Fertility Projection Using Hazard-Based Microsimulation Models, with an Application to Belgian Fertility, 1960-2015.

Karel Neels , University of Antwerp
Jonas Wood, University of Antwerp

Several factors have been identified as potential drivers of fertility postponement and the decline of period fertility in recent decades, with the introduction and diffusion of contraceptive technology, rising education and female labour force participation, ideational change, variation in economic context and (lack of) supportive family policies figuring prominently among them. Despite this abundance of candidate causal factors, little quantitative evidence is available on the contribution of specific determinants to aggregate change in tempo and quantum of period fertility by birth order. Using exhaustive longitudinal microdata from the 2001 Belgian census and the national population registers, this paper uses discrete-time hazard models to document the contributions of rising educational enrolment, changing patterns of union formation as well as variation in both economic conditions and family policies, to explaining aggregate change in tempo and quantum of first births as well as progression to second and higher-order births between 1960 and 2000. Subsequently, discrete-time hazard models for entry into parenthood and progression to second and higher-order births are integrated into a microsimulation framework to generate prospective out-of-sample predictions of aggregate fertility trends by birth order, and assess whether and to what extent variation in the aforementioned factors is capable of accounting for trends in Belgian order-specific fertility between 2000 and 2015.

See extended abstract

 Presented in Session 127. Fertility Timing