The discussion about bimodal pay distribution is interesting, but I think there's a much more parsimonious explanation to it. Most of the companies that are going to be paying software developers $150k+ are based in or around SV, where rent is abnormally high.
But what's even more interesting to me is another question the author discusses halfway through the article:
> How is it possible that programmers are paid so well without these other barriers to entry that similarly remunerative fields have?
The author posits three hypotheses: 1) supply of programmers is naturally low, 2) immigration laws keep the supply artificially low, and 3) the overall relative efficiency of American companies create a productivity force multiplier for everyone at the company, including programmers.
I want to introduce a fourth hypothesis, which is that programming skills are highly commoditized, which causes demand for those skills to increase.
If I work for OmniCorp using 100% proprietary OmniCorp technologies, they have no reason to pay me more than the bare minimum because I'm effectively married to them. If I work for OpenCorp on a 100% open-source stack, then the buzzwords listed on my resume prove I can be productive at thousands of other companies, giving them all incentive to compete for me.
But isn't OmniCorp married to their devs too? Those devs probably took a while to get up to speed on the proprietary stack, and outside devs would have a ramp up time. Not to mention outside devs would be hesitant to take the job seeing as they'd get stuck learning proprietary technology making them less mobile.
There's actually a practical case study in this. Epic Systems, which uses MUMPS, is reasonably well known for hiring naive college grads, teaching them MUMPS, and then exploiting the crap out of them.
Tech is a vast field. We are having trouble finding a Django developer where we are. Project managers decision to hire based on this, my opinion is we should look for a good all rounder, as Django knowledge only makes up around 25% of the knowledge that is needed. Someone that knows HTML, some JavaScript, any other framework, how to set up remote servers and use a database effectively. If your only skill is Omnicore, your suggestion makes sense, but there are so many different things to learn, that having a wide knowledge base makes sense.
I agree with all of that. It's a complex issue. And it's compounded by the general "we have no idea how to hire developers" that is found throughout the industry.
Do they? If you're the one developer in the world keeping their tech going, and the nearest next developer is 6 months of ramp up time away (plus the time for the hiring process), you'd have some leverage.
In this hypothetical scenario, the developer has no leverage to renegotiate, because they have no job options outside of OmniCorp, hence, no BATNA (best alternative to negotiated agreement). If the developer was homeless or independently wealthy, and not in need of a job, they might have the ability to terminate the agreement without reaching a new one. Then the company would truly have something to lose.
But OmniCorp knows that the developer has to pay rent, and no other company is going to offer them anything but an entry level job, because the only experience they have is OmniCorp this, OmniCorp that.
It may seem far-fetched, but believe me, this discussion is hitting close to home for me. My team is currently moving to a highly customized tech stack that is open-source, but has zero or near-zero adoption outside of the company. It's scary; I've been in software for eight years and I feel like I'm fresh out of college. If you're in software I hope it never happens to you.
It happened to me. I got out of it by studying open source tech stacks and building side projects until I could get a job away from the proprietary stuff. It wasn't ideal, but I managed to pull it off.
I'm mostly playing devil's advocate with my replies. It's generally better to stick with mainstream -- and ideally open source -- tech. But I don't entirely agree that proprietary tech is only beneficial to the company in these scenarios. They are taking risks that the employee can take advantage of. Heck, for me, that proprietary company still contacts me on occasion to see if I'll come back.
dont forget, when you can easily apply for 1000s of other jobs, they can also easily find 1000s of ppl out there to replace you. and as far as OmniCorp goes, you know how much IBM is paying now for one of their mainframe programmer? those guys... you lose one, you wont get another.
But what's even more interesting to me is another question the author discusses halfway through the article:
> How is it possible that programmers are paid so well without these other barriers to entry that similarly remunerative fields have?
The author posits three hypotheses: 1) supply of programmers is naturally low, 2) immigration laws keep the supply artificially low, and 3) the overall relative efficiency of American companies create a productivity force multiplier for everyone at the company, including programmers.
I want to introduce a fourth hypothesis, which is that programming skills are highly commoditized, which causes demand for those skills to increase.
If I work for OmniCorp using 100% proprietary OmniCorp technologies, they have no reason to pay me more than the bare minimum because I'm effectively married to them. If I work for OpenCorp on a 100% open-source stack, then the buzzwords listed on my resume prove I can be productive at thousands of other companies, giving them all incentive to compete for me.