Domestic violence and decision-making power of married women in Myanmar: analysis of a nationally representative sample

Russell Kabir,Mainul Haque,Masoud Mohammadnezhad,Nandeeta Samad,Shabnam Mostari,Shiny Jabin,Md Anwarul Azim Majumder,Md Golam Rabbani
DOI: https://doi.org/10.5144/0256-4947.2019.395
Abstract:Background: Women in Myanmar are not considered decision makers in the community and the physical and psychological effect of violence makes them more vulnerable. There is a strong negative reaction, usually violent, to any economic activity generated by women among poorer and middle-class families in Myanmar because a woman's income is not considered necessary for basic survival. Objective: Explore the relationship between domestic violence on the decision-making power of married women in Myanmar. Design: Cross-sectional. Setting: National, both urban and rural areas of Myanmar. Patients and methods: Data from the Myanmar Demographic and Health Survey 2015-16 were used in this analysis. In that survey, married women aged between 15 to 49 years were selected for interview using a multistage cluster sampling technique. The dependent variables were domestic violence and the decision-making power of women. Independent variables were age of the respondents, educational level, place of residence, employment status, number of children younger than 5 years of age and wealth index. Main outcome measures: Domestic violence and decision-making power of women. Sample size: 7870 currently married women. Results: About 50% respondents were 35 to 49 years of age and the mean (SD) age was 35 (8.4) years. Women's place of residence and employment status had a significant impact on decision-making power whereas age group and decision-making power of women had a relationship with domestic violence. Conclusion: Giving women decision making power will be indispensable for the achievement of sustainable development goals. Government and other stakeholders should emphasize this to eliminate violence against women. Limitations: Use of secondary data analysis of cross-sectional study design and cross-sectional studies are not suitable design to assess this causality. Secondly the self-reported data on violence may be subject to recall bias. Conflict of interest: None.
What problem does this paper attempt to address?