I guess one question I have re the first study you posted is: just because you get a certified LCA for level 1 or 2, does that mean the employer is going to actually pay what corresponds to a level 1 or level 2 wage? Isn't that the floor? LCA happens before I-129, so the employer is just certifying they need someone, anyone, to take the role.
For instance, a promotion or regular pay raise doesn't trigger a new LCA. That only happens when you change employers. As an employer, if it didn't affect the likelihood of an approved LCA (and the article does itself state that they're largely pro forma, and won't get denied unless there's an obvious error) why would the employer not pick the lowest class to maximize flexibility?
I actually have only read the synopsis of the first one. I only included that link because it was mentioned in the article. And I had quoted from the article's interview with that author.
The one that convinced me previously was the second link, which was a pretty convincing natural experiment leveraging the fact that the visas are awarded by (randomized) lottery.
For instance, a promotion or regular pay raise doesn't trigger a new LCA. That only happens when you change employers. As an employer, if it didn't affect the likelihood of an approved LCA (and the article does itself state that they're largely pro forma, and won't get denied unless there's an obvious error) why would the employer not pick the lowest class to maximize flexibility?