> Additional evidence is that changing jobs results in much higher wage increase than staying put, if wages were connected to skill this would not happen
This can definitely happen if wages are linked to value.
You are not identically valuable to all companies, and since changing jobs is something that usually has some general cost/risk associated with it, it's regularly done for larger rather than smaller salary increases.
>You are not identically valuable to all companies
This is very true and certainly affects the very theoretically top limit of the job offer. But when deciding compensation the actual calculation is only remotely based on the value the employee brings your particular problem.
In practice an offer is mostly based on the employee's current market value (ie how easy it is to get someone cheaper) and how much he is currently making (ie the minimum amount I can offer and still get him to accept) and some social acumen around ability to negotiate a salary.
So yes, when you change jobs your salary is more likely to rise if an organization can use you more effectively. But it will always max out based on how easy you are to replace. This and only this is the pay you will eventually reach if you change jobs enough. And that has no connection to the value you produce but is solely connect to the number of people with your skill vs number of openings.
I don't see this argument in the article.
> Additional evidence is that changing jobs results in much higher wage increase than staying put, if wages were connected to skill this would not happen
This can definitely happen if wages are linked to value.
You are not identically valuable to all companies, and since changing jobs is something that usually has some general cost/risk associated with it, it's regularly done for larger rather than smaller salary increases.