Mothers’ Employment and their Children’s Schooling: A Joint Multilevel Analysis for India

Abstract

This paper studies the relationship between mothers' employment and children's schooling in India. Using the second National Family Health Survey, the results of a multilevel probit model show that the correlation between mothers' employment and their children's schooling is negative. Women in poorer households are more likely to work but, given the negative correlation, their additional income does not seem sufficient to enable children's school attendance. A sensitivity analysis on wealth deciles shows that this negative relationship disappears in urban areas and becomes weaker in rural areas at the top wealth deciles. JEL Classification: J13; J22; O15; O18.

Household Enterprise in Base-of-the-Pyramid Markets: the Influence of Institutions and Family Embeddedness

“…The effect of family embeddedness also plays out via the mother's employment, with research suggesting both substitution and complementarity effects (DeGraff & Levison, 2009;Francavilla, Giannelli, & Grilli, 2013;Self, 2011). More specifically, as the mother works in the household enterprise or undertakes income diversification activities in the broader labor market, children often (1) have to act as substitutes by performing key household duties (i.e., cleaning, cooking, etc.…”

“…In the very low-income context, children's well-being does not benefit from income diversification. Indeed, children instead become a tool used by households to diversify income and supplement older family members who work in the household enterprise (DeGraff & Levison, 2009;Francavilla et al, 2013). In such household enterprises characterized by family over-embeddedness, children's well-being suffers, increasing the potential that the family will remain in its cycle of poverty in future generations.…”

Does Community-Based Rehabilitation Enhance the Multidimensional Well-Being of Deprived Persons With Disabilities? A Multilevel Impact Evaluation

“…A random effects model explicitly accounts for the multilevel structure of the data, thus the inferential results are adjusted for the within-cluster correlation (Rabe-Hesketh & Skrondal, 2012;Snijders & Bosker, 2012). For example, a random effects logit model has been used by Francavilla, Giannelli, and Grilli (2013) to analyze the probability of attending school in India, where clusters are sets of children with the same mother. In the present application, clusters are villages, which are indexed by j = 1,. .…”

Manure and adoption of modern seeds in cereal‐based systems in West African drylands: linkages and (non)complementarities

“…However, joint determination does not necessarily mean the two technologies are correlated. As noted by Francavilla et al (2012), the SUR procedure has the advantage of enabling the assessment of whether the regressors have opposing or similar effects on the two adoption decisions, but is not capable of assessing the effects of one technology on the other. Neither does it make it possible to test and/or to address the issue of endogeneity.…”