Abstract
India presents itself as a paradox with low infant mortality and high malnutrition. This paper provides survival bias as an explanation of the paradox. Using pooled health surveys from 1993 to 2005 and a pseudo-panel selection model, this study finds that the change in Height-for-Age Z-Scores (HAZ scores) can be explained by mortality selection. Specifically, children with sample average characteristics that survive have 17.4% less HAZ scores than a child randomly drawn from the population indicating an overestimation of malnutrition in India. This is consistent with the hypothesis of weaker children surviving due to skilled delivery which pulls down the overall HAZ scores. The results are robust to controls for unobservable characteristics of groups of women. Son preference is also apparent in the results. The selection is more evident among male children and in the states where sex selection is historically seen as a problem in India.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
For example, Chad had infant mortality rate of 124 vis-a-vis 50 for India in 2009. While, Chad had 44.8% children below the age 5 as stunted while 47.9% in India for the same period (World Health Organization 2011).
HAZ Scores are the most common anthropometric measure to track malnutrition and is used by UN and WHO. I focus on height instead of weight because height is considered as a long-run measure of an individual's health (Behrman and Deolalikar 1988). But, I do also check for robustness of results with WAZ scores.
One problem that can be raised with the recall data is the measurement error problem. Since the birth histories do not go too much into the past, it is lesser of a problem in this case. Moreover, since deaths of a child are important in a mother’s life, this variable should be recorded without much measurement error.
I also check for the mortality selection effect if I constrain the results to children born under 36 months across different survey years in Table 6. The results are unchanged.
If the selection process is identical over time, then the fixed effects estimator will remove the selection bias in a panel data.
I use the 2006 WHO standards for HAZ, computed using the Stata package “haz06" which requires the data for age, height, and gender from NFHS. I have also run this using the HAZ scores reported in the NFHS. But, it does not change the results qualitatively.
Including different characteristics with missing values, increases the number of cohorts but decreases the observations within cohorts; which is not desirable computationally and otherwise for consistency (Borjas and Sueyoshi 1994).
Correlation of inverse Mills ratio with other independent variables are provided in the Appendix Table 8.
Skilled delivery assistance is measured by delivery being assisted by doctors, nurse/midwife, auxiliary midwife, ayurvedic doctor, and any other India-specific health professional. It is not considered as skilled assistance if the baby is born with the help of trained birth attendant, traditional birth attendant, relatives, other persons, or no one.
A z-score of zero indicates the median of gender and age specific reference population, − 1 is 1 standard deviation below and + 1 is 1 standard deviation above the reference median population.
Similarly, percentage change in HAZ also differs by gender and birth order, as shown in Appendix Fig. 8.
Since mother cohorts are defined by the place of residence, heterogeneity for urban and rural areas cannot be carried out at the mother cohort level. The analysis for the heterogeneity has been performed with state fixed effects.
Graph in Appendix Fig. 5.
References
Akachi, Yoko, and David Canning. 2010. Health trends in Sub-Saharan Africa: Conflicting evidence from infant mortality rates and adult heights. Economics and Human Biology 8 (2): 273–288.
Alderman, Harold, Michael Lokshin, and Sergiy Radyakin. 2011. Tall claims: Mortality selection and the height of children in India. Economics and Human Biology 9 (4): 393–406.
Behrman, J.R. 1988. Nutrition, health, birth order and seasonality: Intrahousehold allocation among children in rural India. Journal of Development Economics 28 (1): 43–62.
Behrman, Jere R., and Paul Taubman. 1986. Birth order, schooling, and earnings. Journal of Labor Economics 4 (3): S121–S145.
Belli, Pedro. 1971. The economic implications of malnutrition: The dismal science revisited. Economic Development and Cultural Change 20 (1): 1.
Boerma, J. Ties, A. Elisabeth Sommerfelt, and George T. Bicego. 1992. “Child Anthropometry in Cross-Sectional Surveys in Developing Countries: An Assessment of the Survivor Bias.” American Journal of Epidemiology 135(4):438–49.
Borjas, George J., and Glenn T. Sueyoshi. 1994. A two-stage estimator for probit models with structural group effects. Journal of Econometrics 64 (1–2): 165–182.
Bozzoli, Carlos, Angus Deaton, and Climent Quintana-domeque. 2009. Adult height and childhood disease. Demography 46 (4): 647–669.
Bushway, Shawn, Brian D. Johnson, and Lee Ann Slocum. 2007. Is the magic still there? The use of the heckman two-step correction for selection bias in criminology. Journal of Quantitative Criminology 23 (2): 151–178.
Caulfield, Laura E., Mercedes de Onis, Monika Blössner, and Robert E. Black. 2004. Undernutrition as an underlying cause of child deaths associated with diarrhea, pneumonia, malaria, and measles. The American Journal of Clinical Nutrition 80 (1): 193–198.
Dancer, Diane, Anu Rammohan, and Murray D. Smith. 2008. Infant mortality and child nutrition. Health Economics 1035 (July): 1015–1035.
