Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/196787 
Year of Publication: 
2019
Series/Report no.: 
IZA Discussion Papers No. 12289
Publisher: 
Institute of Labor Economics (IZA), Bonn
Abstract: 
The poor state of child health in India has generated a number of puzzles that have received attention in the literature. A recent focus on birth order has produced contradictory results. Coffey and Spears (2019) document an early-life survival advantage in India accruing to later birth orders, which they interpret as the result of a pattern of improving maternal nutrition over mothers' childbearing careers. In apparent contrast, Jayachandran and Pande (2017) show, using the same set of demographic surveys, a disadvantage in child height for later birth orders in India relative to Africa's birth order gradient. They interpret this pattern as discrimination against later birth-order children in India. This paper resolves the apparent contradiction, showing how differing correlations between sibsize (a child's number of siblings) and household wellbeing can account for the empirical findings of both studies: A mother having higher fertility, rather than lower, implies more socioeconomic disadvantage within India than within Africa. Accounting for sibsize reverses the apparent Indian laterborn disadvantage in child height, reversing the interpretation of Jayachandran and Pande. In short, a child's sibsize (or, equivalently, its mother's fertility) is an omitted variable in Jayachandran and Pande's analysis of birth order effects. Resolving these puzzles is critical for human development policy to combat the enduring challenges of disproportionately high rates of stunting and neonatal death in India, where one-fifth of global births occur.
Subjects: 
child height
birth order
India
DHS
high-dimensional fixed effects
JEL: 
O15
I15
Document Type: 
Working Paper

Files in This Item:
File
Size
827.14 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.