This paper empirically investigates the determinants of fertility drawing upon large household data sets in India constructed by the National Sample Surveys (NSS) and National Family Health Surveys (NFHS) over the period 1992-2006. Broadly similar and consistent results are found for the two different surveys over the different years. First, we have found a negative and significant association between the number of children and mother’s education even if the latter is instrumented by (the proxy for) the pre-generation access to primary school at village level, or if parental wage equations are incorporated into the fertility equation. Both direct and indirect effects are observed for mother’s education which not just directly reduces fertility but also increases mother’s potential wages or opportunity costs which would deter her from having a baby. Second, father’s education became increasingly important in reducing fertility in the last two rounds. Third, the negative effect of expenditure on fertility is found when it is treated as exogenous, but not once instrumented. Fourth, pseudo panel models for three rounds of NSS and NFHS are estimated to confirm the negative effects of parental education. Finally, state-wise regression results show that fertility determinants are different in different states. Our results suggest that national and state governments should improve social
infrastructure, such as school at various levels, promote both male and female education, and facilitate female labor market participations to speed down the population growth. These policies would be particularly important in backward states or for socially disadvantaged groups (e.g. Scheduled Castes) which still have higher fertility as well as poverty rates.