[swombat]

In this discussion you're assuming the "employer" as a separate entity from the "employees". This was an enormously hard dichotomy for us to overcome, because it required me and my cofounder to let go of control over the salary structure and hand that over to the company as a whole. That requires a huge amount of trust that "the employees" care about the company as much as we do, really want it to be a success, are capable of informing themselves to make good decisions on behalf of the company.

It is a very worthwhile path, but it certainly is not easy for the founders (and is probably impossible in a large traditional, top down, public company where "the owners" are a faceless group of shareholders who definitely want to believe they are in control - even though they definitely are not).

At this point, we (my cofounder and I) have pretty much zero control over how much people are paid in the company. We provide advice to people making those decisions but we are not the decision-makers. And the conclusion of that experiment is that people actually seem to care far more about getting the arbitrary salary formula right than we do, so I believe it was very much the right choice to let go of this lever and let the company design its own salary structure.

Thanks to this situation (however difficult it was to get there), the problem you outline is not there because there is no "employer" entity separate from the employees that could collude or have incentives different from the employees.