|
Karel Neels , University of Antwerp
Leen Marynissen, University of Antwerp
Jonas Wood, University of Antwerp
The period TFR is the most widely used aggregate-level indicator of period fertility, which feeds directly into commonly used aggregate-level models of population dynamics (e.g., cohort-component projection). Although a myriad of factors have been identified that affect the tempo and quantum of fertility, the aggregate-level nature of the period TFR and the fact that it solely considers age as the relevant time clock make it ill-suited to quantify the impact of these factors on trends in period fertility, let alone the impact of anticipated changes in these factors on future trends in period fertility. The widespread use of hazard models to link tempo and quantum of various life course transitions to potential determinants at the individual, household and contextual levels suggests that individual-based models may in fact be better suited to model aggregate-level fertility trends. Using population-wide longitudinal microdata for Belgium and a late entry design for the period 2001-2010, this paper develops a compartmental hazard model of entry into parenthood and subsequent parity progression which feeds into a dynamic microsimulation model that generates maternity histories for individual women aged 15-50 between 2011-2070. By aggregating individual trajectories into aggregate-level fertility indicators, the event logs from microsimulation models incorporating different determinants of fertility are used to quantify the impact of anticipated changes in these determinants on aggregate-level fertility measures such as the period TFR. In line with previous research, we consider variation in the period TFR associated with economic prospects, ongoing educational expansion and unfolding population heterogeneity by migration background.
Presented in Session 115. Novel Applications of Traditional Tools in Fertility Studies