Postponement and recuperation of cohort fertility

Karel Neels, Vrije Universiteit Brussel

Period circumstances are widely considered to present the prime source of variation in fertility rates. Evidence of cohorts effects being limited, cohort completed fertility is considered as a moving average of the period circumstances that cohorts experience throughout their lives. The assumed absence of cohort effects allows to adjust period measures for period-induced variation in the timing of fertility. Using data from the 2001 Belgian census this paper reconstructs first birth hazards of women aged 15 to 45 between 1960 and 2000. The results reveal age*period interaction with decreasing birth hazards among young women aged a being compensated for x years later by increasing birth hazards among women aged a+x, similar to the ‘postponement-recuperation’ mechanism suggested by Hajnal. This paper introduces specificity for education into birth-order specific and tempo-adjusted period measures to allow for lagged period effects such as the rapid increase of educational attainment that have affected cohort tempo.