Participation in household decision making and justification of wife beating: Evidence from the 2018 Mali Demographic and Health Survey

International Health, Volume 14, No. 1, Year 2022

Interpretation

Background: We assessed the association between women’s participation in household decision making and justification of wife beating among married women ages 15-49 y in Mali. Methods: We employed a cross-sectional study design among 7893 women of reproductive age involving a two-stage sampling technique using version 6 of the Mali Demographic and Health Survey (MDHS) data, which was conducted in 2018. Results: Approximately 37% participated in at least one household decision while 23.4% reported that they would not justify wife beating in any of the stated circumstances. Women who participated in at least one household decision had lower odds (adjusted odds ratio [AOR] 0.834 [confidence interval {CI} 0.744 to 0.935]) of justifying wife beating. With respect to the covariates, we found that women 45-49 y of age had lower odds of justifying wife beating compared with those ages 15-19 y (AOR 0.569 [CI 0.424 to 0.764]). Women with higher education (AOR 0.419 [CI 0.265 to 0.662]) and those whose husbands had secondary education (AOR 0.825 [CI 0.683 to 0.995]) had lower odds of justifying wife beating. Women who lived in urban areas were less likely to justify wife-beating (AOR 0.328 [CI 0.275 to 0.390]) compared with those who lived in rural areas. Conclusion: This study suggests that participation in household decision making is associated with a significantly lower rate of justifying wife beating in Mali. These results underscore the need for various interventions to empower women to increase women’s participation in decision making to reduce justification of domestic violence. The data supporting this study were obtained from the version 6 of the Mali Demographic and Health Survey (MDHS), which was conducted in 2018. Specifically, the women recode file was used for the study. The MDHS forms part of the Demographic and Health Surveys (DHS) Program. DHS aims at monitoring health indicators in >85 LMICs globally. The survey captures a wide range of information on sexual and domestic violence as well as maternal and child health issues. The study has a two-stage sampling design. At the first stage, 379 primary survey units (PSUs) or clusters (104 in urban and 275 in rural areas) were systematically drawn with a probability proportional to their size in households from the list of enumeration sections (ESs) established during the general census of population and housing conducted in 2009. A household mapping and enumeration operation in the clusters was organized to draw an updated list of households in each ES to be used as a basis for stage sampling. In the regions of Kidal, Gao and Timbuktu, the mapping and enumeration of households was carried out just a few days before the data collection for the main survey. In the rest of the regions, this operation was carried out well before the main survey, from 25 May to 8 July 2018. After this, they compiled an updated list of households of each ES, a sample of 35 households in the Kidal, Gao and Timbuktu regions and 26 households in all the other regions with a systematic draw with equal probability. In households selected for the survey, all women 15–49 y of age usually living in selected households or present the night before the survey were eligible to be surveyed. For the purpose of this study, we dropped observations with missing information for the variables included in the analysis, which left data for 7893 currently married women as our analytical sample. Justification of wife beating was the dependent variable for our study. It was derived from five questions. Specifically, female survey respondents were asked if they would justify domestic violence under these five circumstances: going out without telling her husband, neglecting the children, arguing with her husband, refusing to have sexual intercourse and burning the food. For each of these circumstances, responses were ‘yes’, ‘no’ and ‘don’t know’. These were coded as no=0, yes=1 and don’t know=8. For the purpose of the analysis, only women who provided confirmatory responses (either yes or no) were included in the study. Following the methodology employed by Alam et al.,7 if a respondent thought beating would be justified, she was assigned a score of 0, but if a respondent thought beating would not be justified, she was assigned a score of 1. The internal consistency among the five variables (i.e. five circumstances) was assessed with Cronbach’s α and a value of 0.8166 was obtained. All five circumstances were used to generate the binary outcome variable: 1 if the respondent thought beatings were justified in any circumstance and 0 if the respondent thought beatings were not justified in any circumstance. The main explanatory variable of the study is self-reported participation in household decision making. This was derived from the responses to three individual questions regarding who within the household makes decisions in three circumstances: own healthcare, major household purchases and visits to family or relatives. For each circumstance, the response categories were as follows: (a) respondent alone; (b) respondent and husband, partner jointly; (c) husband/partner alone; (d) someone else and (e) other. The category (e) was deleted since there were few responses to that category (0.003%). These variables were dichotomously coded to be full or partial participation, described in options (a) and (b) and assigned a score of 1, and no participation, described in options (c) and (d) and assigned a score of 0. The internal consistency among the three variables (i.e. three circumstances) was Cronbach’s α=0.7479. The predictor variable is equal to 1 if the respondent participated in any of the decisions and 0 if the women did not participate in any of the decisions. We included a number of control variables due to their association with either the outcome or predictor variables.7,27–29 These included current age, respondent’s and husband’s education, respondent’s work status, respondent’s religion, parity, place of residence, wealth status and exposure to mass media (radio, television and newspaper). In the DHS, wealth is a composite measure computed by combining data on a household’s ownership of carefully identified assets including a television and bicycle, materials used for house construction, sanitation facilities and type of water access. Principal component analysis was used to transform these variables into a wealth index by placing individual households on a continuous measure of relative wealth. The DHS segregates households into five wealth quintiles: poorest, poorer, middle, richer and richest. Some of these variables were recoded for easy interpretation and analysis. Religion was recoded as Christian, Islam and other. Parity was recoded (0, 1, 2, 3 and ≥4) and occupation was recoded as working and not working. The data were analysed using Stata version 14.2 for MacOS (StataCorp, College Station, TX, USA). Our analysis began with a descriptive investigation into the key sociodemographic characteristics and their relationship to justification of domestic violence. We then conducted a χ2 test to ascertain the relationship between participation in household decision making, sociodemographic characteristics and justification of sexual violence. Afterwards we conducted a χ2 test to ascertain the relationship between participation in household decision making, sociodemographic characteristics and justification of wife beating. This was done to identify significant variables to be considered for the inferential analysis. All these are reported in Table 1. At the inferential level, two binary logistic regression models were fitted. The first one (model I) accounted for only women’s participation in household decision making and justification of wife beating, while the second (model II) controlled for the effect of the significant sociodemographic variables. Results for model I were presented as crude odds ratio (CORs) while adjusted odds ratios (AORs) were reported for model II with their respective confidence intervals (CIs) at a 5% margin of error. All analyses were performed considering the probability sample design. The svy commands were used in descriptive and bivariate analyses and probability weight, proposed by Rabe-Hesketh and Skrondal,30 was applied to the binary logistic regression analysis. Characteristics of study participants and percentage of females who do not justify wife beating, by sociodemographic characteristics Source: 2018 MDHS. The survey reported that ethical approval was granted by the Institutional Review Board of ICF International.31 Informed consent was sought from all the participants during the data collection exercise. We further obtained permission from the DHS Program for use of these data for the study.

