = Yes, but only if they 0 = No provide financial have custody of the 66 = Other support for his/her children 98 = Don’t know spouse in case of a 3 = Depends on who is at divorce? fault or who initiated the divorce Q11: W ho would be likely 1 = Myself 6 = Husband and wife get what belongs to to get immobile 2 = My spouse them assets, such as the 3 = Other family 7 = Husband and wife get what they brought house and land, in members into marriage the case of a divorce? 4 = Divided in half 8 = Depends on who is at fault or who between husband initiated the divorce and wife 66 = Other 5 = Divided 98 = Don’t know depending on custody of children 313 Q12: W ho would be likely 1 = Myself 6 = Husband and wife get what belongs to to get mobile assets, 2 = My spouse them such as vehicles, 3 = Other family 7 = Husband and wife get what they brought furniture, savings, members into marriage livestock, in the case 4 = Divided in half 8 = Depends on who is at fault or who of a divorce? between husband initiated the divorce and wife 66 = Other 5 = Divided 98 = Don’t know depending on custody of children Q13: Fo r men: Did your family pay a 1 = Yes >>> Q41 bride price? 0 = No Q14: Fo r men: Why not? 1 = We could not 3 = We are still in the process of buying the Then next module afford it items 2 = We do not follow 4 = Other this system Q15: Fo r men: What did the bride 1 = Cash 6 = Jewellery price/dowry include? 2 = Livestock 7 = Suitcase 3 = Land 8 = Furniture 4 = Cloth/clothes 9 = Bible or Quran 5 = Drinks 10 = Other gifts Q16: Fo r men: Did your partner 1 = Yes contribute to the bride price? 2 = No >>>> Then next module Q17: Fo r men: What did your partner 1 = Cash 6 = Jewellery contribute? 2 = Livestock 7 = Suitcase 3 = Land 8 = Furniture 4 = Cloth/clothes 9 = Bible or Quran 5 = Drinks 10 = Other gifts Q18: Fo r women: Did your family 1 = Yes >>> Q46 receive a bride price on your 0 = No behalf? Q19: Fo r women: Why 1 = My partner’s family 3 = My partner’s family are still in the not? could not afford it process of buying it Then next module 2 = We do not follow this 4 = Other system Q20: Fo r women: What did the bride 1 = Cash 6 = Jewellery price/dowry include? 2 = Livestock 7 = Suitcase 3 = Land 8 = Furniture 4 = Cloth/clothes 9 = Bible or Quran 5 = Drinks 10 = Other gifts Q21: Fo r women: Did you contribute to 1 = Yes the bride price? 2 = No >>>> Next module Q22: Fo r women: What did you 1 = Cash 6 = Jewellery contribute? 2 = Livestock 7 = Suitcase Then next module 3 = Land 8 = Furniture 4 = Cloth/clothes 9 = Bible or Quran 5 = Drinks 10 = Other gifts 314 Q23: Fo r women whose husband has RANK _______ more than one wife: Are you the first, second, third wife? Q24: Ar e you engaged? 1 = Yes 0 = No Q25: Ar e you in the process of a 1 = Yes customary marriage? 0 = No For those who are currently not in a relationship: Q1: If you are not currently in a relationship, 1 = Yes, but we are now separated/broke up have you had a relationship in the past? 2 = Yes, but s/he left/disappeared IF RESPONDENT HAS HAD MORE 3 = Yes, but we are now divorced THAN ONE RELATIONSHIPS IN THE 4 = Yes, but I am now widowed PAST, PLEASE REFER TO THE MOST 0 = No >>>>>> NEXT MODULE RECENT ONE Q2: If 4 to Q52, were you ever asked to marry 1 = Yes the brother/sister of your spouse who 0 = No passed away? Q3: If 1 to Q53, did you agree to marry a 1 = Yes brother /sister of your spouse who passed 0 = No away? Q4: A fter your spouse passed 1 = Yes, it was all passed on to 4 = No, it was given to away, were your spouse's me my spouse’s family assets (land, property, 2 = Yes, some of it was passed 5 = No, it was given to money…) passed on to you? on to me someone else 3 = No, it was given to our children I would like to ask you some questions about your most recent relationship. Q5: W ere you in a relationship with more 1 = Yes than one person? 0 = No >>>>>> Q60 Q6: If yes, how many persons were you in a relationship with? __________________ number THIS INCLUDES ALL RELATIONSHIPS, INCLUDING THE PRIMARY RELATIONSHIP Q7: W ere you married to any of these other 1 = Yes persons? 0 = No >>>>>> Q60 Q8: If yes, how many of these persons were you married to? __________________ number THIS INCLUDES ALL MARRIAGES, INCLUDING THE PRIMARY MARRIAGE For the remainder of these questions I would like to ask you about your most recent primary/main partner. This could be the person that you live with. Q9: W ere you in a relationship with a man or 1 = Man a woman? 2 = Woman Q10: H ow old were you when you entered this relationship? ___________ years 315 Q11: W as this an approximation? 1 = Yes 0 = No Q12: H ow old were you when you ended this relationship? ____________ years Q13: W as this an approximation 1 = Yes 0 = No I would like to ask you some questions about your most recent partner. Q14: H ow old was your partner when you started your relationship? Q15: H ow old was your partner when you ended your relationship? Q16: W hat was the highest level of 1 = Pre-school 5 = TECH/VOC education that your partner 2 = Primary 6 = Higher had received? 3 = Middle/JSS/JHS 98 = Don’t know 4 = Secondary/SSS/SHS Q17: W hat would best describe 1 = Self-employed 6 = Student your most recent partner’s 2 = Employee 7 = Apprentice work situation at the time? 3 = Casual work 8 = Retired/too old to work 4 = Not working and not 9 = Disabled/sick looking 10 =Too young to work 5 = Not working and 98 = Don’t know looking for work Q18: H ow did you get to know your most 1 = Arranged meeting recent partner? 2 = We are family (e.g. cousins) 3 = Friends/classmates/colleagues/neighbours/cus tomer 4 = At party, festival or other social event 5 = Other Q19: W ho influenced the decision 1 = Myself 4 = Church/Mosque/cult/sect of choosing your partner? 2 = Parents 5 = Other MULTIPLE ANSWERS 3 = Extended family If respondent is man skip to Q 73 ALLOWED Q20: Fo r Women: Did your partner have any wives 1 = Yes in addition to your relationship? 0 = No >>>>> Q73 98 = Don’t know >>>>> Q73 Q21: Fo r Women: How many wives did your partner have? ______________ THIS INCLUDES ALL MARRIAGES, EVEN 98 = Don’t know THE MARRIAGE WITH THE RESPONDENT Q22: D id your partner have any partners/lovers in 1= Yes addition to yourself? 0 = No >>>>> Q75 98 = Don’t know >>>>> Q75 Q23: H ow many partners/lovers did your partner have? ______________ THIS INCLUDES ALL RELATIONSHIPS, 98 = Don’t know EVEN THE RELATIONSHIP WITH THE RESPONDENT 316 Q24: In general, did you and your most recent partner discuss Yes No Can’t say or the following topics together: don’t want to say a) Things that had happened to your partner during the 1 0 3 day b) Things that had happened to you during the day 1 0 3 c) Your worries or feelings 1 0 3 d) Your partner’s worries or feelings 1 0 3 Q25: In your relationship with 1 = Less than once per month (once 5 = 2–3 times a week your most recent partner, in a while) 6 = Almost every day how often would you say 2 = Once per month 7 = Every day that you quarrelled? 3 = 2–3 times per month 4 = Once a week 1 = Yes, all 2 = Yes, 3 = No, the time sometimes never Q26: In your relationship with your most recent partner, did your partner ever get angry if you spoke with another man/woman? Q27: In your relationship with your most recent partner, did you ever get angry if your partner spoke with another man/woman? Q28: In your relationship with your most recent partner, 1 = Very often 3 = Rarely did your partner ever suspect that you were 2 = A few times 4 = Never unfaithful, and not trust you? Q29: In your relationship with your most recent partner, 1 = Very often 3 = Rarely did you ever suspect that your partner was 2 = A few times 4 = Never unfaithful, and not trust him/her? Q30: W ere you living together with your most 1 = Yes >>>>> Q83 recent partner? 0 = No Q31: W hy 1 = I was not the first wife 6 = In this part of Ghana, we don’t have not? 2 = S/he was travelling for work or to live with our partner when we are education married 3 = Too early in the relationship 7 = S/he was migrating for work or 4 = We were waiting to be married education 5 = Needed a bigger house 8 = I had migrated for work or education 9 = We had separated 10 = Other Q32: W ere you married to your most recent 1 = Yes partner? 0 = No >>> Q103 Q33: H ow many years were you married for? Q34: W hich type of marriage did you have? 1 = Customary marriage 2 = Mohammedan/Islamic 3 = Ordinance 317 I would like to ask you a few questions about what would happen if you were to divorce or if you have had a divorce. I know this might not have happened to you, but your opinion will help us better understand family life in Ghana. Q35: W ho would be likely to 1 = Myself 6 = Girls will go to the mother and get custody (or who got 2 = My ex-spouse boys to the father custody) of your children 3 = Other family 7 = Divided depending on other (0–4 years) in the case of a members assets divorce? 4 = Children choose 8 = Depends on who is at fault or 5 = Divided in half who initiated the divorce between husband and 66 = Other wife 98 = Don’t know Q36: W ho would be likely to 1 = Myself 6 = Girls will go to the mother and get custody (or who got 2 = My ex-spouse boys to the father custody) of your children 3 = Other family 7 = Divided depending on other (5–14 years) in the case of members assets a divorce? 4 = Children choose 8 = Depends on who is at fault or 5 = Divided in half who initiated the divorce between husband and 66 = Other wife 98 = Don’t know Q37: W ould you expect the 1 = Yes 4 = Depends on the main breadwinner to 2 = Yes, but only if they have financial situation provide financial support custody of the children 0 = No for the children, or did the 3 = Depends on who is at fault or 66 = Other main breadwinner who initiated the divorce 98 = Don’t know provide financial support for the children, in case of a divorce? Q38: W ould you expect the 1 = Yes 4 = Depends on the main breadwinner to 2 = Yes, but only if they have financial situation provide financial support custody of the children 0 = No for his/her spouse, or did 3 = Depends on who is at fault or 66 = Other the main breadwinner who initiated the divorce 98 = Don’t know provide financial support for his/her spouse, in case of a divorce? Q39: W ho would be likely to 1 = Myself 6 = Husband and wife get what get (or who got) immobile 2 = My ex-spouse belongs to them assets, such as the house 3 = Other family members 7 = Husband and wife get what and land, in the case of a 4 = Divided in half between they brought into marriage divorce? husband and wife 8 = Depends on who is at fault 5 = Divided depending on or who initiated the divorce custody of children 66 = Other 98 = Don’t know 318 Q40: W ho would be likely to 1 = Myself 6 = Husband and wife get what get (or who got) mobile 2 = My ex-spouse belongs to them assets, such as vehicles, 3 = Other family members 7 = Husband and wife get what furniture, savings, 4 = Divided in half between they brought into marriage livestock in the case of a husband and wife 8 = Depends on who is at fault divorce? 5 = Divided depending on or who initiated the divorce custody of children 66 = Other 98 = Don’t know Q41: Fo r men: Did your family pay a bride price? 1 = Yes >>> Q94 0 = No Q42: Fo r men: Why not? 1 = We could not afford it 3 = We were still in the process of Then next module 2 = We do not follow this buying it system 4 = Other For men: What did 1 = Cash 6 = Jewellery the bride 2 = Livestock 7 = Suitcase price/dowry 3 = Land 8 = Furniture include? 4 = Cloth/clothes 9 = Bible or Quran 5 = Drinks 10 = Other gifts Q1: Fo r men: Did your partner contribute to 1 = Yes the bride price? 2 = No >>>> Next module Q2: Fo r men: What did your partner 1 = Cash 6 = Jewellery contribute? 2 = Livestock 7 = Suitcase 3 = Land 8 = Furniture Then next module 4 = Cloth/clothes 9 = Bible or Quran 5 = Drinks 10 = Other gifts Q3: Fo r women: Did your family receive a 1 = Yes >>> Q99 bride price/dowry on your behalf? 0 = No Q4: Fo r women: Why 1 = My partner’s family could 3 = My partner’s family were still not? not afford it in the process of buying it Then next module 2 = We do not follow this system 4 = Other Q5: Fo r women: What did the bride 1 = Cash 6 = Jewellery price/dowry include? 2 = Livestock 7 = Suitcase 3 = Land 8 = Furniture 4 = Cloth/clothes 9 = Bible or Quran 5 = Drinks 10 = Other gifts Q6: Fo r women: Did you contribute to the 1 = Yes bride price? 2 = No >>>> Next module Q7: Fo r women: What did you contribute? 1 = Cash 6 = Jewellery 2 = Livestock 7 = Suitcase Then next module 3 = Land 8 = Furniture 4 = Cloth/clothes 9 = Bible or Quran 5 = Drinks 10 = Other gifts Q8: Fo r women whose husband has more than one wife: Were you the first, second, RANK _______ third wife? Q9: W ere you engaged to your most recent 1 = Yes partner? 