In order to understand what is driving the change in

This preview shows page 130 - 132 out of 177 pages.

In order to understand what is driving the change in the anthropometric outcomes, we perform the decomposition analysis. We decompose the change into its constituent components at the mean, using the OB method, and across the distribution using QR based MM method. We find for the HAZ measure, the results of OB and MM are in consonance with each other with coefficient effect contributing relatively more to the overall change than the covariate effect. The disaggregate OB of HAZ, however, reveals that bulk of the coefficient effect is coming from the intercept. The positive intercept term (picks up the effect of all other covariates not included in the model), while more difficult to interpret, suggests that the improvement in HAZ is likely to pick up the effect of policies and interventions related to food, nutrition and sanitation that were put in place years ago. Another important excluded covariate is mother’s health which is related to child’s height through the genetics pathway. We can argue that our sample of interviewed mothers in NFHS-3 are relatively healthier compared to those interviewed in NFHS-1. The NFHS-3 parents were by and large a part of post Green Revolution period and were less likely to be food insecure, whereas NFHS-1 parents were young children
The Delhi University Journal of the Humanities and the Social Sciences Vol. 4, 2017 126 before the Green Revolution, and may have been both nutritionally deprived, and also, may have faced severe droughts. It is possible that this affected their heights more permanently than weights, and part of this is showing up in the positive intercept contribution in HAZ decomposition, but not in WAZ decomposition. Thus, the cohort of children born to NFHS-3 mothers are relatively taller which gets reflected in the positive and dominant contribution of the intercept to the coefficient effect. Turning to the decomposition of the WAZ indicator, we find that it is the improvement in the covariates, mostly related to improved endowment of mother’s education, number of vaccines received, improved toilets and household wealth that explains the improvement in the mean WAZ. The results are, however, enriched by MM decomposition which suggests that the coefficient effect is equally important for the bottom half of the distribution, that is, for the relatively undernourished. An important finding of this analysis is that improvements have been higher for the most vulnerable girls and for them both the covariate and coefficient effects are important. We are also able to highlight the factors associated with poor nutrition as well as improved outcomes which can be targeted through specific interventions and policy initiatives: health and hygiene infrastructure; mother’s education; and household’s economic status. Therefore, efforts should be directed in designing and implementing policies targeting these factors such that both the covariate and coefficient effects work complementarily to each other in lowering and ultimately eradicating the problem of girl-child undernutrition in India.

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture