摘要：This paper examines the predictors of immunization coverages of children across Indian states and evaluates the role
of Universal Immunization Program (UIP) - a comprehensive policy of the Indian government in that light. Employing
Machine Learning methods such as LASSO and hierarchical LASSO, we find that not the UIP expenditure by itself,
but health infrastructure turns out to be a robust predictor of immunization coverage. The policy prescription that
follows from our study is that the immunization program should focus on promoting the required health infrastructure
in addition to monitoring the usage of funds closely for facilitating effective usage of the money. We also scrutinize
performances of ‘BIMARU', states that are considered traditionally underperforming states in terms of health and
education.