0 = No Q10: W ere you in the process of a customary 1 = Yes marriage with your most recent partner? 0 = No 319 Reproductive decisions I would like to ask you some questions about your sexual health and life. Please remember that all of this information will be kept confidential, and your name will be stored separately to the rest of the information that we gather. Q1: How old were you when you first had 99 = no answer _____________ sexual intercourse? 88 = no sexual experience AGE >>>> Q17 Q2: Did you give consent to your first 1 = Yes sexual partner? 0 = No Is the respondent in a relationship? 1= Yes ENUMERATOR TO CHECK 0 = No >>>>> Q19 RELATIONSHIP STATUS For those in a relationship: Now we would like to ask you a few questions regarding sexual and reproductive health. Q3: For women: Are you pregnant at the moment? 1 = Yes 0 = No >>>>> Q12 Q4: Before you were pregnant did you and your partner 1 = Yes use birth control (methods to delay or avoid 0 = No >>>>> Q7 pregnancy?) Q5: Who made this decision? 1 = Yourself 2 = Your partner 3 = Self and partner 4 = Someone else Q6: What type of birth 1 = Condom 9 = LAM control did you and 2 = Female condom 10 = Foam/jelly your partner use? 3 = Hormonal pill 11 = Diaphragm (MULTIPLE ANSWER 4 = Hormonal implant 12 = Rhythm ALLOWED) 5 = IUD (coil) 13 = Emergency contraception 6 = Sterilisation 14 = Injectable FOR ALL, GO TO Q12 7 = Abstinence 15 = Other AFTER 8 = Withdrawal Q7: Why did you and your 1 = Partner didn’t allow 10 = We are trying to have partner not use birth 2 = Someone else did not children control? allow 11 = Me/partner post- 3 = Makes me feel weak/ill menopausal (MULTIPLE ANSWER or other side effects 12 = Homosexual relationship ALLOWED) 4 = Makes sex less 13 = Me/partner is infertile pleasurable 14 = Tubes are tied, FOR ALL, GO TO Q12 5 = Didn’t feel the need to reproductive organs removed AFTER 6 = Religious beliefs or vasectomy 7 = It is expensive 15 = We had not heard of 8 = I felt embarrassed contraceptive methods 9 = It is not available 16 = We don’t have sex 17 = I was not menstruating 18 = Other Q8: Do you and your partner use birth control (methods 1 = Yes to delay or avoid pregnancy?) 0 = No >>>>> Q11 320 Q9: Who made this decision? 1 = Yourself 2 = Your partner 3 = Self and partner 4 = Someone else Q10: What type of birth 1 = Condom 9 = LAM control do you and your 2 = Female condom 10 = Foam/jelly partner use? 3 = Hormonal pill 11 = Diaphragm (MULTIPLE ANSWER 4 = Hormonal implant 12 = Rhythm ALLOWED) 5 = IUD (coil) 13 = Emergency contraception 6 = Sterilisation 14 = Injectable FOR ALL, GO TO Q12 7 = Abstinence 15 = Other AFTER 8 = Withdrawal Q11: Why are you and your 1 = Partner didn’t allow 10 = We are trying to have partner not using birth 2 = Someone else did not children control? allow 11 = Me/partner post- 3 = Makes me feel menopausal (MULTIPLE ANSWER weak/ill or other side 12 = Homosexual relationship ALLOWED) effects 13 = Me/partner is infertile 4 = Makes sex less 14 = Tubes are tied, pleasurable reproductive organs removed 5 = Didn’t feel the need to or vasectomy 6 = Religious beliefs 15 = We have not heard of 7 = It is expensive contraceptive methods 8 = I felt embarrassed 16 = We don’t have sex 9 = It is not available 17 = I am not menstruating 18 = Other Q12: Do you and your partner use a safe sex method to 1 = Yes avoid sexually transmitted diseases and infections? 0 = No >>>>> Q14 Q13: Which method do you and your partner use? 1 = Condom (MULTIPLE ANSWER ALLOWED) 2 = Female condom FOR ALL, GO TO Q15 AFTER 3 = Abstinence 4 = Other Q14: Why don’t you and your 1 = Partner didn’t allow 6 = Religious beliefs partner use any method? 2 = Someone else did not 7 = It is expensive allow 8 = I felt embarrassed (MULTIPLE ANSWER 3 = Makes me feel 9 = It is not available ALLOWED) weak/ill or other side 10 = We are trying to have effects children 4 = Makes sex less 11 = We have not heard of safe pleasurable sex methods 5 = Didn’t feel the need to 12 = We don’t have sex because we trust each 13 = Other other Q15: Over the last six months, or since you 1 = Never met your partner, how often do you and 2 = Every day your partner engage in sexual 3 = Several times a week intercourse? 4 = Once a week 5 = Several times a month 321 6 = Occasionally Q16: Do you think this is adequate for you? 1 = Yes 2 = No, it should be more 3 = No, it should be less Q17: Has your partner ever refused to have 1 = Yes sexual intercourse with you for long 0 = No periods of time? Q18: Who makes decisions about how many 1 = Yourself children to have? 2 = Your partner 3 = Self and partner 4 = Someone else 5 = We have not discussed this (it is too early) 6 = We cannot have children Q19: Have you ever heard of female 1 = Yes >>>> if woman skip to q 21. If man circumcision? skip to next module 0 = No Q20: In a number of countries, there is a 1 = Yes >>>> If man skip to next module practice in which a girl may have part of 0 = No >>>> Skip to next module her genitals cut. Have you ever heard about this practice? For women: Q21: Have you yourself ever had your 1 = Yes genitals cut? 0 = No >>>> Next module Q22: Could you please tell me what was done 1 = Yes to you at this time. 0 = No Was any flesh removed from the genital 98 = Don’t know area? Was the genital area just nicked without 1 = Yes removing any flesh? 0 = No 98 = Don’t know Was your genital area sewn closed? 1 = Yes 0 = No How old were you when this occurred? Age in completed years ____________ 95 = During infancy 98 = Don’t know Who cut (or nicked) the genitals? 1 = Traditional circumciser 2 = Traditional birth attendant 3 = Other traditional person 4 = Doctor 5 = Trained nurse/midwife 6 = Other health professional 7 = Other 98 = Don’t know 322 Please can you tell me if anyone has ever done or tried to do the following things to you: Clarify: Anyone includes your partner, members of your household, family, work colleagues etc. Q1: 1 = Yes 2 = Someone tried >>> NEXT ITEM 0 = No >>> NEXT ITEM Q2: Di d this happen when you were younger than 15 years old? 1 = Yes, 0 = No Q3: H as this happened in the last 12 months? 1 = Yes, 0 = No >>> NEXT ITEM 323 a) Made inappropriate sexual comments to you that made you feel uncomfortable b) Touched you in an inappropriate and sexual way that made you feel uncomfortable c) Physically forced you to have sexual intercourse or perform a sexual act when you did not want to? d) Otherwise forced you to have sexual intercourse or perform a sexual act when you did not want to? This could be by blackmailing, threatening or scaring you. e) Had sexual intercourse or other sexual act without you being able to give your permission? f) Not used protection even after you asked? g) A sexual partner did not reveal to you that he/she had HIV (and they knew about it)? h) Had sexual intercourse or performed a sexual act with someone because you felt like you did not have a choice or you were worried about the reaction. i) Penetrated you with an object against your will? Q4: In the last 12 months, was the person who did this to you from your current household? 1 = Yes all of them, 2 = Some of them, 0 = No >>Q6 Q5: W ho did this to you from your current household? Use IDs from Household roster Then skip to Q8 Q6: W ho did this to you? Use codes above Q7: W hat was the sex of this person? 1 = Male, 2 = Female Q8: I a m sorry to ask details about this experience, but could you tell me how they physically forced you? Q9: Ov er the last 12 months, how often has this happened? Use codes above Q10: Di d this result in any injuries? 1 = Yes, 0 = No >>>>> Q12 324 Q11: W hat type of injury? Q12: Ho w has this affected your daily activities? USE CODES BELOW Q13: If 6 to Q12: How many days were you not able to work over the last 12 months? Q14: If 5 to Q12: How many days were you not able to go to school over the last 12 months? Q15: W hen you were a child, did you ever observe this kind of behaviour between your 1 = Yes, parents parents or other members of your household? 2 = Yes, other members 0 = No Q16: Th ink of 10 of your closest neighbours, how many of them do you know, or think, experienced this kind of behaviour over the last 12 months? 98 = Don’t know _______________ out of 10 CODES FOR Q5 and 36: WHO DID THIS TO YOU? AND WHO DID YOU DO THIS TO? CODES FOR Q9 and 33: HOW OFTEN HAS THIS HAPPENED? 1 = Partner 11 = Other relatives who do not live in this household 1 = Once only 6 = Once a week 2 = Former partner but in the community 2 = Once in a while 7 = 2–3 times a 3 = In-laws/partner’s family 12 = Other relatives who live outside this community 3 = Less than once per month week 4 = Parents 13 = Friend 4 = Once per month 8 = Almost every 5 = Sons, daughters 14 = Teacher 5 = 2–3 times per month day 6 = Brothers, sisters 15 = Employer 9 = Every day 7 = Uncles, aunts 16 = Classmates 8 = Grandparents 17 = Work colleagues 9 = Cousins, nieces, nephews, other 18 = Police relatives 19 = Other people in the community 10 = Priest or religious leader 20 = Other people outside the community 325 CODES FOR Q11: WHAT TYPE OF INJURY CODES FOR Q12: HOW HAS THIS AFFECTED YOU? CODES FOR Q8: HOW (select multiple) WERE YOU PHYSICALLY FORCED? 1 = Non-genital injuries 6 = Migraines and other 0 = Not at all 6 = I could not go to 1 = Pinning you down, so (Cuts, punctures, bites, frequent headaches 1 = Unable to concentrate at school that you could not get up Scratches, abrasion, bruises) 7 = Sexually transmitted work 7 = I couldn’t do 2 = Beating you up if you 2 = Genital injuries (bruising, infections 2 = Unable to concentrate at domestic work tried to move inflammation, tenderness, 8 = Cervical cancer school 8 = Lost confidence in 3 = Beating you until you abrasions, or lacerations) 9 = Loss of pregnancy 3 = Unable to concentrate at own ability became unconscious 3 = Chronic pain 10 = Premature birth domestic work 9 = Left work 4 = Otherwise hurting you 4 = Gastrointestinal 4 = Could not take care of the 10 = Left school so that you would not disorders children 11 = Live in constant move 5 = Gynaecological 5 = I had to miss work fear complications 12 = Other 326 Q17: Ov er the last 12 months, how many times 1 = Once/twice were you injured due to sexual violence? 2 = 3–5 times 3 = Many (more than 5) times Q18: Ov er the last 12 months, were you ever hurt 1 = Times needed health care: badly enough due to sexual violence that ________ you needed health care? 2 = Yes, but don’t know IF YES: How many times? 3 = Not needed >>> NEXT MODULE Q19: Di d anyone stop you from trying to receive 1 = Yes health care for these injuries? 0 = No >>>>> Q24 Q20: W as the person who stopped you from your 1 = Yes household? 0 = No >>>>> Q22 Q21: W ho stopped you from your household? Use Household IDs Q22: W ho stopped you from outside your Use perpetrator codes household? Q23: Ho w many times has someone stopped you from receiving health care over the last 12 __________________ months? Q24: Ov er the last 12 months, have you ever 1 = Times needed health care: received health care these injuries? ________ IF YES: All of the time, or sometimes? 2 = Yes, but don’t know 3 = Not needed >>> NEXT MODULE Q25: Ho w much did you spend on health care for your injuries including medicines, fees and transport (IF MULTIPLE VISITS: during your last health visit)? Q26: Fo r your injury, did you have to spend any Number of nights in health facility: nights in any health facility over the last 12 ________ months? If none, enter ‘00’ IF YES: How many nights (in total)? Q27: Du ring your last health care visit, did you 1 = Yes >>> NEXT MODULE tell a health worker the real cause of your 0 = No injury? Q28: Du ring your last health care visit, why did 1 = I was afraid they would tell someone you decide not to tell the real cause of your 2 = I was afraid of repercussions from the injury? person who hurt me 3 = I was afraid they would not believe me 4 = I was afraid they would not care MULTIPLE ANSWERS ALLOWED 5 = I was afraid they would treat me badly 6 = Some of the people who work there know me or my family 7 = Other 327 If respondent has reported no sexual violence, Read out loud: I know that these questions are very personal and some people would rather not reveal this sensitive information to a stranger. So we are giving everyone, no matter what they said before, the chance to reveal this information privately and discreetly without me watching you. No matter what you answered before, I am going to give you the questionnaire, and let you choose your responses without me looking. You can turn the page over as soon as you are done Q29: R ead to respondent: Before you were 15 years old, had anyone ever touched you sexually or made you do something sexual that you did not want to do? Now hand questionnaire to respondent. Q30: N ow I would like you to answer the same question, but when you were older than 15 years old. Now hand questionnaire to respondent. 