Deaton, Angus. 1985. Deaton_Panel_Data_from_Time_Series_of_Cross_Sections.Pdf. Journal of Econometrics 30: 109–126.
Deaton, Angus. 2007. Height, health, and development. Proceedings of the National academy of Sciences of the United States of America 104 (33): 13232–13237.
Horton, Sue. 1988. Birth order and child nutritional status: Evidence from the Philippines. Economic Development and Cultural Change 36 (2): 341–354.
Jayachandran, Seema, and Rohini Pande. 2017. Why are Indian children so short? The role of birth order and son preference. American Economic Review 107 (9): 2600–2629.
Little, Roderick J.A. 1985. A note about models for selectivity bias. Econometrica 53 (6): 1469–1474.
Moradi, Alexander. 2010. Selective mortality or growth after childhood? What really is key to understand the puzzlingly tall adult heights in Sub-Saharan Africa. Working paper 1–23.
Panagariya, Arvind. 2013. Does India really suffer from worse child malnutrition than Sub-Saharan Africa? Economic and Political Weekly 48 (18): 98–111.
Pelletier, D.L., E.A. Frongillo, D.G. Schroeder, and J.P. Habicht. 1995. The effects of malnutrition on child mortality in developing countries. Bulletin of the World Health Organization 73 (4): 443–448.
Pinchinat, Sybil, Catherine Enel, Gilles Pison, Géraldine Duthé, Emmanuel Lagarde, François Simondon, and Kirsten B. Simondon. 2004. No improvement in weight-for-age of young children in Southern Senegal, 1969–1992, despite a Drastic Reduction in Mortality. Evidence from a Growth Monitoring Programme. International Journal of Epidemiology 33 (6): 1202–1208.
Pitt, Mark M. 1997. Estimating the determinants of child health when fertility and mortality are selective. The Journal of Human Resources 32 (1(Winter)): 129–158.
Rice, Amy L., Lisa Sacco, Adnan Hyder, and Robert E. Black. 2000. Malnutrition as an underlying cause of childhood deaths associated with infectious diseases in developing countries. Bulletin of the World Health Organization 78 (10): 1207–1221.
Rodríguez, Jhon James Mora, and Juan Muro. 2014. Consistent estimation in pseudo panels in the presence of selection bias. Economics: The Open-Access, Open-Assessment E-Journal 8 (2015): 1–26.
Savage, Tim, José G.B. Derraik, Harriet L. Miles, Fran Mouat, Paul L. Hofman, and Wayne S. Cutfield. 2013. Increasing maternal age is associated with taller stature and reduced abdominal fat in their children. PLoS ONE 8 (3): e58869.
Spears, Dean, Arabinda Ghosh, and Oliver Cumming. 2013. Open defecation and childhood stunting in India: An ecological analysis of new data from 112 districts. PLoS One 8 (9): 1–9.
Vella, Francis. 1998. Estimating models with sample selection bias: A survey. The Journal of Human Resources 33 (1): 127–169.
World Health Organization. 2011. World health statistics report, 2011.
Acknowledgements
I am grateful to Anil Deolalikar, Mindy Marks, Joseph Cummins, and Aman Ullah for their comments. This paper has benefitted from presentation at the South Asian University Seminar series, Population Association of America (PAA) Annual Meeting in 2016, and CSWEP workshop in WEAI meetings. This paper was developed in University of California, Riverside and I am grateful for the funding provided by the graduate division. All remaining errors are my own.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
See Figs. 5, 6, 7, 8 and Tables 8 and 9.
Mean HAZ scores and IMR, by survey years. These bar graphs plot the mean HAZ Scores and mean infant deaths by the 3 survey years—1993, 1998 and 2005 in our sample
Infant mortality change, by child characteristics. The first graph shows how change in infant mortality varies by birth order. The changes in infant mortality are calculated by changes in mean infant deaths by birth order over 1993–2005, divided by mean infant mortality over that cell in 1993 and multiplied by 100. The second graph plots a bar graph of infant mortality changes for males and females where male is denoted by female = 0
Infant mortality trends, by gender and birth order. This graph shows the decline in infant deaths overtime (1980–2005) by gender. Panel 2 plots the mean infant deaths by birth order for the two survey years, 1993 and 2005
Changes in HAZ (1993–2005), by gender and birth order. The first graph shows how change in HAZ varies by birth order, where change is defined as the difference between mean HAZ by birth order between 1993 and 2005, divided by HAZ in 1993 and multiplied by 100. The second graph plots a bar graph of HAZ changes for males and females where male is denoted by female = 0
Rights and permissions
About this article
Cite this article
Panda, P. Selective Mortality and Malnutrition in India. J. Quant. Econ. 18, 861–890 (2020). https://doi.org/10.1007/s40953-019-00194-8
Published:
Issue Date:
DOI: https://doi.org/10.1007/s40953-019-00194-8