Search
  • Andrew Walsh MD PhD

Have the protests started the second wave of Covid-19? (update after 10 June data)

#secondwave #covid19 #georgefloyd


Data is from day 15 after George Floyd's death. There was a big spike in US daily incidence today which is not yet sufficient to alter the overall flat trend. I note this report in the Washington Post, "Coronavirus hospitalizations rise sharply in several states following Memorial Day", which relates the hospitalizations to early relaxations of stay at home orders in certain states and does not implicate the protests. I think it is too early for the protests to result in hospitalizations. At this stage most rally positives would be starting to feel ill and not yet need a hospital bed.


The New York Times states

At least 15 cases nationally have been linked to protests, including five National Guard members and one police officer in Nebraska. Health officials inParsons, Kan., andStevens Point, Wis., on Tuesday also announced new cases involving people who had attended protests.

I have several points to make:

  • I am not saying there is zero transmission outdoors, just that risk appears to be drastically reduced in comparison with indoors, especially when precautions (like mask-wearing) are taken.

  • At least one of the 15 cases reported by the Times was positive for Covid-19 at the rally, not as a result of the rally (the Parsons case). The NY Times report did not make this clear. Indeed I would expect the number of positive people attending protests nationwide was far larger than 15.

  • On a day when there were 27 234 new cases, 15 represents just a drop in the ocean (0.055%) and if anything supports my case that outdoor risk is very low.


Hennepin had a minor increase in daily incidence but overall trend is still downward.


Washington DC minor daily drop and still showing downward trend.



Los Angeles appears to have started a downward trend as I suggested it would a few days ago. Still early days though and could change.


Atlanta is not really changing much at all either way and has been hovering around 50 new cases per day for the last month.


New York City is on its same trajectory of about 500 new cases daily.


Fifteen days after George Floyd's death with resulting massive protests we are still not seeing any increase in incidence in the cities I am looking at.


Just today I saw this story in Al Jazeera, which used a similar approach to me (looking at specific cities with large protests) and they have reached the same conclusion, that there is no increased incidence in the cities they have looked at. However they were more cautious in their conclusions:

In such an analysis, the best we can do is point out whether there is a correlation between the two data sets. This is very different from proving cause and effect - especially given the multitude of factors that contribute to the spread of the virus.

This is a fair point to make but I have the following comments:

  • I predicted there would be no observable increase in incidence after the protests, based off the scientific evidence available, which is the reason I started this blog

  • What irks me about the "correlation does not imply causation" argument is that correlation is essential for causation to be true. In addition, if you predict something beforehand based off scientific models and your prediction eventuates, then it is very likely that causation was the reason rather than blind luck. Al Jazeera looked at the data post hoc, which is NOT what I have been doing.


I am becoming more confident here - hope I don't end up with egg on my face but will be the first to admit I am wrong if things change.

8 views0 comments

©2020 by Covid-19 Models. Proudly created with Wix.com