I'm of course concerned about the protests contributing to epidemic resurgence of #COVID19. They are outside, but close contact and large crowds are risk factors and if they continue day after day with the same group of people you risk amplification. 1/8
We'll be able to measure this to some degree by looking for increasing case counts in counties with sustained protests and by looking for increased case counts in younger demographic. 2/8
But this may very well be confounded with transmission in these counties linked to arrests as @zeynep mentions here. 3/8
And everything is playing out on a background of increased mobility as cities and states open up. Figure from apple.com/covid19/mobili…. 4/8 
I'm not a trained epidemiologist, but I'd think the best way to address this is to have public health across the country ask "have you attended a protest" during case investigation phone calls following up with positive cases. 5/8
This should (once controlling for confounding variables) give a direct estimate of the increase in exposure risk from attending a protest. 6/8
Questions about travel, workplace, community or social event, etc... are standard practice in case investigation (see excellent overview by @CDCgov (cdc.gov/coronavirus/20…) 7/8
However, I haven't seen sharing of large-scale deidentified datasets of these case investigations with general risk factors. I'm more familiar with more detailed outbreak investigations in the published literature. 8/8
Follow up: Multiple people have pointed out the risk of answering this question truthfully. I don't know how to make this safe to answer. We want to know if the activity is dangerous from a public health perspective. But absolutely don't want to add risk for those protesting.
I like this idea
Another clever idea to help preserve anonymity
