Understanding the Effects of Older Siblings on Institutional Delivery of Subsequent Births in India: An Application of Multivariate Multilevel Models

Priyanka Dixit, International Institute for Population Sciences (IIPS)

Over past decade indeed, there is a near-absence of studies that have identified determinants of consistent use of institutional delivery services for all births within a woman. Hence, objective of this study is to understand effects of place of delivery of older siblings on subsequent institutional delivery in India using NFHS 2005-06 data. To fulfil objective multilevel binary logistic regression model and multivariate multilevel analytical approach has been applied. Result shows that children born 2-4 years back from date of survey were less likely to deliver in an institution than recent born child. Unconstrained binomial variance parameter shows that data constructed in a multivariate multilevel model have followed binomial assumption. Finding shows that correlations between successive pregnancies were high compared to alternative pregnancies estimate. This suggests that once decision has been made for the place of delivery of one birth, behavior of mother is unlikely to change for next and, to a lesser extent, subsequent birth.