@article{Wood_31_46,
author = {Wood, Jonas and Neels, Karel and Kil, Tine},
title={{The educational gradient of childlessness and cohort parity progression in 14 low fertility countries}},
journal = {Demographic Research},
volume = {31},
number = {46},
pages = {1365--1416},
doi = {10.4054/DemRes.2014.31.46},
year = {2014},
abstract = {Background: Although the association between fertility and education is central to several theories of fertility behaviour and is frequently explored in empirical work, educational differentials in childlessness and cohort parity progression have been scarcely documented and few cross-country comparisons have been made. 

Objective: This article explores educational gradients with respect to entry into parenthood and parity progression for cohorts born between 1940 and 1961 in 14 low-fertility countries. 

Methods: Using longitudinal microdata, discrete-time event history models for repeated events are estimated for first, second, and third births including a random effect at the level of individual women (shared frailty). Subsequently, estimated hazards are used to calculate cohort parity progression ratios by level of education. 

Results: Educational gradients in fertility differ strongly between countries whereas change over time within countries is limited. In all countries childlessness is more frequent among highly educated women, suggesting that negative effects of opportunity costs outweigh positive income effects. The effect of unequal selection into motherhood across educational groups on educational gradients in higher order births through unobserved time-invariant characteristics is limited. For second births, Central and Eastern European countries (Bulgaria, Estonia, Georgia, Romania, Russia, Hungary) show negative educational gradients, whereas the educational gradient is neutral or positive in other countries (Norway, Australia, UK, Netherlands, Belgium, France, Spain, Italy). For third births results show that Central and Eastern European countries more often display negative educational gradients, whereas other European regions and Australia show negative gradients, positive gradients, and U-shaped patterns of association. The strong differences between countries suggest that context plays an important role in shaping educational gradients in childlessness and parity progression. },
URL = {https://www.demographic-research.org/volumes/vol31/46/},
eprint = {https://www.demographic-research.org/volumes/vol31/46/31-46.pdf}
}