328 If respondent has been injured (Yes to ANY IN Q10) READ: I would like to learn more about the injuries resulting from the sexual violence you have experienced Please can you tell me if you have a) Made b) Touched c) d) e) Had f) Not used g) Not h) Felt that i) ever done or tried to do the inappropri someone in Physically Otherwise sexual protection reveal to a someone Penetrated following things to someone else: ate sexual an forced forced intercourse even after sexual had sexual someone Clarify: Anyone includes your comments inappropri someone to someone to or someone partner intercourse with an to someone ate and have have performed asked? that you or object partner, members of your family, that made sexual way sexual sexual a sexual act had HIV performed against work colleagues etc. them feel that made intercourse intercourse with (and you a sexual act their will? uncomfort them feel or perform or perform someone knew with you able uncomfort a sexual act a sexual who was about it)? because able when they act? This not able to you felt like did not could be give their they did not want to? by permission have a blackmaili ? choice or ng, were threatening worried or scaring about your them reaction 329 Q1: 1 = Yes 2 = I tried >>> NEXT ITEM 0 = No >>> NEXT ITEM Q2: Ha s this happened in the last 12 months? 1 = Yes, 0 = No >>> NEXT ITEM Q3: Ov er the last 12 months, how often has this happened? Above Q4: Ov er the last 12 months, was the person you did this to from your household? 1 =Yes, 0 = No >>>>Q36 Q5: O ver the last 12 months, who did you do this to from your household? Use Household IDs Q6: O ver the last 12 months who did you do this to? Use codes above Q7: W hat was the sex of this person? 1 = Male, 2 = Female 330 Please can you tell me if anyone a) Slapped b) Pushed c) Hit you d) Kicked e) Choked f) Burnt g) Used a h) Poured i) FOR has ever done, or tried to do, the you or you or with you, or you on gun, knife hazardous MEN following things to you: thrown shoved his/her fist dragged strangled purpose? or other chemicals ONLY: Clarify: Anyone includes your something you? or with you or you on weapon or Kicked or partner, members of your at you that something beaten purpose? against substance pulled household, family, work colleagues could hurt else that you up? you? s (e.g. your etc. you? could hurt acid) on external you? you? genitalia? Q1: 1 = Yes 2 = Someone tried >>> NEXT ITEM 0 = No >>> NEXT ITEM Q2: Di d this happen when you were younger than 15 years old? 1 = Yes, 0 = No Q3: Ha s this happened in the last 12 months? 1 = Yes, 0 = No >>> NEXT ITEM Q4: In the last 12 months, was the person who did this to you from your current household? 1 = Yes all of them, 2 = Some of them, 0 = No >>Q6 Q5: In the last 12 months, who did this to you from your current household? Use IDs from Household roster Then skip to Q8 331 Q6: In the last 12 months, who did this to you? Use codes below Q7: W hat was the sex of this person? 1 = Male, 2 = Female Q8: Ov er the last 12 months, how often has this happened? Use codes below Q9: Ov er the last 12 months did this result in any injuries? 1 = Yes, 0 = No >>>>> Q11 Q10: W hat type of injury? Q11: Ov er the last 12 months, did this ever end in a fight where you also hit back and retaliated? Clarify: to hurt the person doing this to you? USE CODES BELOW Q12: Ho w has this affected your daily activities over the last 12 months? USE CODES BELOW Q13: If 6 to Q12: How many days were you not able to work over the last 12 months? 332 Q14: If 5 to Q12: How many days were you not able to go to school over the last 12 months? Q15: Fo r women, if answered yes to any of the above: During the times you were hit, 1 = Yes, 0 = No >>> Q17 were you ever pregnant? Q16: Fo r women, if answered yes to any of the above: If yes, did something happen 1 = Yes, I had a miscarriage 5 = Neurological to your pregnancy? 2 = Still birth complications to foetus 3 = Premature delivery which include long term 4 = Neurological disability complications to foetus 6 = Other options 0 = Nothing Q17: W hen you were a child, did you ever observe this kind of behaviour between 1 = Yes, parents 0 = No your parents or other members of your household? 2 = Yes, other members Q18: Th ink of 10 of your closest neighbours, how many of them do you know, or think, experienced this kind of behaviour over the last 12 months? _______________ out of 10 98 = Don’t know CODES FOR Q6 and 37: WHO DID THIS TO YOU? AND WHO DID YOU DO THIS TO? CODES FOR Q8 and 34: HOW OFTEN HAS THIS HAPPENED? 1 = Partner 11 = Other relatives who do not live in this household but in the 1 = Once only 6 = Once a week 2 = Former partner community 2 = Once in a 7 = 2–3 times a week 3 = In-laws/partner’s family 12 = Other relatives who live outside this community while 8 = Almost every day 4 = Parents 13 = Friend 3 = Less than 9 = Every day 5 = Sons, daughters 14 = Teacher once per month 6 = Brothers, sisters 15 = Employer 4 = Once per 7 = Uncles, aunts 16 = Classmates month 8 = Grandparents 17 = Work colleagues 5 = 2–3 times per 9 = Cousins, nieces, nephews, 18 = Police month other relatives 19 = Other people in the community 10 = Priest or religious leader 20 = Other people outside the community 333 CODES FOR Q10: WHAT TYPE OF CODES FOR Q11: DID YOU CODES FOR Q12: HOW HAS THIS AFFECTED YOU? (mark all INJURY EVER FIGHT BACK? that apply) 1 = Cuts, punctures, 6 = Ear or eye injury 1 = Never 0 = Not at all 6 = I could not go to school bites 7 = Fractures, broken 2 = Once or twice 1 = Unable to concentrate at work 7 = I couldn’t do domestic 2 = Scratch, bones 3 = Several times 2 = Unable to concentrate at school work abrasion, bruises 8 = Broken teeth 4 = Many times/most of the time 3 = Unable to concentrate at 8 = Lost confidence in own 3 = Sprains, 9 = Other domestic work ability dislocations 4 = Could not take care of the 9 = Left work 4 = Burns children 10 = Left school 5 = Penetrating 5 = I had to miss work 11 = Live in constant fear injury, deep cuts, 12 = Other gashes 334 If respondent has been injured (Yes to ANY IN Q9) READ: I would like to learn more about the injuries resulting from the physical violence you have experienced. Q19: Ov er the last 12 months, how many 1 = Once/twice times were you injured due to 2 = 3–5 times physical violence? 3 = Many (more than 5) times Q20: Ov er the last 12 months, were you 1 = Times needed health care: ________ ever hurt badly enough due to 2 = Yes, but don’t know physical violence that you needed 3 = Not needed >>> NEXT MODULE health care? IF YES: How many times? Q21: Di d anyone stop you from trying to 1 = Yes receive health care for these injuries? 0 = No >>>>> Q26 Q22: W as the person who stopped you 1 = Yes from your household? 0 = No >>>>> Q24 Q23: W ho stopped you from your Use Household IDs household? Q24: W ho stopped you from outside your Use perpetrator codes household? Q25: Ho w many times has someone stopped you from receiving health __________________ care over the last 12 months? Q26: Ov er the last 12 months, have you 1 = Times needed health care: ________ ever received health care these 2 = Yes, but don’t know injuries? 3 = Not needed >>> NEXT MODULE IF YES: All of the time, or sometimes? Q27: Ho w much did you spend on health care for your injuries including medicines, fees and transport (IF MULTIPLE VISITS: during your last health visit)? Q28: Fo r your injury, did you have to Number of nights in health facility: spend any nights in any health ________ facility over the last 12 months? If none, enter ‘00’ IF YES: How many nights (in total)? Q29: Du ring your last health care visit, 1 = Yes >>> NEXT MODULE did you tell a health worker the real 0 = No cause of your injury? Q30: Du ring your last health care visit, 1 = I was afraid they would tell someone why did you decide not to tell the 2 = I was afraid of repercussions from the person real cause of your injury? who hurt me 3 = I was afraid they would not believe me 4 = I was afraid they would not care MULTIPLE ANSWERS ALLOWED 5 = I was afraid they would treat me badly 6 = Some of the people who work there know me or my family 7 = Other 335 If respondent has reported no physical violence, Read out loud: I know that these questions are very personal and some people would rather not reveal this sensitive information to a stranger. So we are giving everyone, no matter what they said before, the chance to reveal this information privately and discreetly without me watching you. No matter what you answered before, I am going to give you the questionnaire, and let you choose your responses without me looking. You can turn the page over as soon as you are done Q31: Re ad to respondent: Has anyone ever hurt you physically by doing any of the things that I have just mentioned? Now hand questionnaire to respondent. 336 Please can you tell me if you have a) b) Pushed c) Hit d) Kicked, e) Choked f) Burnt g) Used a h) Poured i) Kicked ever done or tried to do the Slapped or shoved someone dragged or someone gun, knife hazardous or pulled following things to someone else: or thrown someone? with your or beaten strangled on or other chemicals someone’s Clarify: Anyone includes your somethin fist or someone someone purpose? weapon or external partner, members of your family, g at with up? on against substance genitalia? work colleagues etc. someone somethin purpose? someone? s (e.g. that could g that acid) on hurt could hurt someone? them? them? Q1: 1 = Yes 2 = I tried >>> NEXT ITEM 0 = No >>> NEXT ITEM Q2: Ha s this happened in the last 12 months? 1 = Yes, 0 = No >>> NEXT ITEM Q3: Ov er the last 12 months, how often has this happened? Use codes above Q4: Ov er the last 12 months, was the person you did this to from your household? 1 =Yes, 0 = No >>>>Q37 Q5: Ov er the last 12 months, who did you do this to from your household? Use Household IDs Q6: Ov er the last 12 months who did you do this to? Use codes above 337 Q7: W hat was the sex of this person? 1 = Male, 2 = Female Q1: Please imagine a nine-step ladder, where on the bottom, the first step, are those who are totally unable __________________ to change their lives, while on step 9, the highest step, stand those who have full control over their own life. On which step are you? How much do you agree with the following 1 = Strongly 2 = Somewhat 3 = 4 = Somewhat 5 = Strongly statements? agree agree Neutral disagree disagree Q2: In uncertain times, I usually expect the best Q3: It’s easy for me to relax Q4: If something can go wrong, it will Q5: I’m always optimistic about my future Q6: I enjoy my friends a lot Q7: It’s important for me to keep busy Q8: I hardly ever expect things to go my way Q9: I don’t get upset too easily Q10: I rarely count on good things happening to me Q11: O verall, I expect more good things to happen to me than bad 338 Please can you tell me if anyone a) Insulted, b) Threatened to c) Done things to scare d) Threatened to e) Threatened to has ever (even when you were a belittled, or abandon you, or intimidate you on use a gun, knife or hurt you or child) done, or tried to do, the humiliated you in ignored you or purpose? (For example other weapon someone you following things to you: private or in front treated you by breaking something against you? care about with Clarify: Anyone includes your of other people indifferently in front of you) something other partner, members of your household, than a weapon? family, work colleagues etc. This includes threats to take away children or someone you care about Q1: 1 = Yes 2 = Someone tried >>> NEXT ITEM 0 = No >>> NEXT ITEM Q2: Di d this happen to you when you were younger than 15? 1 = Yes, 0 = No Q3: Ha s this happened in the last 12 months? 1 = Yes, 0 = No >>> NEXT ITEM Q4: Di d this behaviour make you feel worthless or make you doubt yourself? 1 = Yes, all the time 2 = Yes, sometimes 3 = No, not at all Q5: In the last 12 months, was the person who did this to you from your current household? 1 = Yes all of them, 2 = Some of them, 0 = No >>Q7 339 Q6: In the last 12 months, who did this to you from your current household? Use IDs from Household roster Then skip to Q9 Q7: In the last 12 months, who did this to you? Use codes above Q8: W hat was the sex of this person? 1 = Male, 2 = Female Q9: Ov er the last 12 months, how often has this happened? Use codes above Q10: W hen you were a child, did you ever observe this kind of behaviour between 1 = Yes, parents your parents or other members of your household? 