Calculating the risk of Thanksgiving dinner

Want to know the risk of being exposed to COVID-19 at your Thanksgiving dinner? In a paper released today by Nature Human Behaviour (part of the Nature family of prestigious journals), a team from Georgia Tech has created a handy-dandy online calculator that can do just that—as well as calculating the risk of being exposed at any other social gathering of any size, anywhere in the U.S. (and in some other countries).

Cutting edge science it is not: the math is basic probability. But it’s a cool website to play around with, and the announcement is very, very timely.

It’s primarily oriented toward policy-makers trying to decide what types of gatherings to allow, but it’s also revealing to any of us who are simply curious…and concerned about stemming the rapid spread of the virus.

Simply use the slider to pick the size of your event, then hover the cursor over your county, and it will tell you, based on current infection rates, the probability that at least one person is infected. (Note: it does not work on all browsers. Here’s the direct link in case you need to copy it to another browser. For me, it works on Firefox, but not Safari. https://covid19risk.biosci.gatech.edu/).

For example, it tells me that in Portland, Oregon, a gathering of 50 people has a 38 percent risk of having one or more infected people present. Ouch. Having ten people over for Thanksgiving isn’t as risky, but still carries a 9 percent chance.

And it’s a lot worse elsewhere. In my brother’s county in Iowa, a 50-person church service carries a stunning 97 percent chance of drawing an infected person, and that not-all-that big Thanksgiving dinner has a 50 percent chance.

There are some caveats.

  • The figures are based on the presumption that cases are under-reported by a factor of 10, something that is backed up by antibody tests in some parts of the country, but not others.
  • It assumes that everyone who tests positive is infectious. I don’t know if that’s true.
  • It does not tell you the chance that you yourself will become infected.

But if you’re looking for a reason to wear a mask, avoid large (or even mid-sized) indoor gatherings (and crowded outdoor ones), and to postpone major holiday celebrations, this is it. Me, I’ll celebrate Thanksgiving on my own, looking forward to next year, hoping to maximize the chances that my friends (and their loved ones) will still be alive to do it.

Risk and the Psychology of the New COVID-19 Surge

The rising number of COVID-19 cases has many causes, including the question of how many are being found due to increased testing capacity. (Likely answer: some, but not all. Hospitalizations are also rising, which means more people are getting sick at levels that would have been detected in the virus’s first surge.)

But another part of the answer has to do with what, in retrospect, should have been a predictable aspect of public psychology: as the pandemic has dragged on, COVID-19 has gone from new and frightening to “normal.”

One of the more esoteric fields I’ve dabbled in over the years is the psychology and economics of risk.

Psychologists have found that our fear reactions do not correlate well to the actual risks posed by a given hazard. Economics has confirmed this by noticing how people, in spending money to guard against risks, tend to spend disproportionate amounts on various types of risks.

Two examples are air travel and bicycle helmets.

Continue reading Risk and the Psychology of the New COVID-19 Surge