An expert on excess mortality in Germany

Herr zur Nieden, the Federal Statistical Office recently presented a special evaluation on excess mortality in November. Why are statisticians so excited about these numbers?

Stefan Tomik

With these data it is possible to assess the development of deaths regardless of the exact cause of death. Although we have seasonal fluctuations in the total number of deaths, there are indications of an expected course that results from the average value of the previous years from 2016 to 2019. And if more people die in 2020 than expected, we will have a relatively clear finding. According to preliminary figures available so far, excess mortality was 11 percent in November; it rose from week to week to 14 percent in the last week of November. That is an unusually high value for this time of year. The results available so far are purely a case count without checking and completeness checks.

Does it even make sense to lump all deaths together? Wouldn’t you at least have to factor out traffic accidents and murders if you wanted to know, for example, how many people died of Corona?

We cannot do that because we only get a message from the registry offices that someone has died, without any information on the cause of death. Excluding individual causes such as traffic accidents would also be difficult because they could be indirect effects of the lockdown and thus the pandemic. You would get a lot of definition problems – what do we include and what not? But the decline in road accidents in spring hardly plays a role in the overall picture.

Felix zur Nieden, specialist in death rates at the Federal Statistical Office

Felix zur Nieden, specialist in death rates at the Federal Statistical Office

Image: Alexa Sommer /

What can we learn from your numbers that we don’t already know from the Covid statistics of the Robert Koch Institute (RKI)?

Our numbers can show whether there is an undiscovered dark figure, i.e. whether all corona deaths have really been recorded. Since deaths include all deaths, they can also reflect indirect effects of the pandemic.

Is there such a number of unreported cases?

In some other countries, excess mortality was much higher than the officially reported Covid deaths. It was concluded from this that too little was tested there. This did not play a major role in Germany, as the findings on excess mortality and the corona death rates already matched quite well in the spring.

What are the limits of the interpretation of excess mortality?

We do not see from the death rates what would have happened if politicians had taken no or fewer protective measures against the corona virus. We only see what actually happened despite all the measures that were aimed at preventing additional deaths. The excess mortality gives an indication of how well this worked, also in an international comparison. According to the figures available so far, Germany is doing comparatively well here. However, you cannot assess the fundamental danger of the coronavirus; you have to use other data, for example from medical studies.

Critics of the corona measures claim that the protective measures led to more deaths than the virus itself. Operations had been postponed, patients no longer dared to see a doctor, and some died of loneliness. Is that valid?

Leave a Comment