As I’ve been pointing out for weeks, not all economic activity is inherently equally risky or safe in terms of spreading this particular infection. For example, a visit to a podiatrist (foot doctor) is likely less risky than a visit to an ears, nose, and throat specialist. (But that’s just my opinion on the relative risks.)
This suggests that some jobs should be reopened quickly and other jobs cautiously.
But how can we tell which is which?
We should look at infection / hospitalization / and death rates by occupation. If people in your type of work were seldom infected or hospitalized or dying, then you should get back to work. In contrast, if your type of worker has been dying like flies, well, you need serious plans to upgrade your activities to make them safer.
This is fair and objective.
But how can we get the data for this?
America has all sorts of privacy laws protecting infection and hospitalization information.
Yet, we also have laws protecting the privacy of your 1040 tax returns, but that hasn’t stopped Harvard economist Raj Chetty from getting his hands on “anonymized” 1040 data that nobody else had the gall to ask for. So, some Chetty-like data maestros (such as, to take a random example, Raj Chetty) should be proposing study protocols that get around HIPAA constraints.
But if these regulations prove insurmountable barriers to epidemiological science, unprivileged outsiders could still datamine newspaper obituaries for the occupations of people who have recently died. For example, in New York City, 9,780 people died in the 30 days ending April 4, 2020, 5,530 more deaths than the average NYC March-April period and almost 2,000 more than the 3,350 deaths attributed officially to coronavirus. (So, NYC appears to be undercounting coronavirus deaths.)
Another study says about 30% of deaths in NYC were people under age 65. Many of them would be retired, but you could probably find a sizable sample of people who were working up until their recent demise. So, look at the obituaries for New York City in March-April 2020 and count how many people died by different occupation categories. Then divide the deaths by occupation for previous years’ average March-April deaths.
The Index would give us a rough indication of how risky each occupation was. For example, it was reported recently that 41 NYC subway workers have died of coronavirus. (Update: now up to 50 deaths.)
Is that a big number? I don’t know. How many NYC subway workers die in the average month of March? If 40 die in the average time period in years past, then that’s not a big number. If the average has been 10, then subway workers were at risk about 4x greater than normal. You could do an apples to apples comparison by reading old obituaries.
For purposes of data analysis, I’d keep two separate counts: people who officially died of coronavirus, and people who just plain died. As we see in NYC, total excess deaths over the previous month have been about 1.6 times the number officially attributed to coronavirus.
This kind of analysis could be performed by amateur outsiders using public records, such as newspaper obituaries and the like. For example, sabermetrician baseball analysts: You got anything else to analyze at the moment?
Update: How about you Wall Street quants? For example, consider how valuable it would be to figure out the relative risk of infection of flight attendants, since their numbers could serve as a proxy for being a paying passenger on an airline. If flight attendants are surprisingly less infected than you would expect, then airline stocks should go up as the rest of the world figures out that flying isn’t as dangerous as they might have assumed.
There is money to be made by figuring out the relative risks of different economic activities.