Stochastic Forecasts of Fertility Based on Quantum-Tempo Decomposition

Jose A. Ortega, Universidad Autonoma de Madrid

Forecasting fertility is one of the weak points of population forecasting. One of the reasons for poor performance of forecasts in the past has been the use of period TFR figures as a base for forecasts without consideration of tempo effects. Recent developments in the analysis of tempo effects provide the basis for improved forecasting methods. In this paper we propose the use of the techniques developed by Kohler and Ortega (2002) to model age and parity specific fertility intensities taking into consideration tempo effects. The use of a multivariate state-space model for modeling and forecasting the tempo adjusted quantum of fertility, the mean age at childbearing and the variance of the intensity schedule lead to stochastic fertility forecasts that capture turning points in fertility better than existing methods. The method is applied to Swedish fertility based on data for the period 1970-2000.