2 = Yes, other members 0 = No Q11: Th ink of 10 of your closest neighbours, how many of them do you know, or think, experienced this kind of behaviour over the last 12 months? _______________ out of 10 98 = Don’t know CODES FOR Q6 and 18: WHO DID THIS TO YOU? AND WHO DID YOU DO THIS TO? CODES FOR Q9 and 15: HOW OFTEN HAS THIS HAPPENED? 1 = Partner 9 = Cousins, nieces, nephews, 13 = Friend 1 = Once only 6 = Once a week 2 = Former partner other relatives 14 = Teacher 2 = Once in a while 7 = 2–3 times a week 3 = In-laws/partner’s 10 = Priest or religious leader 15 = Employer 3 = Less than once per 8 = Almost every family 11 = Other relatives who do not 16 = Classmates month day 4 = Parents live in this household but in the 17 = Work colleagues 4 = Once per month 9 = Every day 5 = Sons, daughters community 18 = Police 5 = 2–3 times per month 6 = Brothers, sisters 12 = Other relatives who live 19 = Other people in the community 7 = Uncles, aunts outside this community 20 = Other people outside the 8 = Grandparents community 340 Please can you tell me if you have a) Insulted, b) Threatened to c) Done things to d) Threatened to e) Threatened to ever done, or tried to do, the belittled, or abandon someone, scare or intimidate use a gun, knife or hurt someone or following things to someone else: humiliated ignored someone or someone on other weapon someone they Clarify: Anyone includes your someone in treated them purpose? (For against someone? care about with partner, members of your family, private or in front indifferently example by breaking something other work colleagues etc. of other people things in front of than a weapon? someone) This includes threats to take away children or someone they care about Q1: 1 = Yes 2 = I tried >>> NEXT ITEM 0 = No >>> NEXT ITEM Q2: Ha s this happened in the last 12 months? 1 = Yes, 0 = No >>> NEXT ITEM Q3: Ov er the last 12 months, how often has this happened? Use codes above Q4: Ov er the last 12 months, was the person you did this to from your household? 1 =Yes, 0 = No >>>>Q17 Q5: Ov er the last 12 months, who did you do this to from your household? Use Household IDs Q6: Ov er the last 12 months who did you do this to? Use codes above Q7: W hat was the sex of this person? 1 = Male, 2 = Female 341 Q1: Do you have a bank account? 1 = Yes, 0 = No >>>>>>Q3 Q2: Ho w much savings do you have in your bank account? _________________ Q3: Do you use mobile banking? 1 = Yes, 0 = No >>>>>Q5 Q4: Ho w much savings do you have in your mobile banking account? _________________ Q5: Do you have any other form of savings? 1 = Yes, 0 = No >>>>>Q7 Q6: Ho w much are these total savings? __________________ Please can you tell me if anyone a) Refused to b) Taken c) d) e) f) Forced you g) Refused to has ever (even when you were a give you cash or Controlled Destroyed Prohibited to work give you or child) done, or tried to do, the enough chop withdrawn your own or damaged you from against your denied you food following things to you: money even money from belonging property working or will? or other basic Clarify: Anyone includes your though you your bank s and/or that you forced you needs? partner, members of your household, think he/she account or your have to quit your family, work colleagues etc. has enough other savings spending material work? money to without decisions? interest in? spend on other permission? things? Q7: 1 = Yes 2 = Someone tried >>> NEXT ITEM 0 = No >>> NEXT ITEM Q8: Ha s this happened in the last 12 months? 1 = Yes, 0 = No >>> NEXT ITEM Q9: In the last 12 months, was the person who did this to you from your current household? 1 = Yes all of them, 2 = Some of them, 0 = No >>Q11 342 Q10: In the last 12 months, who did this to you from your current household? Use IDs from Household roster Then skip to Q7 Q11: In the last 12 months, who did this to you? Use codes below Q12: W hat was the sex of this person? 1 = Male, 2 = Female Q13: Ov er the last 12 months, how often has this happened? Use codes above Q14: W hen you were a child, did you ever observe this kind of behaviour between 1 = Yes, parents your parents or other members of your household? 2 = Yes, other members 0 = No Q15: Th ink of 10 of your closest neighbours, how many of them do you know, or think, experienced this kind of behaviour over the last 12 months? _______________ out of 10 98 = Don’t know CODES FOR Q11 and 21: WHO DID THIS TO YOU? AND WHO DID YOU DO THIS TO? CODES FOR Q13 and 18: HOW OFTEN HAS THIS HAPPENED? 1 = Partner 9 = Cousins, nieces, nephews, 15 = Employer 1 = Once only 6 = Once a week 2 = Former partner other relatives 16 = Classmates 2 = Once in a while 7 = 2–3 times a 3 = In-laws/partner’s 10 = Priest or religious leader 17 = Work colleagues 3 = Less than once per week family 11 = Other relatives who do not 18 = Police month 8 = Almost every 4 = Parents live in this household but in 19 = Other people in the community 4 = Once per month day 5 = Sons, daughters the community 20 = Other people outside the 5 = 2–3 times per month 9 = Every day 6 = Brothers, sisters 12 = Other relatives who live community 7 = Uncles, aunts outside this community 8 = Grandparents 13 = Friend 14 = Teacher 343 Please can you tell me if you a) Refused to b) Taken c) d) Destroyed e) Prohibited f) Forced g) Refused have ever done, or tried to do, give someone cash or Controlled or damaged someone someone to to give or the following things to someone enough chop withdrawn someone’s property that from work against denied else: money even money from belongings someone has working or their will? someone Clarify: Anyone includes your though you someone’s and/or their a material forced food or other partner, members of your family, have enough bank account spending interest in? someone to basic needs? work colleagues etc. money to or other decisions? quit their spend on savings work? other things? without permission? Q16: 1 = Yes 2 = I tried >>> NEXT ITEM 0 = No >>> NEXT ITEM Q17: Ha s this happened in the last 12 months? 1 = Yes, 0 = No >>> NEXT ITEM Q18: Ov er the last 12 months, how often has this happened? Use codes above Q19: Ov er the last 12 months, was the person you did this to from your household? 1 =Yes, 0 = No >>>>Q7 Q20: O ver the last 12 months, who did you do this to from your household? Use Household IDs Q21: O ver the last 12 months who did you do this to? Use codes above Q22: W hat was the sex of this person? 1 = Male, 2 = Female 344 Q1: Over the last 12 months, have you ever left your home, even if 1 = Yes only for the night, because of a problem or situation at home? 0 = No >>> NEXT MODULE Q2: Was this following any of the treatment or behaviour we have 1 = Yes previously discussed? 0 = No >>>>Q4 Can you tell me which Produce list of violence and behaviour that respondent treatment or behaviour in has said have happened to him/her over last 12 months particular? Mark all that apply Q3: What were the reasons you left 0 = No particular reason the last time? 1 = Encouraged by friends/family 2 = Could not endure more MARK ALL MENTIONED 3 = Badly injured 4 = Afraid person/persons would kill me 5 = The person/persons who did this to me threatened or tried to kill me 6 = The person/persons who did this to me threatened or hit the children 7 = Saw that children are suffering 8 = Thrown out of the home 9 = Encouraged by organisation 10 = Encouraged by health workers/teachers 11 = Encouraged by police 12 = Other Q4: How long did you stay away 1 = Number of days (if less than one month) the last time? _____________ RECORD NUMBER OF 2 = Number of months (if 1 month or more) DAYS OR MONTHS _____________ 3 = Left or did not return to person who did this to me >>>> Next module Q5: Why did you return? 1 = Didn’t want to leave children 2 = Sanctity of marriage MARK ALL MENTIONED 3 = For sake of family/children 4 = Couldn’t financially support children 5 = Loved person who did this to me 6 = Person who did this to me asked me to come back 7 = Encouraged to return by family 8 = Encouraged to return by health workers/teachers 9 = Encouraged to return by police 10 = Forgave person who did this to me 11 = Thought person who did this to me would change 12 = Person who did this to me threatened partner/children 13 = Ran out of money 14 = Do not have another home to go to 15 = I was concerned about safety of people I was with 16 = I was concerned about safety of people left behind 17 = Other 345 Verify if respondent has ever experienced any 1 = Yes violence? 0 = No >>>>> Q8 Q1: After experiencing any of the types of violence 1 = Yes >>>>>>>Q3 and treatment discussed previously, have you 0 = No ever tried to approach a person or organisation for help or support? Q2: Why not? 1 = Did not know who to ask 8 = Violence normal/not 2 = Person/organisation too far serious THEN GO TO 3 = Nobody can help me 9 = Embarrassed/ashamed Q12 4 = I don’t need help 10 = Afraid would not be 5 = If my partner finds out it will believed cause trouble 11 = Afraid would be blamed 6 = If other family member finds out 12 = Afraid would end it will cause trouble relationship 7 = I don’t trust anyone 13 = Afraid would lose children 14 = Bring bad name to family 15 = Other Q3: What made you 1 = Encouraged by friends/family 6 = Partner threatened or hit approach 2 = Could not endure more the children someone? 3 = Badly injured 7 = Saw that children are 4 = Afraid partner would kill me suffering 5 = Partner threatened or tried to kill 8 = Thrown out of the home me 9 = Encouraged by organisation 10 = Other Q4: Q5: Q6: Q7: Who did you approach? How satisfied were Did you have How much you with the way to pay did you have that they dealt with anything to to pay? your case? this person or 1 = Completely organisation? satisfied 1 = Yes, I paid 2 = Somewhat a fee satisfied 2 = Yes, I paid 3 = Somewhat a bribe dissatisfied 0 = No >>>> 4 = Completely Next Item dissatisfied A A family member or friend B Health centre/hospital/other health service provider C Police D Domestic Violence and Victim Support Unit (DOVVSU) E Shelter F NGO, CSO or social worker G Lawyer or member of the Court 346 H Legal aid group I Queen mother of your community J Traditional leaders K Religious leaders L Religious group M Community group N Other Thank you very much for answering these questions. We are almost at the end, and we have only three more topics we would like to briefly discuss. First, we would like to ask you about access to services in your area. Q8: How long does it take you to get to your ________ minutes nearest police station? 98 = Don’t know Circle one: 1 = Foot 2 = Car 3 = Bicycle 4 = Motorcycle 5 = Tricycle 6 = Boat/canoe 7 = Public transport (e.g. bus or tro-tro) 8 = Truck 9 = Animal drawn cart 10 = Tractor Q9: Does your nearest police station have a 1 = Yes >>>>Q11 domestic violence victim support unit 0 = No (DOVVSU)? 98 = Don’t know Q10: D o you know where the closest domestic 1=Yes violence victim support unit (DOVVSU) is? 0=No Q11: I f not, how long would it take you to get to ______________ minutes the nearest domestic violence victims support 98 = Don’t know unit (DOVVSU)? Circle one: 1 = Foot 2 = Car 3 = Bicycle 4 = Motorcycle 5 = Tricycle 6 = Boat/canoe 7 = Public transport (e.g. bus or tro-tro) 8 = Truck 9 = Animal drawn cart 10 = Tractor How much would you trust the following if a friend 1 = Completely trust needed help after being physically assaulted in the home 2 = Somewhat trust by their partner? 3 = Somewhat distrust 4 = Completely distrust 98 = Never heard of them 347 Q12: H ealth centre/hospital/other health service provider Q13: P olice Q14: D omestic Violence and Victim Support Unit (DOVVSU) Q15: S helter Q16: N GO, CSO or social worker Q17: L awyer or member of the Court Q18: L egal aid group Q19: Q ueen mother of your community Q20: T raditional leaders Q21: R eligious leaders Q22: R eligious group Q23: C ommunity group Q24: D oes Ghana have a law 1 = Yes 4 = No >>>>>>NEXT against domestic violence? 2 = Yes, but I do not MODULE understand it 98 = Don’t know 3 = Yes, but I do not know >>>>NEXT MODULE what is in there exactly Q25: I f yes, where have you heard 1 = Friends/family 7 = School about it? 2 = Neighbours 8 = TV 3 = Community leaders 9 = Radio 4 = Religious leaders 10 = Newspaper 5 = Organisation 11 = Books 6 = Police 12 = Other I would now like to ask you a few questions about your health and use of health services. Q1: In general, would you describe your 1 = Very good 4 = Poor overall health as …? 2 = Good 5 = Very poor 3 = Fair Q2: Do you have any chronic conditions? 1 = Yes; 0 = No >>>>>Q4 Q3: Please indicate which 1 = Diabetes 6 = Cancer chronic condition(s) 2 = Asthma 7 = High blood pressure you have? 3 = Lung disease 8 = Physical disability which restricts 4 = Heart disease certain activity MULTIPLE 5 = Arthritis or other 9 = Mental disability which restricts ANSWERS rheumatic disease certain activity 10 = Other chronic condition Q4: Which of the following describes your 1 = I used to smoke but not anymore smoking habit? 