The article has a bunch of speculation that’s not in the paper: https://www.nber.org/system/files/working_papers/w35387/w353...

One key point:

> Nor do I find evidence of upward job mobility among incumbent workers. Already-employed women do not increase hours, income per household member, or job-to-job moves. If anything, employer changes decline, possibly due to a desire to retain access to health insurance that covers these drugs. These null effects help distinguish a new-match channel from a general productivity or well-being channel

This complicates the idea that it’s just people treating lighter people better. It could well be that a lot of people who take GLP-1s are obese enough where it affects their ability to go out in the world and do stuff, and losing the weight enables people to work.