Marital Homogamy, Maternal Empowerment, and Child Health in Ghana
Faculty Mentor: Renata Forste, Sociology
Current research shows that marital homogamy, where spouses are similar in attributes, increases gender equality in marriage. Greater age and education homogamy within marriage is associated with greater gender equality between husbands and wives (Pyke & Adams). When women have equal status and are more empowered in the home, they are more able to invest in the health and development of their children. As women control earnings or home finances, funds are allocated for the health and nutrition of their children, rather than alcohol and status consumer goods (Hoddinott & Haddad, 1995). This is crucial as the nutrition and health care a child receives at an early age influences cognitive ability later in life (Berkman et. al, 2002; Pelletier et al., 1995). To further understand this relationship, I utilize the Demographic and Health Survey of Ghana, 2014 to examine how age and educational homogamy within marriage affect child nutrition and health. I hypothesize, based on previous literature, that greater parental age and educational homogamy will be associated with lower stunting or chronic child malnutrition, and that this effect will be mediated by maternal empowerment.
Data for this study come from the Demographic and Health Survey of Ghana, 2014. The survey asked women of childbearing age several questions about different aspects of health and daily life, such as sanitation, diet, health knowledge and practices, decision-making, female empowerment, and so forth. Detailed information, including height and weight data, were collected from children five years and under at the date of the survey.
Children who are 2 standard deviations or more below the median height for their age are classified as stunted according to the World Health Organization. To measure child malnutrition, I considered children who were classified as stunted vs. children who were not. To measure parental age homogamy, I subtracted the age of the wife from the age of her husband/partner. Similarly, to measure educational homogamy, I subtracted the education level of the wife from that of her husband/partner. To measure female empowerment, I utilized several indicators, including plural marriage, acceptance of wife beating, decision making in the home, and the wife’s earnings relative to her husband’s. Control variables include: gender, child age, ethnicity, religion, location, household size, wealth, maternal age at first birth, child health, health knowledge, and sanitation. Given the dependent variable (child stunting) is a dichotomous measure, my models are estimated using logistic regression. The results are reported as odds ratios and indicate the odds of the child being stunted versus not for a unit change in the independent variables. I first model the relationship between age and education homogamy and stunting. Next, I consider the mediating influence of women’s empowerment, and last I include the control variables.
19.2% of children within the sample are stunted. Husbands on average are 7.3 years older than their wives and .27 levels higher in education. On average, husbands and wives achieve primary education. Wives are 31 years old on average, and husbands are 38 years old, but there is greater variation in ages among husbands.
In relation to wife-beating, on a scale of 0 to 5, 0 being never acceptable and 5 being acceptable in all situations referenced, wives on average scored 1.05, or that it was acceptable in 1 to 2 of the 5 situations for a wife to be beaten by her husband. Decision-making ability varies, with most women having a say in visiting relatives (84.9%), health care (73.7%), and large purchases (71.5%), but less than half (43.1%) indicate any say in how the husband’s earnings are spent. Very few women earn more or equal to their husband (9.6%) and about 1 in 5 respondents indicated they participate in plural marriage or are unsure of how many wives their husband has (19.8%).
First, I examined the effects of parental age homogamy, education homogamy, mother’s age and mother’s education on child stunting. For every education level that the husband was higher than the wife, the child was 16.5% less likely to be stunted. But, for every level of education that the wife had regardless of the husband’s education level, the child’s odds of stunting was reduced by about 43%.
Next, I included measures of female empowerment. As hypothesized, the addition of these variables reduced the association between educational homogamy and maternal education and child stunting. To clarify, female empowerment is partially the reason why marital homogamy affects child stunting. Finally, I included control variables, to prove that the model is not affected by other variables.
From these results, we can see that educational homogamy is a significant predictor of child stunting, mediated by female empowerment. However, surprisingly, it is the father’s higher education relative to the mother that decreases the likelihood of stunting. This is possibly due to the effect that father’s education has on wealth, which can play a crucial role in the family’s diet and health. Preliminary logistic regression models show that this may be the case; further research is needed to establish this relationship.
However, after taking into account wealth, father’s education level was still mediated through female empowerment, in particular acceptance of wife-beating. This could be due to the effect speculated by Basu (2002): educated men are saying something about themselves and their views of women by marrying educated women. Educating men may also reduce domestic violence in general, as found in other studies (Takyi & Mann, 2006). Further research is needed to establish the relationship between husband’s education level on wealth, acceptance of wife-beating, and child health.
One limitation of this study, however, is that due to child age within the sample, diet and nutrition are not able to be adequately measured or included. Further studies should focus on including measures of diet.
Regardless, my research underscores the importance of mother’s education for child health, regardless of female empowerment, wealth, region, or any other control variable included. Governmental and non-governmental actors should value educating both men and women and consider this information when moving forward with plans and policies to empower women and to improve child nutritional status.