5 ms·
If you have someone worth twice as much as your other employees at the same level, then you can promote that person or perhaps figure out why your other employe
by hallihax 8y ago
If you have someone worth twice as much as your other employees at the same level, then you can promote that person or perhaps figure out why your other employees aren't as productive.
I don't see why such a situation should need to affect salary transparency. It's also useful for employees to know the standards required to earn more.
- weliketocode 8y agoThis is the correct answer in theory, but in practice, I've never seen a company that was very good at paying/promoting people based on output. It is difficult enough to measure output, especially when there's often many issues around politics and luck-of-the-draw for project assignments. Compound this with the fact that employees don't always scale with the company, or contribute at a high-level when promoted.
- kibwen 8y agoAnd yet people still earn salaries regardless of our ability to accurately gauge output. A salary is our most useful estimation of output. If salaries were known, people could look at their colleagues and say, "well, I think I'm more productive than they are, so I should be paid more", and take that to their manager, who can then counter with "actually, here's our point of view on your colleagues' productivity, and what we would expect from you in order to give you a promotion". This is a modestly effective weapon against pay that is based on internal politics (e.g. nepotism), because it forces managers to explain pay disparities in a way that employees can at least somewhat validate. IOW, if the boss wants to pay his worthless nephew more to sit around and do nothing, employees who recognize that he does nothing can effectively challenge their boss to either admit that they should be paid more, or alternatively lie to their faces in order to justify the nephew's salary, which, IMO, is a great excuse to begin seeking employment elsewhere.
- hallihax 8y agoI agree in principle - but the hypothetical was about someone who was (presumably measurably) worth twice as much as other employees. I agree that it's difficult to measure productivity / output - but it's by no means impossible - and sharp discrepancies in productivity between employees are often highly noticeable anyway.