2 = I currently smoke >>> SKIP TO Q6 0 = I never smoked >>> SKIP TO Q7 1 = Daily 2 = Occasionally Q5: Did you use to smoke daily or occasionally? Q6: Do you smoke daily or occasionally? Q7: Do you drink alcohol? 1 = Yes; 0 = No >>>>if in relationship Q9 if not Q11 Q8: How often do you drink 1 = Every day or nearly every day 4 = 1–3 times a month alcohol? 2 = More than half the week but not 5 = Less than once a everyday month 3 = Once or twice a week 348 Q9: Does your partner drink alcohol? 1 = Yes, 0 = No >>>> Q11 Q10: H ow often does your 1 = Every day 4 = 1–3 times a month partner drink alcohol? 2 = More than half the week but not 5 = Less than once a every day month 3 = Once or twice a week Q11: H as anyone ever stopped you from or did not allow you to receive 1 = Yes health care when you needed it? Clarify: This could be by not allowing 0 = No >>>>> respondent to visit health care facility or refusing to pay the fees or Q17 costs of medicine. Q12: H as this happened over the last 12 months? 1 = Yes; 0 = No >>>> Q17 Q13: W as the person who stopped you from your household (over 1 = Yes; 0 = No >>>>> Q15 the last 12 months)? Q14: W ho stopped you from your household (over the last 12 months)? Use Household IDs Q15: W ho stopped 1 = Partner 10 = Other relatives who do not live in this you from 2 = Former partner household but in the community outside your 3 = In-laws/partner’s family 11 = Other relatives who live outside this household 4 = Parents community (over the last 5 = Sons, daughters 12 = Friend 12 months)? 6 = Uncles, aunts 13 = Teacher 7 = Grandparents 14 = Employer 8 = Cousins, nieces, nephews, 15 = Work colleagues other relatives 16 = Police 9 = Priest or religious leader 17 = Other people in the community 18 = Other people outside the community Q16: H ow many times has someone stopped you from receiving health care over the last 12 months? ___________ Q17: I n the last 30 days, have you felt sick or unwell? 1 = Yes; 0 = No >>>Q20 Q18: I n the last 30 days, did you personally 1 = No one 7 = Midwife consult a doctor or other professional 2 = Doctor 8 = Traditional birth or health worker or traditional healer 3 = Nurse attendant because you were sick? Clarify: This 4 = Counsellor 9 = Herbalist does not include health check-ups. 5 = Traditional healer 10 = Spiritualist IF YES: Whom did you consult? 6 = Pharmacist/chemical 11 = Prayer camp MULTIPLE ANSWERS ALLOWED seller 12 = Other Q19: W ere you unable to work because of your sickness? IF YES: How many days? _________ During the last 30 days, about how All the Most Some A little None often did … time of the of the of the of the time time time time Q20: y ou feel nervous? Q21: y ou feel hopeless? Q22: y ou feel restless or fidgety? Q23: y ou feel that everything was an effort? Q24: y ou feel so depressed that nothing could cheer you up? Q25: y ou feel worthless? Q26: h as the thought of ending your life been on your mind? If there are children in the household aged 14 and younger 349 Adults use certain ways to teach children the right behaviour or to address a behaviour problem. I will read various methods that are used and want you to tell me if you have used this method with any children (aged 14 and younger) in your household in the past month. In the last month, have you or anyone else in your household 1 = Yes, 0 = No Q1: Taking away privileges Q2: F orbidding something that he/she liked Q3: Not allow him/her to leave the house Q4: Explaining why the behaviour was wrong Q5: Giving him/her something else to do Q6: I gnoring/refusing to communicate to the child Q7: Shouting, yelling at or screaming at him/her Q8: Calling him/her dumb, lazy or another name like that Q9: Locking child up in room Q10: S haking him/her Q11: S panking, hitting, pushing or slapping him/her on the bottom with bare hand Q12: H itting him/her on the bottom or elsewhere on the body with something like a belt, comb, hairbrush, cane, stick or other hard object Q13: H itting or slapping him/her on the face, head or ears Q14: H itting or slapping him/her on the hand, arm or leg Q15: B eating him/her up, that is hitting him/her over and over as hard as one could Q16: P hysically hurting them in any other way (e.g. burning) Finally, I would like to ask you a few questions about how the children of this household spend their time. Q17: H ow many hours do you spend with your child/children (aged 14 and below) each day? 98 = Don’t know 99 = Not applicable Q18: H ow many hours do you spend supervising your child/children (aged 14 and below) during homework 98 = Don’t know each day? Should be less than Q17 99 = Not applicable Q19: H ow many hours does your child spend with other adult household members? 98 = Don’t know 99 = Not applicable Q20: D o you regularly attend Parent-Teacher Association 1 = Always (PTA) meetings? 2 = Often 3 = Sometimes 4 = Rarely 5 = Never 99 = Not applicable Q21: D oes someone else from this household regularly 1 = Yes attend PTA meetings? 0 = No 350 Thank you very much for your time! We know it is not easy to answer all these questions, and we really appreciate that you took the time and patience. Since we have talked about issues related to violence, we recommend that you do not speak to others about this part of the questionnaire. Otherwise, some people in the community might get upset with you or others who answered the survey. Are there any questions you have for us, or is there anything that you would like to let us know? _________________________________________________ 351 Q3: Qualitative instruments 1. Focus Group Discussion Guideline – Domestic Violence Ghana 2015 Approach: For four separate and mixed FGDs with men and women. 1:30 hours /8–12 people Group 1: Married/living together men (aged 18–50) Group 2: Married /living together women (aged 18–50) Group 3: Unmarried/youth (aged 18-25) Group 4: Mixed (aged 50+) Facilitator introduces the team, explains consent and the purpose Thanks everyone for coming. Good morning/afternoon/evening and welcome. Thank you all for taking the time to be part of this discussion. We very much appreciate your willingness to participate. My name is [name of facilitator], and I will serve as the moderator for today’s focus group discussion. Assisting me is [name of assistant], who will take notes. This study is part of the National Survey on Domestic Violence 2015 being conducted by the Institute of Development Studies, University of Sussex (IDS) for the Ministry of Gender, Children and Social Protection and DfiD, UK. Participation in this research is voluntary. The discussion may last for about 45 to 90 minutes. All information provided will be kept secret. We will not ask your real name and/or your address. You can withdraw from this interview anytime you want, or refuse to answer any questions that you are not comfortable with. The purpose of today’s discussion is to learn from you about:  The causes and consequences of domestic violence  Which services survivors of domestic violence use  What works well, what does not  How these services can be improved or strengthened  The findings will also inform civil society organisations and help them to reach out to survivors of domestic violence and lobby for more programmatic interventions and changes in laws.  There will be a report that will be published internationally. We will be spending a few days in the community trying to find out about the different roles of women and girls and men and boys in the community and in the family, and discuss whether and how views and experiences with domestic violence has changed compared to the past. 352 Facilitator explains the consent forms and the value of recording, and checks permission. Facilitator and Note-taker help participants complete consent forms. Note-taker explains that we give people a number and use the numbers for the note- taking. Facilitator asks group for permission to record the discussion. Rules We would like you to feel safe here. We are here to learn from you; you have the knowledge to help us understand these issues. We are not going to judge you and will listen to everything you have to say. To do this, we would like to employ some guidelines for the session to make sure we are safe and have a good discussion; the guidelines will apply to all of us. 1. THERE ARE NO RIGHT OR WRONG ANSWERS Every person’s experiences and opinions are important. Speak up whether you agree or disagree. It’s ok to have a different opinion; but everyone’s opinion matters equally. We are interested to hear everybody’s opinion. 2. WE WOULD LIKE YOU TO DO THE TALKING. Don’t feel like you have to respond to me all the time. However, we would like everyone to participate. So if you’re talking a lot, I may ask you to give others a chance. And, I may call on you if I haven’t heard from you in a while. 3. WHAT IS SAID IN THIS ROOM STAYS HERE Your experiences and input is very important and will feed into the research report – but not what you say individually, unless you would like to be named. We have name tents here in front of us today so that we can all know who’s sitting here with us, but no names will be included in any reports! Discussion starts. Facilitators start with the questions. They do not read the headings. Community’s perception of violence: What is violence? 1. Is it acceptable if a man beats his wife for disagreeing with him? (Allow different opinions and allow participants to state reasons for their answers.) If yes, why? ………………………………… If no, why? …………………………… 353 2. Is it acceptable if a woman beats her husband for having a child with someone outside the family? (Allow different opinions and allow participants to state reasons for their answers.) If yes, why? ………………………………… If no, why? …………………………… 3. Is it acceptable if a man beats his wife for refusing to have sex with him? (Allow different opinions and allow participants to state reasons for their answers.) If yes, why? ………………………………… If no, why? …………………………… 4. Is it acceptable if a man refuses to give ‘chop money’ to his wife? (Allow different opinions and allow participants to state reasons for their answers.) If yes why? ………………………………… If no, why? …………………………… 5. Is it acceptable for men to tell women they cannot work? If yes why? ………………………………… If no, why? …………………………… 6. Is it acceptable for men to beat children and youth in the home? If yes why? ………………………………… If no, why? …………………………… 7. Is it acceptable for women to beat children and youth in the home? If yes why? ………………………………… If no, why? …………………………… 8. Is it acceptable for fathers to stop their daughters from attending school? If yes, why? …………………………………………………….. If no, why? ……………………………………………………….. When should one intervene? Allow different opinions and allow participants to state reasons for their answers. 9. What would you do if you were passing along a courtyard in your neighbourhood and saw a man and woman shouting at each other? 10. What would you do if you saw the man now beating the woman? 11. What would you do if you saw a woman now beating the man? 12. What would you do if you saw the man/woman take out an object? What would you do if you saw the woman/man now hurting the man/woman with the object? Causes of domestic violence 13. Do we know whether there is disagreement and fights by people/men and women from the same family in this community? If Yes, probe: [Who is involved in these fights? What is the fighting about? How often do we hear or see these fights?] 14. So, why is there fighting within our families in this community 15. What are the various types of disagreement within families? 16. What typically happens during these fights/disagreements? (Describe these fights.) 354 17. Do men also experience violence in the family? What does this look like? How is this different from how women experience violence in the domestic context? 18. Are weapons involved? (If yes, what type of weapons?) 19. Who is affected by these fights? How? Why? (Probe: children, elderly members, others in family?) 20. What is the immediate impact? And what is the long-term impact on those affected? Intervention and mediation 21. How do these domestic fights come to an end? Who do people go to for help? Who is involved? 22. In your community, what kind of services are there to help people who are hurt by their family? (Probe: list the services to understand the ‘in principle’ facets of government support.) 