The Journal of Public Economics has an interesting paper by Jonathan I. Dingel and Brent Neiman on How many jobs can be done at home. The TL;DR is that they looked at broad job categories and by using surveys on what doing those jobs require they determined what jobs could be done entirely at home (spoiler: 37% in the U.S.).
The research is presented in the context of COVID-19 social distancing requirements but it’s also interesting to those of us who have been interested in working from home long before COVID-19. As expected, those who do the sort of work that most Irreal readers do are at the top of the list: we can do 100% of our work from home. Obviously, if you’re serving people at a restaurant, you can’t do any of your work from home. What makes the paper interesting is that it looks at jobs between those two endpoints and considers what percentage of their work can be done from home.
The other interesting result from the paper is that the ability to work from home is positively correlated with wages. That’s not too surprising, of course. A programmer or mathematician makes more than a waiter. What’s surprising is that the graph or wages versus ability to work from home is essentially linear.
It’s a bit of a slog to get through the paper for those of us with STEM backgrounds but that’s only because some of the terms and references are unfamiliar. There’s a lot of information in it even though it’s fairly short. Worth a look if you’re interested in the work-from-home phenomenon.