Sturgis Rally caused "266,000 Cases" - Bad Statistics Alert!

The Sturgis Motorcycle Rally sounded like a really bad idea from the start.  Covid has been politicized and bad info abounds.  But a recent study (it even made Fox News) claims that 266,000 cases of Covid are attributable to the event.  The authors of the study claim the rally can be linked to 20% of the cases in the US.  Just google Sturgis Covid and you'll see links to all the articles.

I don't think this study passes the "smell test"!

Fox News loves to play up stories that appeal to the right.  This one appeals to the left.  My guess is that the story is so attention getting to anyone, that Fox ran it and continues to run it since it is an attention getter.

Why don't I like this study?  First, I must admit I didn't go deep into the weeds trying to understand the details of the analysis.  But I don't think I need to due to the study's overall methodology.

Here are my concerns with the study in no particular order.

1)  The study inherently assumes correlation equals causation.  It looks at a period from mid to late August.  During this period, Covid cases were starting to soar in many parts of mid and rural America.  Now it isn't a mystery that wearing face masks, socially distancing, and avoiding crowds are going to bring down Covid spread.  And not wearing masks, not socially distancing, and attending large events are going to increase Covid spread.

Areas having a portion of residents with behaviors more conducive to Covid spread are also areas more likely to have residents willing to travel to an event where Covid rules are broken.  So, when those areas have high rates of Covid spread, is that because many non-event-attending residents were breaking the rules, or because the actual attendees spread the virus?  In any area, the number of event attendees is a small portion of the total population. 

2)  Around 460,000 folks attended Sturgis.  That is less than two tenths of one percent of the US population.  I know the Sturgis event was "bad", but was it so bad that they caused 20% of infections? Think of it this way - let's suppose three percent of Sturgis attendees were Covid positive.  That would be just under 14,000.  Each attendee would need to infect over 19 people to get to the 266,000 infection number the study attributes to the event.  Typical R values for areas with a lot of infection are probably around two to three.  Areas with stable levels of infection have R values of one.  This doesn't pass the sniff test.

3)  The study identified source communities for Sturgis visitors, and then looked at their infection rates in order to do their extrapolation of infection.  The county that had the highest number of source visitors was Adams county in Colorado according to the study.  But I pulled up the infection data for that county, and I could discern no Covid infection spike during the study's period of analysis.  And I didn't see a rise in other Colorado counties that were also identified as significant source counties for Sturgis attendees. There may be reasons for this buried in the study's methodology.

So perhaps I'm missing something, but the study's claim of 266,000 Covid infections due to the Sturgis rally seems false.  That doesn't mean it was OK that it happened, and it doesn't mean a number of folks won't suffer terrible consequences.  But I like accuracy and honesty, and this study seems to have neither as best as I can tell.