4 ms·
My wife is a pharmacist. Over the past two decades the number of pharmacy schools has increased and some schools have gotten bigger classes, such that now every
by mandevil 4y ago
My wife is a pharmacist. Over the past two decades the number of pharmacy schools has increased and some schools have gotten bigger classes, such that now every year there are about 25% more Pharm.D's made than in 2005. The result has been a predictable decrease in pharmacist quality of life- the pay is basically the same number of dollars today that you would get in 2005 so inflation has made them poorer, the hours are worse, the tech hours are worse, etc.- because pharmacy school is selling the same dream of respectable, well paying work. Law school has already done one cycle of boom, massive overproduction, bust, and is starting to boom again. I know a couple of airline pilots who face similar problems- very slippery rungs at the lowest ends of the ladder, with a lot of the feel of a pyramid scheme (generally to get enough hours to get to the higher rungs you need to be an instructor teaching other people to fly) and for the few who can grab one of the highest rungs (senior pilot at a major) quite a bit of job security and well-paid respect.
If she hadn't become a pharmacist, my wife's back-up plan was to be a chemical engineer, which generally means she would be working for an oil company baby-sitting a machine in a refinery. Over the same period that pharmacists job quality declined, ExxonMobil went from the largest company in the US by market capitalization to the 12th biggest in the US, and from what I've read the quality of life of ChemE's has declined a lot in step with that. So this clearly isn't something where "STEM jobs will save you."
I remain puzzled as to why software engineering seems to be immune to this overproduction problem. Is it just that since software is still eating the world, we haven't found a natural limit to the demand for our kind of engineers yet? Is it that we are so unproductive and failure-prone that we guarantee ourselves full employment?
- paxys 4y ago> I remain puzzled as to why software engineering seems to be immune to this overproduction problem. Few reasons: - While an ever increasing number of prospective students enter the field, a much lower number actually graduate. - People that do graduate quickly get weeded out in technical interviews. This is why you see a high disconnect between unemployed devs and companies who claim they can't find talent. - Of course the number of competent engineers in the field is going up as well, but that is matched by increased demand, as pretty much every company is now employing programmers for some business process.
- lordnacho 4y agoSoftware engineering isn't tied to one particular business model, like pharmacy, oil, or airlines. There's the "software industry" and there's all the industries who also have software developers. Big tech has had a massive boom, pushing up salaries. I'm not so sure devs outside of the official software industry are doing great (are they? Someone comment), though of course there's an evaporation of talent to the higher paying area.
- chasd00 4y agoI work in consulting ( more of a manager now but still save myself at least some code to write ). The software we develop is whatever we're assigned to. It could be finance, hi-tech, biomedical, resource (oil/gas), public service, whatever. There certainly seems to be plenty of work as we're always trying to hire and always short on staffing projects. Software Engineering is probably the widest and deepest career choice out there. For example, Web development is just a tiny piece of it.
- cwp 4y agoI'm sure that we haven't found the limit yet. Usually this kind of discussion turns to supply-and-demand economics and somebody says that if you want to hire software engineers you just have to offer more. That's not wrong, but it too narrow a view. A free labour market ensures that skilled people get allocated to the highest value work, which is a great outcome for everyone. But there are so many projects that never happen at all because software engineers are so scarce. The cost of a software project is not just paying the engineers to do the work, it also has to get over the hurdle of finding, evaluating and enticing them to join the project in the first place.
- nix0n 4y ago> I'm sure that we haven't found the limit yet. I agree with that. > A free labour market ensures that skilled people get allocated to the highest value work, which is a great outcome for everyone. Value in terms of dollars is not really correlated with productive value to society. I think a lot of the software that's being built, shouldn't be built.
- cwp 4y ago> Value in terms of dollars is not really correlated with productive value to society. No correlation? No, I don't think that's true. Sure, there are externalities, but if something truly isn't valuable, you can't make money from it. Also, what do you think shouldn't be built? I have a hard time coming up with "a lot" of software that has zero or negative value. Sometimes a startup will build something that doesn't end up being valuable, but it was at least potentially valuable, or it wouldn't have been funded.
- sumy23 4y agoObviously the software industry is growing much faster than the oil industry, so yes, the industry has not yet satiated its demand for software engineers. That time is coming, of course. You do seem to be discounting a STEM career / education, however. E.g. stating that "STEM jobs [won't] save you." Yes, there isn't much growth in the oil industry for chemical engineers or in the construction industry for civil engineers, but that doesn't necessarily mean the work or jobs are bad. You can still earn a decent living doing fun, mentally challenging work. You might not make 600k a year doing it, but you can probably get close to 400k household income if your partner also does it.