AI Digest

Study Justification:

– The study aims to assess the association between women’s participation in household decision making and the justification of wife beating among married women in Mali.
– This topic is important because it addresses the issue of domestic violence and the role of women’s empowerment in reducing such violence.
– Understanding the factors that contribute to the justification of wife beating can inform interventions and policies aimed at reducing domestic violence.

Study Highlights:

– The study found that approximately 37% of women participated in at least one household decision, while 23.4% reported that they would not justify wife beating in any circumstances.
– Women who participated in household decision making had lower odds of justifying wife beating.
– Other factors associated with lower odds of justifying wife beating included older age, higher education for women, secondary education for husbands, and living in urban areas.

Recommendations for Lay Readers:

– Encourage women’s participation in household decision making to reduce the justification of wife beating.
– Promote education for women and their husbands to empower them and reduce the acceptance of domestic violence.
– Advocate for policies and interventions that address gender inequality and promote women’s empowerment.
– Raise awareness about the negative consequences of domestic violence and the importance of gender equality.

Recommendations for Policy Makers:

– Develop and implement programs that promote women’s participation in household decision making.
– Invest in education and vocational training for women to enhance their empowerment and reduce the acceptance of domestic violence.
– Strengthen laws and policies that protect women from domestic violence and ensure their access to justice.
– Allocate resources for awareness campaigns and community-based interventions to address gender inequality and promote women’s empowerment.

