But no, the reason the protests may have led to net fewer transmissions is simpler, according to the authors of a new study. Namely, Americans are scared of violence. Since the protests began they’ve seen police precincts burn, shop windows smashed, and mayhem various and sundry committed. That risk of rioting gave them an extra incentive to hunker down at home at night when protests sprang up in their own backyard on top of the preexisting incentive not to socialize for fear of being infected.

If you’re afraid that going out will lead to you being caught in the middle of an angry mob, you won’t go out. Result: Less social distancing.

Specifically, after the outbreak of an urban protest, we find, on average, an increase in stay-at-home behaviors in the primary county encompassing the city. That overall social distancing behavior increases after the mass protests is notable, as this finding contrasts with the general secular decline in sheltering-at-home taking place across the sample period (see Appendix Figure 6). Our findings suggest that any direct decrease in social distancing among the subset of the population participating in the protests is more than offset by increasing social distancing behavior among others who may choose to shelter-at-home and circumvent public places while the protests are underway. This latter behavioral pathway is consistent with studies that find perceived safety concerns, crime, and violence to be significant barriers to participation in physical activity and spending time outdoors (Janke, Propper and Shields 2016; Roman and Chalfin 2008; Foster and Giles-Corti 2008). Moreover, to the extent that the killing of George Floyd, and the ensuing civil unrest and mass protests, also impacted residents’ mental health and stress, this may further lead to decreased activity levels and an increase in stay-at-ho me behaviors.

“Okay,” you’re thinking, “but couldn’t this effect be explained by protest-related curfew orders?” The authors thought of that:

In addition to their direct intended effects, curfews could also signal to residents the potential for the eruption of violence, and may further elicit a behavioral response due to this information signal. We find increases in staying-at-home behaviors across both cities with and without curfews, though effect magnitudes are expectedly larger when the city has imposed a curfew order at some point during the protest period.

That reminds me of the data from early on in the pandemic showing restaurant reservations in cities like New York dwindling to zero even before the lockdown orders went into effect. Most people don’t wait until the local government springs into action to begin conducting threat assessments of their environment and adjusting their behavior accordingly.

The “hunker down” effect detected by the study leads me to wonder if this is why we’re seeing infections trend younger in hot spots like Florida now. Older people had reason enough to stay home and away from crowds before the protests, but if they’ve become convinced that some rampaging Antifa mob’s going to steamroll them once they open their front door then they’re *really* not going out.

The most important COVID story right now is the age shift.

In Texas: Young adults driving the spike.https://t.co/7LbB7MWBDR

In Arizona: COVID cases growing 2X faster among ages 20-44 than 65+.

In Florida: Median age of new COVID cases fell from 65 in March to 35 this week —> pic.twitter.com/UOcaeMN0fx

— Derek Thompson (@DKThomp) June 22, 2020

It also occurs to me that these two factors, fear of infection and fear of mob violence, may have been especially potent deterrents to Trump fans who might otherwise have been inclined to attend the Tulsa rally on Saturday. Public health officials spent the last few weeks before the rally warning locals that a mass gathering indoors was a death trap. (Most of those who ended up attending didn’t bother wearing masks. Oh well.) Meanwhile, according to reporters like Dave Weigel who were on the scene, the anxiety about lefties using Trump’s rally as a pretext to make a violent show of force in Tulsa was real:

Something that kept occurring to me: Even after the George Floyd protests became peaceful, Fox was playing and replaying footage of a store being burned in Minneapolis. How do you over-hype the threat of protests then ask people to crowd into downtown?

— Dave Weigel (@daveweigel) June 21, 2020

Result: Just 6,200 people showed up. I wonder if Tulsa’s growing epidemic ironically might see a bit of cooling off in the next week or two due to extra social distancing driven by fear surrounding the Trump rally.

That said, just because some cities have fared well lately amid the protests doesn’t mean all have. A noteworthy quote from the LA Times:

For third day in a week, L.A. County reports more than 2,000 new coronavirus cases…

It’s “highly likely” that the surge is connected to mass protests that erupted in recent weeks over the death of George Floyd, L.A. County Public Health Director Barbara Ferrer said.

Here are the latest daily national numbers, by the way. We continue to follow a strange, seemingly paradoxical two-track system. Cases keep climbing, dragging the positivity rate up with them, but deaths keep dropping:

US daily numbers via @COVID19Tracking:

Newly reported deaths
Today: 285
Yesterday: 297
One week ago (6/15): 375

Newly reported cases
T: 28K
Y: 26K
6/15: 19K

Newly reported tests
T: 475K
Y: 499K
6/15: 448K

Positive test rate
T: 5.9%
Y: 5.3%
6/15: 4.2%

— Nate Silver (@NateSilver538) June 22, 2020

July is shaping up to be a significant month. We’d expect to see the higher number of infections happening right now generate an uptick in deaths eventually. But there’s a chance that we won’t see that, that cases will continue to spike or plateau at some elevated level but deaths will remain low. What will explain that? Some seasonal effect from the heat? The virus losing potency as it mutates? Stay tuned.

One more point to consider. It’s almost universally assumed that Americans won’t stand for another lockdown as cases increase. Given the evidence we’ve seen in the study above of people adjusting their behavior without orders from the government, it’s probably beside the point whether new stay-at-home orders are issued. People will increasingly self-isolate as the threat in their community increases whether the mayor says anything or not. But there may be stronger support for lockdowns than we think if Morning Consult is correct:

People are willing to see nonessential businesses shuttered again by a margin of better than two to one. Huh.

In lieu of an exit question, go read about how dismally America has failed to control infections relative to another hard-hit country, Italy. We had 32,000 new infections on Sunday. They had 264. Even adjusting for population differences, they’re seeing less than five percent of the number of new cases we’re seeing. More than half of the 50 states are now experiencing a transmission rate above 1.0 according to best estimates.