Smoothing trends and seasonality in short-term mortality forecasting

Ainhoa Elena Leger , Interdisciplinary Centre on Population Dynamics, University of Southern Denmark
Silvia Rizzi, Interdisciplinary Centre on Population Dynamics, University of Southern Denmark
Ugofilippo Basellini, Max Planck Institute for demographic Research

Excess mortality is a useful measure to quantify the death toll of various health shocks, from the impact of seasonal influenza to the effects of heatwaves, and pandemics of infectious diseases. Methodological choices about the model to forecast baseline mortality in the absence of a shock can lead to varying estimates of excess death. Starting from modulation models for seasonal death counts developed by Eilers et al. 2008 and combining them with the P-splines forecasting approach developed by Currie et al. 2004, a forecasting strategy is proposed for baseline seasonal patterns. Different specifications of the model are compared. An illustration is shown for Sweden using data from Statistics Sweden from 2008 to February 2020 to forecast mortality from March 2020 to 2021 by age groups. Preliminary results show the forecasts obtained modelling smooth trends and smooth seasonality components.

See extended abstract

 Presented in Session 96. Modelling and Forecasting Mortality