You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
Using @OpenTable's data, we look at states that have reopened (>5% restaurant occupancy compared to 2019) and their change in reported cases from 15 days ago.
There seems to be some weak correlation between normal restaurant occupancy and increase in cases (r = 0.49)
(1/5)pic.twitter.com/BjONZ6gxds
Out of the 5 states with >20% normal occupancy, 3 of them (Arizona, Alabama, Texas) have seen a 20%+ increase in cases over the past 15 days. The other two, Oklahoma and South Carolina, have seen their new cases remain flat. (2/5)
While Georgia's reopening have been widely covered by the media, restaurants are still operating at only ~13% of normal, which is behind 12 other states. It could be that the media attention has caused fewer people to venture out -> fewer new infections (which is good). (3/5)
I believe this OpenTable data is more valuable than looking at mobility data. Mobility is not the best indicator of how people are coming into close contact with other individuals. People can drive to the park more, but that doesn't mean they're more likely to be infected. (4/5)
Caveats: - Correlation does not imply causation - This is still very preliminary data as many states have just reopened, so new cases may not be reflected yet - This is only a subsample of restaurants covered by OpenTable: http://opentable.com/state-of-industry … (5/5)
One more caveat: - Testing capacities may have increased from 15 days ago, thus skewing the results. Nevertheless, if you assume independence, this phenomenon should be not significantly affect the correlation.
In the past 15 days Texas has DOUBLED its testing capacity that has more correlation than what you are saying
I understand that, but so has many of the other states in the plot. So the pattern would still hold if you assume independence of testing increases.
Data makes wonders
Probably no way of knowing whether the seating is indoor or outdoor. No way will I go to a restaurant and sit inside. I would consider outdoor seating with adequate distance between tables.
Can you do this with positive test rate? Or, even better, hospitalizations (but compare it to further back than 15 days to account for the hospitalization delay)
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
