Friday, May 15, 2020

Very Very Brief Blog: WSJ and Weekly Death Data, March to April

WSJ publishes data (through week of April 11) on weekly deaths from all causes in the US.

All deaths per week are normally about 58,000 per week (yellow line).  2020 deaths by week were flat to March 21, just elevated March 28, and then spiked over the next two weeks to April 11.  


The raw death rate per week rose from about 1 per 6000 to 1 per 4500.  That's an a raw death rate per year rising from 1:115 to 1:88.

Note that death rate is a lagging indicator to COVID infection events; many of the people people dying April 1-11 were likely infected in mid-March (say, March 25).  Known death certificates and public health data directly attribute about 2/3 of the increase to COVID-19.   To me, this shows the speed of essentially uncontrolled COVID spread up to the 3rd week of March.   With states "reopening" (which often means restaurants at 50% seating or less, theaters closed, etc) the increase over 5-6 weeks would be much less than the uncontrolled rate shown here.   It will be interesting to have weekly death rate data for the weeks in 2H April, which will come out closer to June 1 and which would be reflective of new COVID events under pretty generalized lockdown (closed theaters, most shopping closed, etc).