Key Role Players:

– Government agencies responsible for women’s empowerment and gender equality.
– Non-governmental organizations working on women’s rights and domestic violence prevention.
– Community leaders and religious leaders who can influence social norms and attitudes towards domestic violence.
– Health professionals and social workers who provide support and counseling to survivors of domestic violence.

Cost Items for Planning Recommendations:

– Funding for education and vocational training programs for women.
– Resources for awareness campaigns and community-based interventions.
– Budget for implementing and enforcing laws and policies related to domestic violence.
– Allocation of funds for support services for survivors of domestic violence, including counseling and shelters.
– Investment in data collection and research to monitor the effectiveness of interventions and policies.

Strength of Evidence

The strength of evidence for this abstract is 7 out of 10.
The evidence in the abstract is relatively strong, but there are some areas for improvement. The study design is cross-sectional, which limits the ability to establish causality. Additionally, the abstract does not provide information on the response rate or potential biases in the sample. To improve the evidence, future studies could consider using a longitudinal design to establish temporal relationships and address potential biases. It would also be beneficial to provide more details on the sampling technique and the representativeness of the sample. Finally, including information on the statistical significance of the findings and effect sizes would enhance the strength of the evidence.

Abstract

The data supporting this study were obtained from the version 6 of the Mali Demographic and Health Survey (MDHS), which was conducted in 2018. Specifically, the women recode file was used for the study. The MDHS forms part of the Demographic and Health Surveys (DHS) Program. DHS aims at monitoring health indicators in >85 LMICs globally. The survey captures a wide range of information on sexual and domestic violence as well as maternal and child health issues. The study has a two-stage sampling design. At the first stage, 379 primary survey units (PSUs) or clusters (104 in urban and 275 in rural areas) were systematically drawn with a probability proportional to their size in households from the list of enumeration sections (ESs) established during the general census of population and housing conducted in 2009. A household mapping and enumeration operation in the clusters was organized to draw an updated list of households in each ES to be used as a basis for stage sampling. In the regions of Kidal, Gao and Timbuktu, the mapping and enumeration of households was carried out just a few days before the data collection for the main survey. In the rest of the regions, this operation was carried out well before the main survey, from 25 May to 8 July 2018. After this, they compiled an updated list of households of each ES, a sample of 35 households in the Kidal, Gao and Timbuktu regions and 26 households in all the other regions with a systematic draw with equal probability. In households selected for the survey, all women 15–49 y of age usually living in selected households or present the night before the survey were eligible to be surveyed. For the purpose of this study, we dropped observations with missing information for the variables included in the analysis, which left data for 7893 currently married women as our analytical sample. Justification of wife beating was the dependent variable for our study. It was derived from five questions. Specifically, female survey respondents were asked if they would justify domestic violence under these five circumstances: going out without telling her husband, neglecting the children, arguing with her husband, refusing to have sexual intercourse and burning the food. For each of these circumstances, responses were ‘yes’, ‘no’ and ‘don’t know’. These were coded as no=0, yes=1 and don’t know=8. For the purpose of the analysis, only women who provided confirmatory responses (either yes or no) were included in the study. Following the methodology employed by Alam et al.,7 if a respondent thought beating would be justified, she was assigned a score of 0, but if a respondent thought beating would not be justified, she was assigned a score of 1. The internal consistency among the five variables (i.e. five circumstances) was assessed with Cronbach’s α and a value of 0.8166 was obtained. All five circumstances were used to generate the binary outcome variable: 1 if the respondent thought beatings were justified in any circumstance and 0 if the respondent thought beatings were not justified in any circumstance. The main explanatory variable of the study is self-reported participation in household decision making. This was derived from the responses to three individual questions regarding who within the household makes decisions in three circumstances: own healthcare, major household purchases and visits to family or relatives. For each circumstance, the response categories were as follows: (a) respondent alone; (b) respondent and husband, partner jointly; (c) husband/partner alone; (d) someone else and (e) other. The category (e) was deleted since there were few responses to that category (0.003%). These variables were dichotomously coded to be full or partial participation, described in options (a) and (b) and assigned a score of 1, and no participation, described in options (c) and (d) and assigned a score of 0. The internal consistency among the three variables (i.e. three circumstances) was Cronbach’s α=0.7479. The predictor variable is equal to 1 if the respondent participated in any of the decisions and 0 if the women did not participate in any of the decisions. We included a number of control variables due to their association with either the outcome or predictor variables.7,27–29 These included current age, respondent’s and husband’s education, respondent’s work status, respondent’s religion, parity, place of residence, wealth status and exposure to mass media (radio, television and newspaper). In the DHS, wealth is a composite measure computed by combining data on a household’s ownership of carefully identified assets including a television and bicycle, materials used for house construction, sanitation facilities and type of water access. Principal component analysis was used to transform these variables into a wealth index by placing individual households on a continuous measure of relative wealth. The DHS segregates households into five wealth quintiles: poorest, poorer, middle, richer and richest. Some of these variables were recoded for easy interpretation and analysis. Religion was recoded as Christian, Islam and other. Parity was recoded (0, 1, 2, 3 and ≥4) and occupation was recoded as working and not working. The data were analysed using Stata version 14.2 for MacOS (StataCorp, College Station, TX, USA). Our analysis began with a descriptive investigation into the key sociodemographic characteristics and their relationship to justification of domestic violence. We then conducted a χ2 test to ascertain the relationship between participation in household decision making, sociodemographic characteristics and justification of sexual violence. Afterwards we conducted a χ2 test to ascertain the relationship between participation in household decision making, sociodemographic characteristics and justification of wife beating. This was done to identify significant variables to be considered for the inferential analysis. All these are reported in Table 1. At the inferential level, two binary logistic regression models were fitted. The first one (model I) accounted for only women’s participation in household decision making and justification of wife beating, while the second (model II) controlled for the effect of the significant sociodemographic variables. Results for model I were presented as crude odds ratio (CORs) while adjusted odds ratios (AORs) were reported for model II with their respective confidence intervals (CIs) at a 5% margin of error. All analyses were performed considering the probability sample design. The svy commands were used in descriptive and bivariate analyses and probability weight, proposed by Rabe-Hesketh and Skrondal,30 was applied to the binary logistic regression analysis. Characteristics of study participants and percentage of females who do not justify wife beating, by sociodemographic characteristics Source: 2018 MDHS. The survey reported that ethical approval was granted by the Institutional Review Board of ICF International.31 Informed consent was sought from all the participants during the data collection exercise. We further obtained permission from the DHS Program for use of these data for the study.

