If you believe that there are those differences, that you can judge those correctly (but elect to not act on it), and that other employers can also judge that well enough to offer differentiated higher salaries to the most valuable employees, then you will be left holding an adversely selected set of employees. (Other employers will cream off the most valuable by being able to offer higher compensation, and you will see more of the left side of the distribution applying to your company once it becomes known that "everyone is paid the same".)
Not in every case (maybe a high-value-creating individual believes in your company and mission enough to forgo a higher salary they could command elsewhere), but in aggregate, I believe you signal to the highest performers that they should look elsewhere and your lowest performers should stay, because you are more interested in equality than fairness.
For example, master these 5 skills at level I, and we then meet to set up a process for learning the next 5 skills needed to advance to level II. Once you advance, you get +20% salary.
I think that seems like a good way of incentivizing performance and continued learning.
Or you can start doing super complicated things like "have 5 out of those 10 skills", but are they as valuable?
Yeah, that's a pain. And then another company may value the skill differently, and you then have to just lose your employee to them because they'll pay them more but you won't because of your rigid structure.
There's so many variables, uniformity here won't get you the best teams.
Every scheme I've heard of to get around that has had had some major flaw or other. Usually by failing to capture some important area of endeavour; or by relying on some combination of easily gamed metrics (like tickets closed or evaluations by co-workers); or requiring employees to capture excruciatingly detailed metrics; or by having incentives too small to be worth pursuing; or by including a large management fudge factor that makes the rest of the system pointless.
The reward usually never materializes. You shouldn't do overwork in the first place.
What end's up happening is people in the same year are sorted by many other things. Better more interesting tasks are given to higher status workers, they avoid more grunt work, get nice perks (at top levels, things like a company car, etc.), and more respect.
It's hard to separate the effects of this on the company from all the other things Japanese companies do which are different from USA/Europe. But you might look into that.
Now this certainly has moral pitfalls, if you use it as a carrot to get people to keep upping the ante on each other, but at some point we have to trust the autonomy of our workers to be able to say no. (I believe that management can help with this as well concurrent with keeping salary-based differentiation, by things such as enforcing not working outside of work hours, mandated vacation time, salary transparency and clear metrics for what achieving more/success means, etc)
I don't understand the problem here. Support agents are unlikely to demand a developer's salary.
This is, in part, because every job will value someone slightly differently depending on the unique situation.
That's still not a reason to be secretive. In fact it was cited by the article as one of the great positives of salary openness: you got much more accurate feedback much more quickly which made it easier to make those tricker decisions.
In my experience, this has never been true. People have different motivations, skills etc.
Although if we can assume that its true, it would be easier to implement govt mandates like Paycheck fairness act.