- thr0wawayf00 4y agoYou're wondering about the software industry specifically, but what you're really asking about is the education industrial complex. Universities are not in the business of properly meeting labor demands of the market with skilled workers, and they never have been. Their goal is to increase enrollment and capture as much grant and tuition money as they can. Universities don't care if too many lawyers, chemical engineers, etc. are being trained. There are several paths into software engineering, many of which bypass the education industrial complex. It's getting easier to access just the education needed to do a technical job like programming. > Is it that we are so unproductive and failure-prone that we guarantee ourselves full employment? I think there's some truth to this, but I also speculate that software devs are much more productive as a cohort than we like to let on. Just think about quickly everything in the industry changes. There's a new way to write JS apps every 3 months and a major breaking, paradigm-shifting change to React every couple years it seems like, and that's just the web side. Most workers don't have to deal with the amount of change throughout their careers as programmers do, and it is absolutely exhausting.
- reidjs 4y agoIt's true the change can be exhausting - but also I kind of enjoy the change of pace. Admittedly I may not enjoy having to learn new stuff all the time once I am older, but I am sort of addicted to novelty. I have worked hourly-type jobs that don't require learning anything new ever and they can be equally exhausting.
- bombcar 4y agoYou also don’t end up as a certificated employee like a nurse or pharm md or even a chemical engineer - so those with “IT/programming” training but end up working elsewhere or in a different line don’t get easily counted.
- htormey 4y agoFlexibility is what keeps software engineering salaries going up. Over the course of my career I’ve done firmware, desktop software, mobile and web development for a variety of industries. All on the back of a 4 year computer science degree. Example industries I’ve worked in: pro audio(Avid), consumer hardware (Apple), social networks (Facebook), medical marijuana, education, crypto currency (Coinbase). My work has impacted tens of millions of people and generated massive revenues for the companies I’ve worked for. Sadly Chemical engineers and pharmacists are very specialized jobs that often require masters degrees and hence don’t have the same career flexibility. This friction in career switching combined with the cost of education in the USA is a big problem. Their are few professions that offer the same bang for your buck as software engineering. We are a very privileged and lucky group of people.
- llaolleh 4y agoDesktop computers were released in the 1980s(Apple II, MS-DOS). If you think about it, that's only 40 years at most where the gates of software programming has been opened. Considering that everyone in the 1st world country has access to a computer, and given that much of our lives are entangled by software, it's just a demand problem. The demand for software completely outstrips the number of software engineers available. Given people's short attention spans, and the incredible number of concentration and patience the discipline requires, I expect the shortage to continue for the foreseeable future.
- _notathrowaway 4y agoI wish I shared in your optimism. From where I am standing I see more and more young people trying to get into our field, and I don't feel there is much of an and entry barrier at all.
- kelseyfrog 4y agoSoftware is eating the world, sure, but consider a parallel: the typist. In 2000 there were 282,000 employed typists in the US[1]. Today it has fallen to 23,000. A 90% decline over 20 years. Why? Everyone knows how to type now. Typing is an assumed skill in other jobs. Twenty two years ago I took a typing class in middle school on clacky old Selectrics[2]. While typing classes aren't taught in school today, I cannot help but look at the pervasive 'teach kids to code'[3][4][5][6] wage and think of the similarities. Future jobs will expect coding as a skill, not an end to itself. You can even see the inklings today with the explosion of data science, analysis, and digital marketing. In two decades, with an assumed coding skill, we'll have orders of magnitude more. Just as we can ask today, "What happened to all the typists?" we'll be able to ask, "What happened to all the coders?" 1. https://fred.stlouisfed.org/series/LEU0254503100A https://fred.stlouisfed.org/series/LEU0254503100A 2. http://ibmselectric.blogspot.com/2016/03/topaz-bronze-selectric-ii-with.html http://ibmselectric.blogspot.com/2016/03/topaz-bronze-select... 3. https://teachyourkidscode.com/ https://teachyourkidscode.com/ 4. https://www.monster.com/career-advice/article/how-to-teach-kids-code https://www.monster.com/career-advice/article/how-to-teach-k... 5. https://www.apple.com/education/k12/teaching-code/ https://www.apple.com/education/k12/teaching-code/ 6. https://www.idtech.com/blog/how-to-teach-kids-to-code https://www.idtech.com/blog/how-to-teach-kids-to-code
- moneywoes 4y agoFor a backend focused dev today what would you suggest to do to be future proof as they say?
- popcalc 4y agoMaster an obscure, esoteric language that has a stronghold on legacy infrastructure.
- kelseyfrog 4y agoAh yes, the Stenographer's Gambit.
- kelseyfrog 4y ago
- moneywoes 4y agoI hate to say it but it’s happening with software and companies like Deel that make outsourcing very easy