23. Which of these services do people use the most? Why? 24. What has been your experience with government support for people hurt by other people in their family? [Probe and discuss satisfaction with police handling of case reporting/help at health centres.] Social cost of domestic violence 25. Are others in the community affected by a family’s occasions of violence/blow-up? Identify who. How are these persons affected? [Probe: different categories of persons identified and how they are impacted by the violence.] 26. What are the costs to the rest of the community by a family’s experience of violence? What would make violence acceptable to people/the community? 27. What would make members of the family turn a blind eye to violence within a family? 28. What are the expectations of masculinity (breadwinner, disciplinarian) that can make men feel they need to behave in a certain dominant or violent way? 29. What are the expectations of femininity (submissiveness) that can make women feel they need to accept certain dominant or violent ways from men? 30. Discuss what are social pressures that can lead to people hurting one another in a family or household – family and community pressure to encourage or discourage prosecution. 31. What are the government pressures to encourage or discourage prosecution of cases? N.B. These are just guidelines. If any interesting additional issues emerge, the discussion will also explore these. Facilitator asks participants if they have any questions for us. Facilitator thanks participants for their time and the interesting information they have shared. 355 2. Observation and In-Depth Interview Guideline on Domestic Violence Ghana: Health Care Providers Approach: Observations and in-depth interviews at health facilities (two per region and >40 staff. Two staff per site). In each urban and rural site, health facilities that provide emergency care for victims of domestic violence are visited. In each of these sites, both observations and interviews should take place. Purposive sample of individuals who are directly involved in domestic violence. These can be leaders or front-line staff or both depending on the facility. Facilities are equally divided across selected urban and rural sites. Facilitator introduces the team, explains consent and the purpose. Thanks for giving us your time. We very much appreciate your willingness to participate. My name is [name of facilitator], and I will serve as the interviewer for today’s discussion. Assisting me is [name of assistant], who will take notes. This study is part of the National Survey on Domestic Violence 2015 being conducted by the Institute of Development Studies, University of Sussex (IDS) for the Ministry of Gender, Children and Social Protection and DfiD, UK. It covers all 10 regions of Ghana, urban and rural areas. Participation in this research is voluntary. The discussion may last for about 30 to 60 minutes. All information provided will be kept secret. We will not ask your real name and/or your address. You can withdraw from this interview anytime you want, or refuse to answer any questions that you are not comfortable with. The purpose of today’s visit and discussion is to learn from you about:  The causes and consequences of domestic violence, and which services are available to survivors of domestic violence  What works well, what does not  How these services can be improved or strengthened  The findings will inform policy recommendations to the Ministry of Gender, Children and Social Protection. It will also enable civil society organisations to reach out more directly to survivors of domestic violence and lobby for more programmatic interventions and changes in laws.  There will be a report that will be published. Facilitator explains the consent forms and the value of recording, and checks permission. Facilitator and Note-taker help to complete the consent form. Note-taker fills out the clip-on sheet. Facilitator asks for permission to record the discussion. 356 Note to Facilitator: You start with broad opening questions to help us understand some more detail about the participant. It’s important to learn more about the person’s work; this will help us to understand what they say, and what they might not feel comfortable saying. It helps to understand the language that the participant will use. Discussion starts. Overview, background: tell me about yourself 1. Can you tell me about the work you do? Probe: a. As a health professional, what are some of your main responsibilities? b. As a health professional, what are some of the main challenges in your work? c. As a health professional, do you see any problems with the way men treat women? With the way women treat men? With the way parents treat their children? Forms and causes of domestic violence 2. Can you tell me what you know about disagreement and fights by people from the same family? 3. What kinds of fights do you see? [Probe to find about forms: a. Are these fights verbal, physical, or sexually related? b. Who is involved? Probe for: partners, parents and children, adults and non- biological children?] 4. Why is there fighting/abuse and violence within families? [Probe to find out about causes: Are they about money, about infidelity, about school etc.?] 5. Based on your observations – as a professional – how often do you think it happens? [Probe: At least once a week or once a month?] 6. Are there specific times when this fighting takes place? [Probe: Any seasons or specific days? Is it linked to school terms? Is it linked to monthly payment cycles?] Impact of domestic violence 7. Who is affected by fights in the family? How? Why these people? [Probe: the range of effects – they include losing confidence, being afraid to leave the house, not being able to work and becoming trapped in a vicious cycle of dependence, for example. How are these effects distributed across members of the household, who are affected by the violence?] 8. Do men also experience violence in the family? What does this look like? How is this different from how women experience violence in the domestic context? 9. What is the immediate impact? And what is the long-term impact on those affected? [Probe for economic, psychological, health impacts.] 357 Acceptability of violence and social norms 10. What would make members of a family turn a blind eye to violence? 11. What are the expectations of masculinity (breadwinner, disciplinarian) that can make men feel they need to behave in a certain dominant or violent way? 12. What are the expectations of femininity (submissiveness) that can make women feel they need to accept certain dominant or violent ways from men? 13. What would make health officers turn a blind eye to reported cases of domestic violence? (i.e. discourage referral to police and other institutions for help) 14. Discuss – what are the social pressures – family and community pressure to encourage or discourage health officers’ referral to the police for prosecution. 15. What are the government pressures to encourage or discourage prosecution of reported cases of domestic violence? Intervention and mediation 16. How do these fights come to an end? Who do people go to for help? Who is involved? [Probe: to find out about government and non-government service providers: a. Are these government-appointed people (nurses, social workers, teachers)? b. Do people feel more comfortable going to community leaders than to police officers to get support? c. Do people feel more comfortable going to certain police (men/women, senior/junior) than other people/service providers? Why? d. What kinds of things matter in how people make decisions to get support? What catalyses these decisions?] 17. What kind of services are there to help people? [Probe to get a list of services to understand the ‘in principle’ facets of government support as the participant understands this.] 18. Which of these services do people use the most? Why? Which are not useful? 19. What has been your experience with government support for people hurt by other people in their family? (Discuss satisfaction with police handling of case reporting/help at health centres.) 20. What do you think needs to happen to: a. Help those people who commit violence to stop? b. Help people who experience violence to get support? c. Help government services reach the right people to end the violence? d. Which services are seen to be most crucial? Why? 21. How many health cases from domestic violence did you received in this health facility in:  2012 ……….………………………………………………………………  2013 ……….………………………………………………………………  2014 ……….……………………………………………………………… 22. How many of these cases were female victims? How many were male victims (Probe: ask for rough estimate on average/percentage.) 358 Availability of logistical support 23. Does the facility have specialised departments offer services to victims of violence such as:  Rape-related services such as post-exposure prophylaxis for HIV prevention (PEP), sexually transmitted infection (STI) referrals, and emergency contraceptives (‘morning-after pills’)?  Legal referrals including prosecution advice for victims of violence (Is this advice also given and tailored to men, and to gay men/women?)  Counselling services, mental health services  Any other (name service)……………………………………………………… Availability of protocols 24. Is domestic violence different from other forms of violence? Does the health unit have formal protocols to follow when dealing with victims of domestic violence?  Yes  No If Yes, name the documents ………………………………………………………….. If yes, what do these protocols deal with?  Emergency response to victims of violence  Priority for victims of abuse  Writing of medical reports  Referral to specialised units (paediatrician/social workers/police) Capacity-building 25. Sensitising the health professionals on domestic violence. Has health centre staff undergone:  Initial training on gender-based violence and on domestic violence  Continuing training on domestic violence  Specialised in listening skills for victims story  Training in writing medical reports Affordable health care 26. Are medical reports free? If No, How much is charged? ………………………………….. Is this a formal, required fee by the government or a personal fee charged by the doctor? …………………………………………………………………….. 27. How are victims who are unable to pay fee for medical reports able to obtain the report to forward to police for prosecution of the case? Continuum of care 28. Does this health site work with any other service providers? If yes, which? (Police, social workers) Name the organisations here ………………………………………………………………………………………..…….. 359 29. How effective is this collaboration? 30. Can you give me an example where this site collaborated with other service providers to give support to someone who had experienced domestic violence? [Probe: Who was involved? What did each ‘sector’ do? What were the outcomes for the person involved? Was the collaboration successful? What made it work well?] OBSERVATIONS While waiting look for: Are files stored away or can the public see these easily? Do you hear staff talk about cases and private affairs of patients in front of you when you are in a public area? Is there a private room or other privacy measures for patients when they speak with medical staff? Ask: Can you show us the building and the services as if we are a victim of domestic violence who comes to report? Can you show us the process step by step? Where would we go first, and what would we do? Who would help us? Where would we sit to report? When and where do I hear about PEP and emergency contraceptives? (Are there rape kits?) Note: If respondents mentioned special services such as rape-related services, legal referrals or counselling services, you can ask them to show you where these are. N.B. These are just guidelines. If any interesting additional issues emerge, the discussion will also explore these. Facilitator asks respondent if (s)he has any questions for us. Facilitator thanks respondent for her/his time and the interesting information they have shared. 