As an opposing anecdote, I've spent a significant percentage of my software development years in the medical arena. The overwhelming majority of doctors that I have worked with do not work as hard, anywhere near as many hours, or under as much stress and the overwhelming majority of software developers I have worked with.
The stressful period for most doctors is getting into medical school, getting through medical school, and getting into and through internships and residencies. That is roughly 10 years or so, and during that time it is absolutely at the pace and pressure I see in software development..but once they finish, for most of them, it is pretty smooth sailing. Those at the top of the game are a different story, but that's true anywhere.
The real difference is that doctors have a powerful lobby that artificially limits their numbers, essentially providing a benefit similar to unionization. You can't massively up the numbers of doctors being trained in the U.S., and you can't import massive numbers of foreign workers to increase supply. No matter how many capable people there are, the numbers will remain artificially constrained.
Software development doesn't have this protection. The only lobby in software development is regularly pushing for massive increases in supply through various channels. Doctors' salaries have continued to rise year after year..not so sure software development salaries will fare as well in the coming years..especially after we convince software devs they are already overpaid ;)
Software will continue its bifurcation into two income distributions, similar to lawyers. No lawyer earns average salary. That industry is split into low wage work and high wage work. I expect software will be the same.
It's hard to tell what the distribution of income is without good data. I'm skeptical that normal, non-startup businesses fall into this pattern. I'd expect more of a log-normal for most industries.
The log-normal also applies somewhat for lawyers. The partners of big firms make millions annually.
As long as we're getting into nuance, I should add that the two-peak distribution is most clear for lawyer starting salary, not long-time professional. You either go top firm or not.
Great points all around! The reason I think small businesses are relevant is that the founders/owners own the entire business. They either lose money, break even, or make money. In the sucessful small business cases where they make money, since they own 100% their profit requirements for a good return are far lower than those of a startup. I'd guess it's still a power law scenario (just with less amplitude than startups).
Very on point about the AMA functioning as a union. I'm not a software developer (and don't even live in the US) so apologies if this question is ridiculous: could software developers form a union, er I mean the 'American Software Developers Association'?
There are strong network effects in many digital markets. In these cases, given enough time, it leads to a monopolist employer in that market. Even in cases where there are a few firms competing (e.g. Google and Apple), they have incentive to form a cartel through 'agreements' on hiring practices (and there have been documented instances of this). The natural way for software developers to avoid being exploited, in either scenario, is to form a labour monopsony...
I'm not a software developer (and don't even live in the US) so apologies if this question is ridiculous: could software developers form a union, er I mean the 'American Software Developers Association'?
Well, the real nuance is that the AMA is not so much a union as a guild. It's not _legal_ to perform medical services without an AMA-blessed licence to practice.
If one could concoct a scheme by which it would be illegal to perform "unlicenced" software engineering work, I suppose there would be the same effect.
It's hard to see this happening since dissatisfied software developers can often fix the problem by finding another job. (Maybe not everyone can do it but that's the perception.)
You'd need a common sense of injustice. It's a lot easier to do this for, say, graduate students who really are underpaid.
There is. It's called the PE (Professional Engineer) license. Today, however, it mostly only applies to civil engineers. Back in the day, if someone wanted a bridge or building (especially for the government) designed, you had to be a PE to get the job. It required a certain amount of hours of work after college and an exam. Doesn't really apply to software though.
--When I got my first job as a software engineer, my dad, who's a civil engineer PE said, "I thought you were going to be an engineer, but it sounds like you're just a computer programmer." :-)
There are plenty of coders in their 60s, I'm one of them. You don't think computer programming was invented in the dot com boom do you? I have colleagues still coding who are in their 70s.
And no we aren't all coding in Cobol or Fortran 4.
I do agree though that it would be difficult for most of us to get new jobs in another company as developers because of the age-ist behaviour of many of our younger colleagues.
I guess your 22 year old founder of a trendy Bay Area startup would have a problem hiring a 60 year old but that's both silly and illegal. I'm not sure trendy Bay Area startups attract the 60+ crowd anyway.
I don't think we'll know the answer to the age question for a little while. The number of coders has been growing over time, so the potential pool of seriously older coders (60+ for sake of argument) is already small to start with.
Anecdotally, three very different examples:
my mother (71) stopped coding in 1969
Guido van Rossum (60) is still coding
Darryl Havens (61) is Distinguished Engineer at Amazon - not per se a coding role, but definitely an engineering role
The stressful period for most doctors is getting into medical school, getting through medical school, and getting into and through internships and residencies. That is roughly 10 years or so, and during that time it is absolutely at the pace and pressure I see in software development..but once they finish, for most of them, it is pretty smooth sailing. Those at the top of the game are a different story, but that's true anywhere.
The real difference is that doctors have a powerful lobby that artificially limits their numbers, essentially providing a benefit similar to unionization. You can't massively up the numbers of doctors being trained in the U.S., and you can't import massive numbers of foreign workers to increase supply. No matter how many capable people there are, the numbers will remain artificially constrained.
Software development doesn't have this protection. The only lobby in software development is regularly pushing for massive increases in supply through various channels. Doctors' salaries have continued to rise year after year..not so sure software development salaries will fare as well in the coming years..especially after we convince software devs they are already overpaid ;)