"If markets were even close to accurate, consider that the range of pay for professional software developers would vary by at least a factor of 10 just around the middle of the bell curve."
That's ludicrous, because you're taking a narrow (and objectively wrong) definition of 'market' and 'market value'. The spread in pay is much lower than you suggest it should be because the market discounts for the difficulty in quantifying the marginal added value between programmers that you suggest exists - in other words, it's too hard for employers to find out about this difference, and hence there is less room in willingness to pay (well that's talking from assumptions that are favorable to your argument - I think the much more rational explanation is that the '10 times difference' myth is a programmer circle jerk) (I'm a programmer myself)
> That's ludicrous, because you're taking a narrow (and objectively wrong) definition of 'market' and 'market value'.
No, I'm just suggesting that 'market value' has little relevance to 'actual value of contributions'.
> The spread in pay is much lower than you suggest it should be because the market discounts for the difficulty in quantifying the marginal added value between programmers that you suggest exists - in other words, it's too hard for employers to find out about this difference
The thing is, if that were true, employers wouldn't be willing to spring for huge pay rises when they realise they are otherwise sure to lose a key developer. However, in my experience, they often are willing to go to those lengths.
I think there is a much simpler explanation for the discrepancy between performance and compensation for programmers: too many good programmers don't realise how much more productive they are than the bad ones, and they often don't have the kind of mindset and/or training to fight for better compensation alone, and in the absence of professional bodies/unions/whatever they will just accept what they are given. This doesn't mean they aren't worth more, it just means that as a profession software developers tend to be lousy negotiators compared to the people who do it for a living.
As for the 10x thing, it's actually much worse than that, because there are a lot of programmers out there who are clearly (to the rest of the programmers on their team) making a net negative contribution. They drain more from other positive contributors than they contribute themselves. In short, you would be better off firing them. This often doesn't happen, whether because employment laws make it prohibitively difficult in your jursdiction or just because management are too incompetent to measure and understand the problem so they can deal with it.
(I always find it odd that managers in software development groups seem convinced of their own worth and that it is higher than most people working under them, yet they rather consistently fail to measure and control even basic productivity and progress within their groups. If managers can't figure out which developers are the 10x guys and who to fire, maybe the managers need to take a pay cut to their own true and very small value until they can.)
"If markets were even close to accurate, consider that the range of pay for professional software developers would vary by at least a factor of 10 just around the middle of the bell curve."
That's ludicrous, because you're taking a narrow (and objectively wrong) definition of 'market' and 'market value'. The spread in pay is much lower than you suggest it should be because the market discounts for the difficulty in quantifying the marginal added value between programmers that you suggest exists - in other words, it's too hard for employers to find out about this difference, and hence there is less room in willingness to pay (well that's talking from assumptions that are favorable to your argument - I think the much more rational explanation is that the '10 times difference' myth is a programmer circle jerk) (I'm a programmer myself)