Explaining Cross-State Disparities in Child Nutrition in Rural India
Source: Flickr, Tom Maisey

A paper published in World Development investigates the cross-state disparities in child nutrition in rural India.  Such a study is relevant as India has the largest number of stunted children in the world, with a child stunting prevalence higher than Africa south of the Sahara, despite India’s rapid economic growth over the past two decades. Evidence is accumulating that there could be severe lifelong economic, health, and cognitive repercussions arising from these early childhood height deficits. There is substantial heterogeneity across states; the National Family Health Survey 2005 data show that under five stunting prevalence ranges from 24 percent in Kerala to 57 percent in Uttar Pradesh.

Despite a growing field of research on child nutrition in India recent studies have generally focused on child nutrition in the nation as a whole but have failed to study the individual particularities, differences and causes of child nutrition in each state. Studies have hypothesized that efforts in combatting low child nutrition rates need to go beyond equalizing the commonly-observed nutrition correlates (such as education, access to the PDS and agricultural production).

The study used data from the nationally representative National Family Health Survey conducted in 2012-2013 which is complemented by UNICEF data on child nutrition in India. The study uses under five height for age (HAZ) as an indicator for child nutrition. A case study approach is used that compares five states with relatively poor child nutrition indicators to one State with relatively good nutrition indicators. The States covered that have the worst child nutrition indicators are: Bihar, Uttar Pradesh, Madhya Pradesh, Odisha and Gujarat; they account for half of India’s stunted children. Under-five stunting prevalence in these states ranges from 45 percent in Odisha to 57 percent in Uttar Pradesh. Tamil Nadu, in comparison, has a stunting prevalence of 31 percent and is considered a relatively good performer on child nutrition indicators.

The study uses a regression analysis to determine the role of different variables in influencing HAZ differences between States. The covariate effects that are analysed are: child characteristics, maternal education and marriage age, mother’s nutrition, sanitation and environment, demographics, non-agricultural assets, and agricultural assets. The paper defines coefficient effects as “returns” to specific endowments, which includes basic endowments that affect nutrition as well as the effect of nutrition policies and quality of institutions.

Regarding results, the study confirms that the average HAZ for children under five is much higher in Tamil Nadu than in each of the comparison states. The study finds that mother’s education is an important factor in HAZ outcomes, and is found to be much higher in Tamil Nadu as compared to other states. Average schooling of mothers in Tamil Nadu is about six years, compared to two to four years of education on average in the other states. Similarly, household sizes are significantly smaller in Tamil Nadu than in the other states. With the exception of Gujarat, households in Tamil Nadu are found on average to have better endowments of assets such as electrification, televisions and motorcycles, and a greater availability of clean cooking fuel. On the other hand, open defecation rates are not found to be substantially different across states. 

Although the analysis shows mixed patterns of significance across the states, regression analysis indicates that child birth weight and age, and education and nutritional status of the mother display particularly consistent and statistically significant associations with HAZ across states. The child’s weight at birth is seen to have a strong positive correlation with current HAZ. Through further analysis  (using the Oaxaca–Blinder decomposition of mean HAZ gap between Tamil Nadu and Bihar, Uttar Pradesh, Madhya Pradesh, Odisha and Gujarat) the paper finds that Tamil Nadu’s endowments of maternal nutrition, education and empowerment put together explain around 16 to 20 percent of the HAZ gap with Bihar, Madhya Pradesh, Uttar Pradesh and Odisha. The other covariate variables studied are found to have a very marginal effect on HAZ rates.

Under further analysis the study finds that coefficient effects, such as State nutrition policies and the effectiveness of State institutions in supporting nutrition programmes, explain 60 to 80 percent of the HAZ gap between Tamil Nadu and Bihar, 55 to 80 percent in the Madhya Pradesh case, over 75  percent  in the Gujarat case, and 67 to 88 percent in the Uttar Pradesh case. This is significant as it suggests that basic endowments, nutrition policies and the effectiveness of State institutions have a large impact on determining HAZ ratios and in explaining the gaps in HAZ rates between States. In the discussion the paper speculates that Tamil Nadu’s record of superior policymaking in this arena is likely to be important and explain the large effect of coefficient differences. Historically, from the early 1980s until the late 90s, the Tamil Nadu Integrated Nutrition Program (TINP) delivered a set of interventions centred on nutrition education, growth monitoring, primary health care, and food supplementation on a state-wide scale. The TINP has been widely praised and was able to reach children under five years of age much more effectively than the ICDS program that was rolled out in the rest of the country in this period. Similarly the implementation of the Public Distribution system is considered to be considerably more effective in Tamil Nadu as compared to the other States covered in this study. From these findings the paper suggests that proven nutrition-specific interventions need to be scaled up and that policy processes and their political underpinnings will be critical for nutrition in the post-2015 era.

The full paper can be accessed here.

By: Bas Paris

Photo credit:Flickr, Tom Maisey