Economists Jonathan Dingel and Brent Neiman have an interesting new white paper out: “How Many Jobs Can be Done at Home?“

Key takeaway: “Our classification implies that 34 percent of jobs can plausibly be performed at home.” Let it suffice to say that this does not bode particularly well for the headline unemployment rate.

Speaking of which, the U.S. unemployment rate stood at 3.5% in February 2020, or 5,787,000 unemployed persons relative to a civilian labor force of 164.5 million. On Thursday we saw initial unemployment insurance claims skyrocket to 3.3 million for the week ending Saturday, March 21. You can think of initial UI claims as a floor for job losses, because not all workers who lose their jobs qualify for UI (e.g., part-time employees or workers who recently moved between states). Holding the size of the labor force constant, if we see, say, 4 million workers transition from employment to unemployment in March, the unemployment rate would surge from 3.5% in February to 5.9% in March (9,787,000/164,546,000=0.059).

For context: The largest monthly increase in the U.S. unemployment rate during the Great Recession was 0.5 percentage points, while the largest monthly increase since WWII was a 1.3 percentage point increase during the recession of 1949. Via FRED: