The age at first birth and completed fertility reconsidered: findings from a sample of identical twins

In this paper we use new methods and data to reassess the relationship between the age at first birth and completed fertility. In particular we attempt to properly estimate the postponement effect, i.e., the reduction in fertility associated with a delay in childbearing, using a sample of Danish monozygotic twins born 1945--60 to control for unobserved heterogeneity. Within-MZ twin pair estimates of the postponement effect indicate that a one year delay in the first birth reduces completed fertility by about 3% for both males and females. The effect is significantly stronger for older cohorts, and it is stronger for females with a late desired entry into parenthood. Analyses that fail to control for unobservables underestimate this postponement effect between 10--25%, and they underestimate the annual decline of this effect by up to 50%. Moreover, our estimates indicate important changes across cohorts in the relevance of child-preferences and ability characteristics for the age at first birth and the pace and level of subsequent fertility. (AUTHORS)