A few months ago we posted on Ariel Karlinsky and Dmitry Kobak’s mortality dataset. Karlinsky has an update:
Our paper and dataset was finally published on eLife. Many more countries since the last version, more up to date data, some discussion and decomposition of excess mortality to various factors, etc.
Thank you for the hard work of collecting all these data and laying it out.
I have a question.
To get the correct view of the matter shouldn’t we compare excess mortality of 2020-2019 to previous years?
In trading for example to get a correct picture one has to zoom out and analyze previous years. In this example how can we get the correct picture without seeing if the war/or the excess mortality between the years 2011-2010, 2012-2011, 2013-2012…2020-2019?
I hope my question is clear.
The excess mortality in the year 2020 to 2019 has ro be compared to previous years (2019-2018, 2018-2019…2011-2010)
I would appreciate your answer,
Tal
When I go to the paper link, the first plot shows “baseline mortality” in black, and “2015-2019 average” in grey, and then excess is relative to this average. So yes, they’re comparing to more than just 2019 vs 2020, it’s an average of many previous years vs 2020.
Excess mortality is mortality above expected mortality. Expected mortality is mortality based on historical data.
For our purposes, we deploy a fairly simple model that accounts both for seasonality and trend, that in many countries is significant due to (mostly) age-structure changes.
This exercise can of course be applied to any year and any country (provided data exists). I conduct this exercise for 2019 here:
https://twitter.com/ArielKarlinsky/status/1413396674966429696/photo/1
You can see that 2019 mortality is completely inline with that from previous years.