Child Schooling in India: An Empirical Analysis with Cluster Based Sample

Realising the significance of education in economic development and growth, the persisting problem of low child schooling in developing countries over decades, has been an area of policy concern. In this process, India is experiencing a combination of problems, with the access to primary schooling and also with the retention and achievement at the primary, secondary and higher secondary levels. In this paper I have used the same set of individual, household and community characteristics to explain these different stages of schooling succession independently with a Heckman Sample Selection Model. Further I have also analyzed the model separately for male, female and rural samples. I have used NFHS, 1998- 99 database on India, and in an effort to account for the cluster design of this database, I have controlled for cluster fixed effects in a separate model. There is significantly strong evidence of gender difference favouring the male child at both the stages of schooling. But, in the cluster fixed effects model, though the gender effect is still very significant, yet it is favouring girls. Inside a cluster girls are estimated to acquire more years of schooling relative to boys provided that they are already enrolled. Parental schooling has a very positive and statistically significant impact on child schooling. There is also very significant wealth effect, community specific effect and regional disparities between states in India in case of child schooling