360 3. Observation and In-Depth Interview Guideline on Domestic Violence Ghana: District Police/DOVVSU Institutions Approach: Observations and in-depth interviews at police facilities (two per region and >20 members of the police). In each urban and rural site, police stations are visited; in each of these sites, both observations and interviews should take place. Purposive sample of individuals who are directly involved in domestic violence. These can be leaders or front- line staff equally divided across selected urban and rural sites. Facilitator introduces the team, explains consent and the purpose. Thanks for giving us your time. We very much appreciate your willingness to participate. My name is [name of facilitator], and I will serve as the interviewer for today’s discussion. Assisting me is [name of assistant], who will take notes. This study is part of the National Survey on Domestic Violence 2015 being conducted by the Institute of Development Studies, University of Sussex (IDS) for the Ministry of Gender, Children and Social Protection and DfiD, UK. It covers all 10 regions of Ghana, urban and rural areas. Participation in this research is voluntary. The discussion may last for about 30 to 60 minutes. All information provided will be kept secret. We will not ask your real name and/or your address. You can withdraw from this interview anytime you want, or refuse to answer any questions that you are not comfortable with. The purpose of today’s visit and discussion is to learn from you about:  The causes and consequences of domestic violence, and which services are available to survivors of domestic violence  What works well, what does not  How these services can be improved or strengthened  The findings will inform policy recommendations to the Ministry of Gender, Children and Social Protection. It will also enable civil society organisations to reach out more directly to survivors of domestic violence and lobby for more programmatic interventions and changes in laws.  There will be a report that will be published. Facilitator explains the consent forms and the value of recording, and checks permission. Facilitator and Note-taker help to complete the consent form. Note-taker fills out the clip-on sheet. Facilitator asks for permission to record the discussion. 361 Note to Facilitator: You start with broad opening questions to help us understand some more detail about the participant. It’s important to learn more about the person’s work; this will help us to understand what they say, and what they might not feel comfortable saying. It helps to understand the language that the participant will use. Discussion starts. Overview, background: 1. Can you tell me about the work you do? Probes: a. As a police officer, what are some of your main responsibilities? b. As a police officer, what are some of the main challenges in your work? i. As a police officer, do you see any problems with the way men treat women? With the way women treat men? With the way parents treat their children? Forms and causes of domestic violence 2. Can you tell me what you know about disagreement and fights by people from the same family? 3. What kinds of fights do you see? [Probe to find about forms: a. Are these fights verbal, physical, or sexually related? b. Who is involved? [Probe for: partners, parents and children, adults and non- biological children.] 4. Why is there fighting/abuse and violence within families? [Probe to find out about causes. Are they about money, about infidelity, about school etc.] 5. Based on your observations – as a professional – how often do you think it happens? [Probe: At least once a week or once a month?] 6. Are there specific times when this fighting takes place? [Probe: Any seasons or specific days? Is it linked to school terms? Is it linked to monthly payment cycles?] Impact of domestic violence 7. Who is affected by fights in the family? How? Why these people? [Probe: the range of effects – they include losing confidence, being afraid to leave the house, not being able to work and becoming trapped in a vicious cycle of dependence, for example. How are these effects distributed across members of the household who are affected by the violence?] 8. Do men also experience violence in the family? What does this look like? How is this different from how women experience violence in the domestic context? 9. What is the immediate impact? And what is the long-term impact on those affected? [Probe for economic, psychological, health impacts, children’s lives.] 362 Acceptability of violence and social norms 10. What would make members of a family turn a blind eye to violence? 11. What are the expectations of masculinity (breadwinner, disciplinarian) that can make men feel they need to behave in a certain dominant or violent way? 12. What are the expectations of femininity (submissiveness) that can make women feel they need to accept certain dominant or violent ways from men? 13. Discuss what are social pressures – family and community pressure to encourage or discourage prosecution. 14. What are the government pressures to encourage or discourage prosecution of cases? Intervention and mediation 15. How do these fights come to an end? Who do people go to for help? Who is involved? [Probe: to find out about government and non-government service providers: a. Are these government-appointed people (nurses, social workers, teachers)? b. Do people feel more comfortable going to community leaders than to police officers to get support? c. Do people feel more comfortable going to certain police (men/women, senior/junior) than other people/service providers? Why? d. What kinds of things matter in how people make decisions to get support? What catalyses these decisions?] 16. What kind of services are there to help people? [Probe to get a list of services to understand the ‘in principle’ facets of government support as the participant understands this.] 17. Which of these services do people use the most? Why? Which are not useful? 18. What has been your experience with government support for people hurt by other people in their family? (Discuss satisfaction police handling of case reporting/help at health centres.) 19. What do you think needs to happen to: a. Help those people who commit violence to stop? b. Help people who experience violence to get support? c. Make sure government services reach the right people to end the violence? Which services are seen to be most crucial? Why? Availability of logistical support 20. Does the unit have specialised departments which offer services such as:  Advocacy for victims of violence  Shelter for victims of violence  Continuing Education Units  Counselling for victims of violence  Self-help advice for victims of violence  Protection for victims of violence  Prosecution advice for victims of violence 363  Batterer’s prevention counsellors  Any gender-specific support services (name service ……………………...) Availability of protocols 21. Does DOVVSU have formal protocols to follow when dealing with reporting of incidences of domestic violence?  Yes  No If Yes, name the documents ………………………………………………………… If yes, what do these protocols deal with?  Dispatch response  Arrests/service of orders  Investigation techniques  Evidence  Reports  Follow-ups  Victim’s rights  Rights perpetuators of violence  Emergency judicial response  Refuge/victim’s safety Capacity-building 22. Have DOVVSU staff received the following specialised training to provide specific help to the victims of violence?  Initial training on gender-based violence and on domestic violence  Continuing training on domestic violence  Domestic violence and sexual assault training  Specialised in listening skills for victim’s story Case processing and pace of justice 23. Does DOVVSU/police station have formal procedures to follow to process domestic violence cases?  If Yes, describe the procedure ..…………………………………………………... ………………………………………………………………………………………..  If No, how are cases of domestic violence processed? ………….…………………………………………………………………………….  How many cases have been reported in  2012  2013  2014  How many cases have been prosecuted?  2012  2013  2014 364 Continuum of care 24. Does this police facility work with any other service providers? If yes, who? (social workers). Name them …………....………………………………………………………………………………….. 25. How effective is this collaboration? 26. Can you give me an example of where this facility collaborated with other service providers to give support to someone who had experienced domestic violence? OBSERVATIONS While waiting, look for: Are files stored away or can the public see these easily? Do you hear staff talk about cases and private affairs of patients in front of you when you are in a public area? Is there a private room where victims can file a complaint? Ask: Can you show us the building and the services as if we are a victim of domestic violence who comes to report? Can you show us the process step by step? Where would we go first, and what would we do? Who would help us? Where would we sit to report? N.B. These are just guidelines. If any interesting additional issues emerge, the discussion will also explore these. Facilitator asks respondent if (s)he has any questions for us. Facilitator thanks respondent for her/his time and the interesting information they have shared. 365 4. In-Depth Interview Guideline – Domestic Violence Ghana Opinion/Community Leaders Approach: For in-depth interviews with opinion leaders, activists and legal experts: 20 community leaders (e.g. religious leaders, and formal and informal political leaders); 10– 20 legal practitioners; 10–20 activists (national or regional level); and 10–20 journalists (national or regional level). Purposive sample in all regions of individuals who have been in current position over two years, equally divided across selected urban and rural sites alike. Facilitator introduces the team (2 people maximum) and explains consent and the purpose. Thanks for giving us your time. We very much appreciate your willingness to participate. My name is [name of facilitator], and I will serve as the interviewer for today’s discussion. Assisting me is [name of assistant], who will take notes. This study is part of the National Survey on Domestic Violence conducted for the Ministry of Gender, Children and Social Protection. It covers all 10 regions of Ghana, urban and rural areas. We use both quantitative and qualitative methodologies. Participation in this research is voluntary. The discussion may last for about 30 to 60 minutes. All information provided will be kept secret. We will not ask your real name and/or your address. You can withdraw from this interview anytime you want, or refuse to answer any questions that you are not comfortable with. The purpose of today’s discussion is to learn from you about  The causes and consequences of domestic violence, and which services survivors of domestic violence use  What works well, what does not  How these services can be improved or strengthened  The findings will also inform civil society organisations and help them to reach out to survivors of domestic violence and lobby for more programmatic interventions and changes in laws.  There will be a report that will be published internationally. Facilitator explains consent forms and the value of recording, and checks permission. Facilitator and Note-taker help to complete the consent form. Note-taker fills out the clip-on sheet. Facilitator asks for permission to record the discussion. 366 Note to Facilitator: You start with broad opening questions to help us understand some more detail about the participant. It’s important to learn more about the person’s work; this will help us to understand what they say, and what they might not feel comfortable saying. It helps to understand the language that the participant will use. Discussion starts. Overview, background: tell me about yourself 1. Can you tell me about the work you do? [Probe: a. As a religious/political/community leader, what are some of your main responsibilities? b. As a religious/political/community leader, what are some of the main challenges in your work? c. As a religious/political/community leader, do you see any problems with the way men treat women? With the way women treat men? With the way parents treat their children?] Forms and causes of domestic violence 2. Can you tell me what you know about disagreement and fights by people from the same family? 3. What kinds of fights do you see? [Probe to find about forms: a. Are these fights verbal, physical, or sexually related? b. Who is involved? Probe for: partners, parents and children, adults and non- biological children] 4. Why is there fighting/abuse and violence within families? [Probe to find out about causes. Are they about money, about infidelity, about school etc.?] 5. Based on your observations – as a professional – how often do you think it happens? [Probe: At least once a week or once a month?] 6. Are there specific times when this fighting takes place? [Probe: Any seasons or specific days? Is it linked to school terms? Is it linked to monthly payment cycles?] Impact of domestic violence 7. Who is affected by fights in the family? How? Why these people? [Probe: the range of effects: losing confidence, being afraid to leave the house, not being able to work and becoming trapped in a vicious cycle of dependence, for example. How are these effects distributed across members of the household who are affected by the violence?] 