Covid-19 Statistics – lies, damned lies etc.

Statistics can be a way to illustrate some the size of some thing or things – in numbers – and variations in this, depending on something else, like time, days, months or hours. But it doesn’t have to. It can also be used to give the reader a false impression of the sizes and variations.

Reporting on the number of deaths in the Covid-19 pandemic could be expected to stand above such petty manipulations. But it does not. The diagram below shows the most recent figures from the Swedish Folhälsomyndigheten (blue pillars) and the page Sweden: Coronavirus Cases in Worldometer (in red)

The beautiful part is that the total number of deaths is the same (ok, the Folkhälsomyndigheten has 18 deaths that are reported to have happened at an unknown date, which are not in the diagram).

The seemingly opposite trends is created by the use of date. Worldometer uses the day the deaths are reported by the Folkhälsomyndigheten, while the Folkhälsomyndigheten uses the day the patient is reported to have died.

Also used is a simple trick of reporting different volumes of deaths in workdays and weekends which makes the difference between the trends greater.

It becomes even more evident when I use the trend-line function for 7-day rolling average

Depending on how the dates are used, the crisis peaked in the middle of April (Folkhälsomyndigheten) or is still growing worse (Worldometer).

This only tells that the Swedish state and its servants probably will not let the statistics be known until afterwards. Unknown date for 18 deaths from Covid-19 also tell us what level of manipulation – by for example claiming lack of knowledge – is accepted by the Swedish media.

Print Friendly, PDF & Email


One Comment

  1. Hi there! This is fascinating stuff! I followed you over from Lars Syll’s site. What I do not get is what is the ”economy” that one is putatively saving by the non-lock-down policy. In a country like Sweden, workers’ incomes will be relatively protected, effective demand can be largely supported by government spending, and exports (hit hard by the worldwide downturn) will still be needed by the rest of the world upon recovery (just as there will still be demand for Canada’s resources and other exports, other than dirty oil). So what are we really talking about here? I think we are talking one about 1) housing values (related to the rise in mortgage credit) and 2) and worries about pressures on the soft Krona-Euro peg. So is the non-lockdown policy actually related to protecting money-managers and Sweden’s institutional investors?