Materials

N/A

Innovations

Based on the provided information, it appears that the study focused on assessing the association between women’s participation in household decision making and justification of wife beating among married women in Mali. The study utilized data from the 2018 Mali Demographic and Health Survey (MDHS) and employed a cross-sectional study design.

In terms of potential innovations to improve access to maternal health based on this study, here are some recommendations:

1. Promote women’s empowerment: Empowering women through education, skills training, and economic opportunities can enhance their decision-making power within households and reduce the justification of domestic violence. This can be achieved through programs that provide education and vocational training for women, as well as initiatives that promote women’s economic independence.

2. Increase awareness and education: Conducting awareness campaigns and educational programs on women’s rights, gender equality, and the harmful effects of domestic violence can help change societal attitudes and reduce the justification of wife beating. These campaigns can target both men and women, with a focus on promoting respectful and non-violent relationships.

3. Strengthen healthcare services: Improving access to quality maternal healthcare services is crucial for women’s well-being and can contribute to reducing domestic violence. This can be achieved by increasing the availability and affordability of maternal healthcare services, ensuring skilled healthcare providers are accessible, and promoting community-based initiatives that address maternal health needs.

4. Enhance data collection and monitoring: Continuously collecting and monitoring data on maternal health indicators, including domestic violence, can help identify trends, gaps, and areas for improvement. This can inform evidence-based decision-making and the development of targeted interventions to address maternal health issues.

5. Foster partnerships and collaboration: Collaboration between government agencies, non-governmental organizations, healthcare providers, and community-based organizations is essential to address the complex issue of domestic violence and improve access to maternal health. By working together, stakeholders can leverage their expertise, resources, and networks to develop comprehensive and sustainable solutions.

It is important to note that these recommendations are based on the information provided and may need to be further tailored and contextualized to the specific needs and challenges faced in Mali.

