I don't deny there's age discrimination, but that survey also says JavaScript is 2x as popular as Java, and 3.4x as popular as C++
Whereas [1] says among job board postings, C++ is about as popular as JS, and Java is 1.7x more popular than JS.
Perhaps the holders of StackOverflow accounts aren't representative of the profession as a whole? I could certainly believe people who were programming professionally before SO existed, and who have decades of experience, would have gotten good at solving problems without asking for help.
> Whereas [1] says among job board postings, C++ is about as popular as JS, and Java is 1.7x more popular than JS.
I think it's questionable whether job posting are more representative of working programmers. Arguably they're less representative. Job postings are unfilled jobs, so technically they represent exactly zero programmers. And many job postings are never filled.
> Perhaps the holders of StackOverflow accounts aren't representative of the profession as a whole?
Agreed, but unfortunately I can't think of a better source of data offhand.
> I could certainly believe people who were programming professionally before SO existed, and who have decades of experience, would have gotten good at solving problems without asking for help.
I don't agree with the assumptions here. As an experienced developer, I ask for help all the time. I'm not an expert on every possible subject, and anyway, tech changes. Also, not everyone on Stack Overflow is asking for help. Some have to be offering help, otherwise nobody would go there.
You're assuming that the ratio is disproportionate because people are being pushed out of the industry against their will. Some people get burned out, some people retire early, and some transition into management or start their own company. The industry has grown exponentially in the 20-40 years those people have been working. If the industry now is 10x the size it was in 2000 (arbitrary number), then 40 year old developers should be 10% of the workforce, even discounting any other factors.
I don't see anything alarming here. I am a 40+ year old developer, and I've never seen anyone my age leave the industry because they were victims of age discrimination. The opposite in fact. I have never had so much clout during job hunting than I do now, with 20 years of on the job experience. Anecdata sure, but that survey is not sound, and it does not attempt to discover whether older developers are actually unable to find work, or what other factors (other than the obvious massive industry growth) might contribute to the imbalance.
I'd say if management routinely burns out programmers and prevents them from having long careers, that in itself is a form of age discrimination.
> If the industry now is 10x the size it was in 2000 (arbitrary number), then 40 year old developers should be 10% of the workforce, even discounting any other factors.
IMO this seems to be an ageist assumption. If an industry is growing, why should the newcomers to the industry be overwhelmingly young? Why shouldn't "older" people join in large numbers too, especially if there's a lot of money to be made? I personally switched careers and didn't get my first tech job until I was over 35.
> I have never had so much clout during job hunting than I do now, with 20 years of on the job experience.
What do you make of the pervasive "quiz show" style job interviews that favor recent CS grads and discount job experience?
In any case, I'd say there's more empirical support for my position than the claim I was arguing against, that people work 80 hours per week and then retire early because they "value their time". Just not their current time, only their time later in life (if they live that long), I guess.
> I'd say if management routinely burns out programmers and prevents them from having long careers, that in itself is a form of age discrimination.
It's not management though that may be a part. Building complex systems with cutting (i.e., bleeding) edge technology is hard. A lot of projects fail, often with career consequences for the people in them. If you experience 2-3 of these you start to look around to see if there's an easier way to make a living.
But that "professional developers" histogram is almost indistinguishable from "all respondents," so really your statement should be "the age distribution of StackOverflow developer survey respondents is different from that of the overall human population."
What's the sampling bias of this survey? How many pro developers over 40 were contacted for the survey but did not respond? Is StackOverflow usage higher among people aged 20-39 compared to 40-59, all else being equal?
https://insights.stackoverflow.com/survey/2020#developer-pro...
Only 17% of professional developers surveyed were age 40+. That vastly disproportionate with the age demographics of the overall human population.