8. Do men also experience violence in the family? What does this look like? How is this different from how women experience violence in the domestic context? 367 9. What is the immediate impact? And what is the long-term impact on those affected? [Probe for economic, psychological, health impacts.] Acceptability of violence and social norms 10. What would make members of a family turn a blind eye to violence? 11. What are the expectations of masculinity (breadwinner, disciplinarian) that can make men feel they need to behave in a certain dominant or violent way? 12. What are the expectations of femininity (submissiveness) that can make women feel they need to accept certain dominant or violent ways from men? 13. Discuss what are social pressures – family and community pressure to encourage or discourage prosecution. 14. What are the government pressures to encourage or discourage prosecution of cases? Intervention and mediation 15. How do these fights come to an end? Who do people go to for help? Who is involved? [Probe: to find out about government and non-government service providers: a. Are these government-appointed people (nurses, social workers, teachers)? b. Do people feel more comfortable going to community leaders than to police officers to get support? c. Do people feel more comfortable going to certain police (men/women, senior/junior) than other people/service providers? Why? d. What kinds of things matter in how people make decisions to get support? What catalyses these decisions?] 16. What kind of services are there to help victims of domestic violence? (Get a list of services to understand the ‘in principle’ facets of government support as the participant understands this.) 17. Which of these services do people use the most? Why? Which are not useful? 18. What has been your experience with government support for people hurt by other people in their family? (Discuss here satisfaction with police handling of case reporting/help at health centres.) Are there interventions to prevent domestic violence in this community? .………………………………………………………………………………………………. 19. Which of these interventions work against domestic violence? Why? 20. What can the government do better, or differently, to make sure that its policy is properly implemented? a. In rural areas? b. In peri-urban areas? c. In urban areas? d. Through: Police? Health workers? Social workers? 21. What do you think needs to happen to: a. Help those people who commit violence to stop? b. Help people who experience violence to get support? 368 c. Help government services reach the right people to end the violence? d. Which services are seen to be most crucial? Why? N.B. These are just guidelines. If any interesting additional issues emerge, the discussion will also explore these. Facilitator asks respondent if (s)he has any questions for us. Facilitator thanks respondent for her/his time and the interesting information they have shared. 369 5. In-Depth Interview Guideline – Victims/Survivors of Domestic Violence Ghana Approach: For in-depth interviews with 45 survivors of domestic violence (three per enumerator) identified during the quantitative research – the survey. Purposive sample in all regions of individuals who self-identify as victims of domestic violence, equally divided across urban and rural areas alike. Facilitator introduces him or herself, explains consent and the purpose. Thanks for giving us your time. We very much appreciate your willingness to participate. My name is [name of facilitator], and I will serve as the interviewer for today’s discussion. My colleague [name of note-taker], will take notes. This study is being conducted for the Ministry of Gender, Children and Social Protection. It covers all 10 regions of Ghana, urban and rural areas. Your participation in this research is voluntary. The discussion may last for about 30 to 60 minutes. All information provided will be kept secret. We will not ask your real name and/or your address. You can withdraw from this interview anytime you want, or refuse to answer any questions that you are not comfortable with. You were invited for this interview because you participated in a survey during which you self-identified as someone who has or is experiencing domestic violence. Thank you for agreeing to speak some more with us about your experiences. The purpose of today’s discussion is to learn from you about:  the causes and consequences of domestic violence;  which services are available to victims/survivors of domestic violence;  what works well for victims/survivors of domestic violence, and what does not; and  how these services can be improved or strengthened. There will be a report that will be published from this country-wide study; however, individual stories will not be identified, so no one will know that you spoke to us or what you told us. Facilitator explains about consent forms and the value of recording, and checks permission. Facilitator helps to complete the consent form. Discussion starts. Background information 1. Can you tell me a little bit about yourself? [Interviewer must ask and record the following background information for each interview: age; sex; marital status (single/married/living together/living together with children); educational level; region; district; city/town/rural site of interview.] 370 2. About your home, and who you live with? 3. About your family, who forms part of the house/household (nuclear or extended family compound household/ rented accommodation etc.); and who do you take care of/takes care of you in the home? 4. About your daily activities, and what takes up your time during the day (earning a living or supplementing income/household chores or combination of these etc.)? 5. About growing up, where you grew up and how you came to live here? Forms and causes of domestic violence 6. You told us that you experienced some violence from other family members of your household in your life? Is that correct? 7. Can you tell me a little bit more about this experience? [Interviewer should give the respondent the opportunity to briefly narrate uninterrupted their experience of domestic violence – notes should be taken of this narration.] 8. Is this situation of violence ongoing? Or was it in the past? When? 9. Let us go into more detail here. What kinds of fights and disagreement did/do you experience/see in your own family? [Probe to find about forms: Are these fights verbal, physical, sexual, economic or emotional/psychological?] 10. Who was/are involved? [Probe to find out about domestic relations: Partners, parents and children, siblings, aunts, uncles, cousins, grandparents, stepchildren and non-biological children, house help.] 11. Why does this fighting/abuse and violence take place in your family? [Probe to find out about causes: Are they about money, about infidelity, about school etc.?] 12. How often does this fighting and violence take place? [Probe for incidence: At least once a week or once a month? Are there specific times when this fighting takes place? Any seasons or specific days? Is it linked to school terms? Is it linked to monthly payment cycles?] Impact of domestic violence 13. Who is affected by fights in your family? How? Why these people? [Probe for the range of effects – they may include: a. Victim losing confidence, being afraid to leave the house, not being able to work and becoming trapped in a vicious cycle of dependence. b. The effects distributed across members of the household who are affected by the violence (by age, sex, biological children/foster children/other kin). c. How are they impacted by the incidences of violence?] 14. What is the immediate impact? And what is the long-term impact on those affected? [Probe for short-/long-term impacts: economic, psychological, health impacts.] Victim responses to violence 15. When you and your family have conflicts, what do you do to cope with the violence? 371 16. When a person is angry or uses violence or threatens to use violence against you, have you ever sought support from other people in the family? Intervention and mediation 17. Does anyone in the family come to intervene/help you in any situations of family violence? Who? How? Why? 18. Is there any one to help reconcile the conflicts between you and the person who has hurt you in the family? Please explain. [Probe for family/extended family intervention and mediation.] 19. Have you been helped by neighbours or local authorities/local organisations (i.e. chief or queen mother, women’s union, reconciliation group)? Please explain the reasons why you have or have not sought help from these people/organisations. [Probe: Community intervention and mediation. a. If no, why did you not seek help from these people community facilities? b. If yes, who/which did you go to for assistance? What were their responses when you told them your story or asked them for help? How did these help/support you? Were you satisfied with the help they provided? Did their support help end the violence you were facing from the family member?] Knowledge/availability of victim support services 20. Can you tell us what you know to do/have been advised to do when you are hurt by members of your family? [Probe for knowledge of rights/resources: a. Do you know about the Ghana 2007 Domestic Violence Act 732/gender equity laws? b. Do you know about any domestic violence victim support services (ask specifically about DOVVSU, CHRAJ, others)?] 21. When you were injured from the violence, did you go to the police for help? [Probe for use/availability of support services and programmes: a. If no, why did you not use the police facilities? b. If yes, how did the police facility help/support you? c. Were you satisfied with the help they provided? d. Did the police support help end the violence you were facing from the family member?] 22. When you were injured from the violence, did you go to health facilities such as hospital, commune health station, pharmacy stores? [Probe: a. If no, why did you not use these facilities? b. If yes, which of the health facilities did you go to (public or private)? c. How did the health providers help/support you? d. Did you have to pay for the services? e. If yes, did you have enough money to pay for the services? f. Did you have someone to accompany you, or did you go by yourself? If yes, who?] 372 23. What do you think local leaders can do to prevent/support victims of domestic violence (to find out about the role of chiefs, community leaders, local assembly/MP and other administrative/political officials play in implementing policy directives)? 24. What do you think the police facilities can do to support victims of domestic violence (to find out about police support and help to prevent domestic violence)? 25. What do you think the health facilities can do to support victims of domestic violence (to find out about health support for victims and help to prevent domestic violence)? 26. What do you think are the most effective ways to stop other people from being hurt like this? [Probe: a. Are these types of support available? Where do you think you can find them? b. How does/did the absence of any of these types of support influence your situation?] 27. Looking back at your situation, what advice would you give another woman or man, or child, who has just started to have these sorts of problems in their home? [Probe: What should these people do? Ask about all the options/choices they have or do not have] 28. How do you think people who hurt other people, like you have been hurt, should be punished? What should happen? [Probe: How do people think about justice, and appropriate punishment?] 29. Are you a member of a group or an organisation? Which (religious, social or traditional group, saving and credit or any group)? 30. If you could advise the government, what would you tell it to do? [This is an open- ended question that can be really useful in drawing out unique responses about government accountability to end domestic violence, so encourage respondent.] N.B. These are guidelines. If any interesting additional issues emerge, the discussion should also explore these. WRAP UP  Facilitator asks respondent if (s)he has any questions for interviewing team.  Facilitator thanks respondent for her/his time. (Thank you for sharing this with us. We have asked very difficult questions, and thank you for being so open. All shared information will remain confidential. What you have told us is very important, and will help us in our work to address domestic violence.)  From the victim’s responses, mention his/her strengths.  Give details of follow-up counselling support available both immediately and later. Give more general information about services available in the community. 373