The thing is a lot of people value the security or aren't aware of the difference between market and their pay. Let's say this is 80% of people.
If you're corporate HR, you can retain 80% of your top talent without big pay increases and promotions while knowing that you'll lose the top 20%. This may be palatable whereas it wouldn't be if you were only retaining 60%.
I think Google and FB were the first to figure out that you could get the top 5%, that they would make a big difference in your org. You'd have to greatly increase compensation but those individuals are probably contributing and worth more than that.
Depending on the job market it can also be very expensive (time vs yield) to hunt for job offers. If there are hardly any offers, and if the companies all look like rather bad employers, or the job looks like it will pay probably less than you already make, then you waste lots of time on nothing. So people just stay at their jobs.
I'd really really like to have something like Stackoverflow jobs (i.e. well-described jobs), but also with information what they would pay (seriously: this should be public; hiring is a business decision, and hiding the price only wastes people's time), and with something like the company's kununu profile linked directly, so I can see how well they treat their employees.
Maybe software development is different from more traditional professions, but if you can hire a top performer for 20% higher salary, they'll probably be as productive as 5-10 average developers, so it's really worth it. Well, over here nobody in HR seems to share that opinion.
I think the people doing the 2-year-hop are the ones with a lot of inbound interest. The last two jobs I've had (including my current job) came to me through a direct inquiry from someone at the company (not a recruiter / headhunter). The thing with consistent inbound interest is that it creates zero incentive for putting up with non-regular advancement / promotion (except the 'devil you don't know' risk, but that can be ameliorated by asking around. It's not hard to avoid a completely blind scenario) because the perception of "job hopping" becomes irrelevant. I suppose at some point the inbound interest might taper off but until it does, it's almost irresponsible to ignore better offers.
You can retain 80% of your talent. You will be able to retain approximately 0% of your top talent, which is certainly a number less than 20% of your total talent pool.
You may be surprised how many top engineers have no idea what they're worth, and can't self-promote worth a damn. I've known some incredibly talented, capable people being woefully underpaid, and not ever trading up.
If you're corporate HR, you can retain 80% of your top talent without big pay increases and promotions while knowing that you'll lose the top 20%. This may be palatable whereas it wouldn't be if you were only retaining 60%.
I think Google and FB were the first to figure out that you could get the top 5%, that they would make a big difference in your org. You'd have to greatly increase compensation but those individuals are probably contributing and worth more than that.