Managers assume that because one good programmer will do the job for $100k that all good programmers will. Many here point out that good programmers will generally find a job that challenges them, and that pays them enough to live. They then conclude that if $100k fills one seat, then $100k is the right price.
This is totally wrong, and the key is "enough to live". I love programming as much as I did when I was 10. My expenses, however, are now somewhat higher than when I was 10. I have a wife, two kids, will not live in a small house, have very expensive car insurance, and I road-race big cars, hard on tires. If you want to hire me, it will cost a lot more than $100k.
Then there is the issue of mobility. Programmers tend to be loyal. They tend to be realists. If I make X here, and I know here to be fun, why would I risk moving for 1.1X, or even 1.5X?
What you have a reversible reaction in equilibrium. You want to bind seats to programmers. Programmers will bind to a seat for a while. But not all programmers will bind for the same price. Sure you can fill a seat for $100k. You might be able to fill a seat for $20k - but you'll have to wait a loooong time. If you want to fill all your seats, you're going to have to appeal to the largest possible pool of good programmers, and some of us need more money.
Even if you ignore the fact that those of us with higher expenses have more experience, the simple math of the equilibrium reaction tells you that if you want to fill more seats, you have to have a bigger pool. Even if we are of no more value than the 20 year old, the question is: Are 5 programmers better than 4 for your business?
Whether or not that makes economic sense for your business is up to you.
So yes RethinkDB, $125k for an Engineer III will get you some hires, but are you going to insist that this is the right price, even if it means that you have empty seats? What is your opportunity cost? It may be the right decision. But it may not.