Tyler Cowen points to a new paper which suggests that working from home during the COVID pandemic produced more deterioration in mental health than working at a workplace. In particular, WFH has negative effects on loneliness, depression, and life satisfaction. This is intriguing, but here's the key sentence:
When considering all three dimensions of mental health together, WFH causes a 0.087 standard deviation increase in the overall measure of mental health deterioration compared to WP.
Hmmm. This is about like finding a height increase from 5'10" to 5'10¼". Even if it's (barely) statistically significant in a technical sense, the result is driven almost entirely by an increase in loneliness—which is hardly unexpected—and the overall effect size is simply too small to take very seriously. That's especially true in a paper like this one, which relies on a mountain of model building, statistical interventions, measurement of unobserved abilities, and variable controls.
There might be something here, but I'd want to see some pretty clear confirmation before I really believed it.
I we really supposed to buy that you can derive a believable statistical value for something like this? And frankly, if Tyler Cowen is siting it, well then forget it. Maybe he can convince his buddy Bronze Age Pervert of it.
Why would anyone assume that *working from home* during COVID caused an increase in loneliness, and not, you know, the overall lockdown.
And how are we to suss out the difference between people having social anxiety choosing work from home?
Seems like underlying assumptions of the group source are poor.
I think that mental stress came not from WFH but from kids at home.