AI Innovations Description

The recommendation that can be developed into an innovation to improve access to maternal health based on the study is to implement interventions that empower women and increase their participation in household decision making. This can be achieved through various strategies such as:

1. Promoting women’s education: Providing opportunities for women to access education and acquire knowledge and skills can empower them to actively participate in household decision making, including decisions related to maternal health.

2. Engaging men as allies: Encouraging men to support and involve women in decision making can help challenge traditional gender norms and promote gender equality within households. This can be done through awareness campaigns, workshops, and community dialogues.

3. Strengthening women’s economic empowerment: Enhancing women’s economic status and financial independence can increase their decision-making power within households. This can be achieved through initiatives such as providing microfinance opportunities, vocational training, and promoting women’s entrepreneurship.

4. Improving access to information: Ensuring that women have access to accurate and comprehensive information about maternal health, including their rights and available services, can empower them to make informed decisions and advocate for their own health needs.

5. Enhancing healthcare services: Implementing interventions that improve the quality, accessibility, and affordability of maternal healthcare services can also contribute to empowering women and increasing their participation in decision making. This can include initiatives such as expanding healthcare facilities, training healthcare providers, and reducing financial barriers to accessing healthcare.

By implementing these recommendations, it is expected that women’s participation in household decision making will increase, leading to improved access to maternal health services and a reduction in the justification of domestic violence.

AI Innovations Methodology

Based on the provided description, the study focuses on the association between women’s participation in household decision making and the justification of wife beating among married women in Mali. The study utilized data from the 2018 Mali Demographic and Health Survey (MDHS) and employed a cross-sectional study design.

To improve access to maternal health, the following innovations could be considered:

1. Mobile Health (mHealth) Applications: Develop and implement mobile applications that provide pregnant women with access to information and resources related to maternal health. These applications can provide guidance on prenatal care, nutrition, and postnatal care, as well as reminders for appointments and medication.

2. Telemedicine Services: Establish telemedicine services that allow pregnant women in remote areas to consult with healthcare professionals through video calls or phone calls. This can help overcome geographical barriers and provide timely medical advice and support.

3. Community Health Workers: Train and deploy community health workers who can provide maternal health education, conduct regular check-ups, and refer women to healthcare facilities when necessary. These workers can play a crucial role in reaching women in underserved areas and improving access to maternal health services.

4. Maternal Health Vouchers: Implement voucher programs that provide pregnant women with financial assistance to access maternal health services. These vouchers can cover costs such as antenatal care visits, delivery services, and postnatal care, ensuring that women can afford essential healthcare services.

To simulate the impact of these recommendations on improving access to maternal health, a methodology could include the following steps:

1. Define the indicators: Identify key indicators that measure access to maternal health, such as the number of antenatal care visits, institutional delivery rates, postnatal care utilization, and maternal mortality rates.

2. Baseline data collection: Gather data on the current status of these indicators in the target population. This can be done through surveys, interviews, or analysis of existing data sources.

3. Intervention implementation: Implement the recommended innovations in a selected sample population or geographical area. Ensure that the interventions are properly implemented and reach the intended beneficiaries.

4. Data collection post-intervention: Collect data on the selected indicators after the implementation of the interventions. This can be done through follow-up surveys, interviews, or analysis of existing data sources.

5. Data analysis: Analyze the collected data to assess the impact of the interventions on the selected indicators. Compare the post-intervention data with the baseline data to determine any changes or improvements.

6. Interpretation and reporting: Interpret the findings of the data analysis and report the results. Highlight the impact of the interventions on improving access to maternal health and provide recommendations for further actions or improvements.

By following this methodology, it is possible to simulate the impact of the recommended innovations on improving access to maternal health and assess their effectiveness in addressing the identified challenges.

Author

Dima Health

Statistics:

Citations: 3

Identifiers:

DOI: 10.1093/inthealth/ihab008

ISSN: 18763413

Research Areas:

Community Interventions, Environmental, Food Security, Health System and Policy, Maternal Access, Maternal and Child Health, Quality of Care, Sexual and Reproductive Health, Social Determinants, Violence and Injury

Study Countries:

Mali

Study Design:

Case-Control Study, Cross Sectional Study, Grounded Theory

Study Approach:

Quantitative

Participants Gender:

Female