If he wants to participate in the annual meeting of his student’s club, he considers how that can be done while maintaining distances. When he walks his dog, he estimates how many people he will meet, when and how well they have followed the regulations. If he wants to buy the goods that used to be sold by the baker, the butcher, the greengrocer and the chemist, he wonders whether or not he will go into the supermarket wearing a mask and gloves or whether he will have them delivered via a web shop. (His own economy is stable as an ABP retiree, it seems.) If he wants to gain knowledge about the pandemic he sees overwhelming amounts of data and stories. Mr. Sum has now been on the pandemic’s track for 37 weeks, in his personal bubble.
It is striking that in the debates an old distinction is often ignored: the distinction between (i) statements that can be tested and falsified by observations and (ii) statements that comply with convictions based on dogmas. What strikes him most is how easily someone like Maurice de Hondt, who should know better, jumps from a tested thesis about aerosols to the belief that keeping distances to curb COVID-19 infections is unnecessary. Such a belief has a tunnel-vision smell and does not explain multitudes of falsifying observations at all.
The observations do not point to simple, one-track inferences at all. On the contrary, thinks Mr. Sum, who gives an overview of data for the Netherlands during / after the first 36 weeks of the pandemic in Fig. 1.
Fig 1 has four plots. The information is too much to discuss here. But Sum does venture to formulate a number of statements that can be tested by observation.
- In the Netherlands, four weeks after the registration of the first Dutch COVID-19 death in week 10, a policy was initiated that could explain the deviation between the ratios of the observed and the R0-based curves. A deviation that becomes apparent through calibration at week 13 (see Graph 0 in Fig. 1). Week 13 ends on March 31, 2020. Rutte and Grapperhaus announced the “intelligent lockdown” in a press conference on March 23.
- In the Netherlands, the numbers of COVID-19 infections and deaths are in line with each other during the first weeks (up to week 20). Thereafter, the numbers of deaths stabilize while the numbers of infections increase sharply (see Graph 1 and Graph 2 in Fig. 1). The factors that contribute to this development are unclear and probably not simple.
- Of the nine jurisdictions that Sum involved in his project, the COVID-19 policies can be compared for efficacy in a general number: the GDD number (the general domestic death toll). How that figure (expressed in the number of deaths per million citizens) developed in the first 36 weeks of the pandemic in the nine jurisdictions can be seen in Graph 3 (in Fig. 1). Two structures of GDD curves can be distinguished: the first tends to horizontal smoothing and the second to linear growth. There are three of the latter type: the USA, Brazil and South Africa. It is remarkable that the economic developments in the latter three are among the most unfavorable (see Fig. 2).
The figures are based on data from two databases, one from the EU and one from the OECD. Fig. 1 and Fig. 2 arouse wonder. Why are there such differences between jurisdictions? And why are the rates of deaths per infections so different in the first and later periods? Answers are likely to be complex. Perhaps something for Maurice de Hondt to focus and to spend his energy on?
What could be taken into account is that the time lag between infection and the occurrence of severe symptoms in the first and second waves will be recorded differently. That could lead to unpleasant surprises. For example, in the first wave, the detection of infections would be limited to and roughly coincide with the appearance of worrisome symptoms – while now, in the second wave, the testing practice has been considerably broadened so that infections become known at a much earlier stage. That could indicate that a severe 2nd wave increase in people with severe symptoms is yet to come.
If not, we have an interesting enigma.