How Does Mother's Schooling Affect Family Health, Nutrition, Medical Care Usage, and Household Sanitation

Citation: Behrman, Wolfe (1987) How Does Mother's Schooling Affect Family Health, Nutrition, Medical Care Usage, and Household Sanitation.
Internet Archive Scholar (search for fulltext): How Does Mother's Schooling Affect Family Health, Nutrition, Medical Care Usage, and Household Sanitation
Tagged: uw-madison (RSS), wisconsin (RSS), sociology (RSS), demography (RSS), prelim (RSS), qual (RSS), WisconsinDemographyPrelimAugust2009 (RSS)

Summary

Introduction

Previous studies indicate that nutrition and health of household members tends to improve with increased education among mothers in developing nations (e.g., World Bank 1980). However, research which considers the endowments of the mother (e.g., family background), however, calls this finding into question. This counter-finding suggests that education may serve as a partial proxy for unobserved family endowments (p. 186). The intent of this study is to analyze simultaneously the effects of women's schooling and background variables on child and maternal health, nutrition, medical care and household sanitation. Method: Data were taken from a household survey of women aged 15-45 in Nicaragua that was conducted in 1977-78. Multiple measures of child health, mother's health, household nutrition, medical care usage, community endowment and mother's childhood endowment were utilized to estimate latent constructs and measurement error via confirmatory factor analyses. Single observed variables on mother's schooling, age and household income were also used in the analyses. Covariance structure models were estimated via LISREL to determine the relative impact of mother's education, mother's endowment and community endowment on child health and mother's health. Also, reduced form equations are used to estimate the effects of exogenous variables (e.g., education) on intermediate, endogenous variables (e.g., sanitation).

Findings

In a model without the endowment effects, mother's education has significant effect on all the dependent variables except her own health. In addition, there are indirect effects on child health through nutrition and water and sanitation (p. 198). When a latent variable for community endowments (e.g., literacy rate) are added, it has a significant positive effect on mother's health, household nutrition, medical care usage, and water and sanitation, as well as an indirect effect on child health by means of nutrition and water and sanitation (p. 199). The inclusion of community endowment improves the explanatory power of the model, but it does not significantly affect most independent variables in either the reduced form or health production equations. That said, it does substantially weaken the effect of mother's education on sanitation and nutrition. Finally, a full model is estimated by including a latent variable for mother's endowment. A woman's childhood endowment has significantly positive coefficient estimates for a woman's subsequent health, household nutrition, medical care usage, and water and sanitation. It thus appears to be an important determinant of health and nutrition (p. 200). Moreover, mother's endowment renders nearly all other independent variables insignificant, and mother's education actually takes on the wrong sign for nutrition and sanitation.

Conclusions

We therefore conclude that . . . the apparently strong positive effects of mother's schooling on health and nutrition in standard estimates are probably substantially overstated because of the failure to control for the mother's endowment of health knowledge and habits and health status formed during her childhood. If this result holds in the context of other developing countries, then the claims of the World Bank and others concerning the strong positive impact of women's schooling on health and nutrition are overstated and misleading (p. 202)