I admittedly haven't read the paper properly, but did study economics at undergrad and read many similar papers. This is how a lot of economics works, you end up with published papers claiming completely different results because of small differences model assumptions. Multiple models can be consistent with observed data while providing vastly different policy implications. If you want to look at a well publicized example (though I think it might have to do more with different econometric assumptions) look at the research published on the Mariel boatlift and its impact on labor markets. See here - https://www.bruegel.org/2017/06/the-mariel-boatlift-controve...
Which shows the original result standing and the "controversy" was caused by the later subsample having a different (racial) composition.
The only issue is people outside your organization, but they have no idea what a given title means. Scrum Lord’s make 500k, wonder what they do?
The dev who can whip out tons of stable code is easily measured. The dev who fights to stop an unneeded re-architecure or veto feature requests that would have maintenance overhead unjustified by their value is harder to judge. And yet that veto may be the equivalent of 2-3 FTEs savings.
They did not seem impressed. I did get hired though...
More seriously, I've met several people who really struggled to get recognized officially for their work. One was someone who was just amazing at cross team collaboration. Specifically, we were a small group, doing an important strategic project, but we needed some work from other teams that were serving the companies larger, single product. It was hard to get them to free up a couple of dev-days, even when we had the explicit support up and down the management chain. She was amazing at getting the ICs on the team to get on board. But really struggled during review cycles because a fair amount of her work wasn't traditional lines-of-code/jiras.
Another example was a developer who was a king at driving customer solutions through the dev team. He was a developer, and even though he's probably responsible for millions in not-lost-sales, he struggled to get the recognition he deserved.
Anyway, just agreeing with some examples from my professional life.
The Indian cricket league IPL has a public auction every year. Salaries range from 2 million at the top to 20/30K at the bottom. And most players are towards the bottom.
If companies could actually reduce wages through pay transparency, they would be giving that information away.
If you are an employer in a jurisdiction with pay transparency, and you need to get a superstar employee, you just create a titled position for them. Then you can justify paying them 350% the average wage.
The paper claims that there will be a wage equilibrium. Yes, for average employees there will be a wage equilibrium, and average employees will see their wages increase because of this.
The paper tries to claim that this wage equilibrium will negatively affect superstar employees. This is complete nonsense. It bases this on the assumption that employers cannot justify discriminating between normal employees and superstar employees. But this is an unwarranted assumption. Literally just give the guy a title. Secondly it assumes that even for these superstar employees their wage would reach an equilibrium. This is another misguided assumption, because these superstar employees are not a fungible resource like the rest of the employees in the same occupation. They very likely possess unique experience and skillsets. So the pay of other superstar employees should not have any bearing on how much they themselves get paid.
Basically the model seems to think that somehow all the employees are equal and entirely fungible, but at the same time some employees somehow manage to negotiate much higher salaries.
Now of course I am not an economist so it is ridiculous to believe that I have found the fatal flaw in a published paper, but because of the political nature of such articles I find it hard to trust the "hard numbers". I remain skeptical.