3 ms·
Salaries are negotiated with far less information on one side and with less vigor than contracting rates, so that doesn't surprise me. If every hire knew what y
by lock-free 6y ago
Salaries are negotiated with far less information on one side and with less vigor than contracting rates, so that doesn't surprise me. If every hire knew what you paid every one of your reports, and they did likewise, the bands would probably be narrower.
- conductr 6y agoInteresting take about information availability on labor supply side. I feel like it's the opposite though. When I've been a part of BigCo salary band creation projects; the information we use is widely available. I've actually been a part of one that started by scraping glassdoor, salary.com, etc. Then we felt it would be good to buy the official data sets and they essentially where the same. However, it's not extremely clear how the equivalent would happen in contracting where it's usually custom bids and rates are much higher because they include non-labor components. What I know for a fact happens, is if (for example) I were Accenture bidding on a project, I know I'm probably up against Deloitte on this project, let me buy my old coworker a beer, the one that now works at a Deloitte, magically both bids have same rate or Accenture slightly undercuts. This is labor intensive though, so what I'd really do is just bid you a price based on what I or my boss already know Deloitte is going to bid because we've already had those beers and the conversations are just common knowledge at this point. So it's not really collusion, but has similar consequences. On the other hand, salary bands typically are referred to as "market rates" which is a misnomer. There's no way an efficient market should have that amount of deviation as you mentioned.