Covid-19 and Trump's Campaign Rallies

The post that was quoted here has been removed

An academic paper concludes that 18 Trump campaign rallies (where most people ignored social distancing or did not wear masks) likely led to more than 700 deaths by Covid-19.

I'm sure this is true, in fact the death toll may be higher than that before we know the full effect of his rally's. I don't know why Trump supporters can't get it through their heads this virus doesn't care which candidate one supports - it's an equal opportunity killer. Maybe they really are "that" dumb!

The post that was quoted here has been removed

Has that paper been through peer review or is it a preprint?

The post that was quoted here has been removed

"Ideology over science" is extremely popular in America. Con men like Trump have taken advantage of this for years.

The post that was quoted here has been removed

A note of caution about that paper. If you look at Figure 1 on page 10 they show the central value for the "treatment effects" for the counties where there were rallies and the "placebo" counties, i.e. ones that did not have rallies. The 95% confidence intervals appear to overlap, although it's difficult to tell because of the way the data is presented.

The measure is the number of new cases, except for the placebo groups they have a negative figure and haven't adequately explained how the number of new cases per 100,000 can come out negative, even as an average. I need to reread it to get a clearer handle on what they've done.

In table 2 they present standard deviations rather than 95% confidence intervals so it's hard to check that they don't overlap. Normally, i.e. in drug trial results, this would be presented as a ratio and a confidence interval presented for the ratio. Instead they've worked out central figures and confidence intervals for the treatment groups and placebo groups separately. Note that the footnote on the Title page says "Department of Economics" so they're not used to presenting medical data - only an economist could think that the number of new cases could come out negative.

I'd take their results as sufficient evidence that large events ought not to take place. The effect is large and we don't need statistical significance at the 95% confidence level for safety purposes. Large gatherings may increase covid-19 deaths is enough not to gather in large numbers. On the other hand if they are trying to prove that the Trump rallies increased Covid deaths then they haven't proved it because we can't be 95% certain their results did not come about by chance.