Navigation and service

Reiner H. Dinkel and Martin Kohls (2006)

"Normal" Seasonal Variations and the Effect of Temporary Peaks in Mortality in Germany*

In: Zeitschrift für Bevölkerungswissenschaft, Vol. 31, 2/2006, p. 163-186, Wiesbaden: VS Verlag für Sozialwissenschaften, ISSN: 0340-2398

Saisonal variations are observed for all demographic events. Already the classical writers of demography identified seasonal variations for all demographic events, including mortality. Extreme fluctuations – occurring particularly in the winter months – in the number of deaths caused by epidemics, still frequent during the 19th century, with a few exceptions disappeared in the 20th century. After the Second World War, extreme mortality values became an exception.

In the summer of 2003, tens of thousands of additional deaths were attributed to the consequences of the summer heat wave in Germany, as in many other countries. To clarify the death burden of such an extraordinary event we have to differentiate between observed and expected mortality. Expected mortality is the number of deaths for a month, taking into account its seasonal component. Thus, official mortality data from 1948 to 2004 were used to calculate the seasonal pattern of mortality in Germany using the Hodrick-Prescott filter.

Even if the number of deaths exceeds the 99 percent confidence interval of its seasonal value, this will not necessarily lead to excess mortality. If the number of deaths falls below its expected value in the following months, this may constitute no more than a short-term variation in the date of deaths, the annual figures remaining unchanged ("harvesting" effect). Careful research of the monthly number of deaths in late 2003 and early 2004 reveals that most of what happened in Germany during the 2003 heat wave was no more than a short-term alteration in the date of death.

* peer-reviewed article (original German title: Die „normale“ Saisonalität und die Auswirkung kurzzeitiger Extremwerte der Mortalität in Deutschland)

This Page

© Federal Institute for Population Research - 2017