Show the Community: Interesting report from PlaceIQ suggesting that cultural events like the NBA cancelling game and Tom Hanks getting sick had a larger impact on stay-at-home behavior and visits to businesses than state-level shelter-in-place policies

<!here>

Interesting report from PlaceIQ suggesting that cultural events like the NBA cancelling games and Tom Hanks getting sick had a larger impact on stay-at-home behavior and visits to businesses than state-level shelter-in-place policies.

Could that be true? Has anyone thought about that?

If anyone can provide a statistically robust version of this, we would promote it heavily with news journalism and state governments; this is a very important public health issue.

https://www.linkedin.com/feed/update/urn:li:activity:6657012734065950720/

The NBA closures happened on 3/11 and 3/13 was the President declaration of national emergency. I think it is pretty hard to separate those two events in time and the latter surely had a huge impact.

Don’t know how accurate it is, but love the storytelling through the trend lines. It could be polished a little bit and would make a beautiful data visualization.

If anyone can provide a statistically robust version of this, we would promote it heavily with news journalism and state governments; this is a very important public health issue.

There are exceedingly many events happening simultaneously, plus potential lagged effects, making it very hard to disentangle which events caused what reactions. That said, it is pretty clear that the spike in visits to grocery/pharmacy/big boxes preceded the declaration of national emergency.

For most diff-in-diff analyses, the Tom Hanks / NBA shift is in the fixed effects (partialled out).

Got curious about this. I tried a specification where I ran post(Tom Hanks news)*google_trends(“tom hanks”|date=03/12/2020). Controlling for county-level shelter-in-place + county/day fixed effects, the Tom Hanks interaction has no effect.