First, hyperspecialization. Instead of a "programming" career we have back-end, front-end, embedded, machine learning/data science, and many other silos to the career, and people are expected to be "at level" (i.e. as good as a typical person of their age and experience) in the specific silo for which they're applying. We have a zillion active languages and database technologies and web frameworks. People who choose the right specialty are in high demand. People who don't, or who fail to specialize, are more likely to lose out. When you see a two-sided scarcity problem, that's usually what's happening: a mismatch between sought skills and what is available on the market.
This is good and bad. Divergence is when exploration happens (people move away from Java and C++ to Python, Ocaml, Scala, Ruby, Haskell... and find out what works well and what doesn't) and convergence is when they move back, hopefully to better mainstays (in PL, Scala and Clojure are the front-runners; it's obvious that dysfunctional programming in the style of Java-esque AbstractVisitorHandlerFactoryFactory patterns is over) than those that were in the mainstream before. People do eventually get sick of having to change tech stacks every time they take a new job. Convergence sets in at some point, and it will, soon. We're late in a divergent phase. Both divergent and convergent phases are necessary and healthy. When we move into a convergent phase, it's going to be easier for typical programmers to find matching jobs (a reduced "curse of dimensionality" as there are fewer dimensions of variation) but some specialties are going to dry up.
Second, short job tenures. Another factor for which both sides are to blame. There are "job hoppers" out there who seek a new environment and a pay raise every 9 months, and there are companies that fire rashly, or that do such a bad job of mentoring their employees that anyone sane will leave in 6 months. Job fluidity is good to a point, but it's passed the optimal point. Now it's in a vicious cycle: an environment of short job tenures discourages companies from investing in their people, which leads to high turnover. Some startups welcome turnover with the self-satisfied and cultish belief that those who leave "weren't good enough" but that just shows a lack of introspection to a degree that's irresponsible. Extremely high turnover is just bad.
Third and most importantly, we need to handle the education problem. We will see technological convergence and increasing job tenures for macroeconomic reasons... to a point. However, technological change is going to remain rapid and continuing education is going to stop being a luxury (for programmers and in general) in no more than 10 years. People will need to access educational resources continually in order to direct their career growth, and they'll need some way to prove they've mastered skills necessary for the transitions they need to make.
The best solution for this problem, for a forward-thinking company, is to drop to 3-4 days of metered work (for most jobs) and require people to pursue some kind of education in the other 1-2 days/week. That reduces unemployment of the "not enough work to go around" sort, and it reduces the friction caused by mismatches in experience. That said, I've seen "20% time" type programs, as wonderful as they are in intention, fail more often than they succeed. Google's 20%T is only used by about 10% of engineers because there's absolutely no anti-retaliation insurance (i.e. managers can punish employees in performance reviews for taking 20%T and employees have no recourse, which means, in effect, that employees only really have it if the manager wants them to have it). Building a robust 20%T program is really not easy.
Tuition reimbursement is a good policy, but people rarely use it because it requires them to plan in ways that most professionals simply can't. How many people can predict that they won't face a badly-timed drop-everything production crisis or other hourage spike over a 15-week semester? Very few. The solution is going to have to be more asynchronous than the typical university schooling model, but also more open and self-directed than most in-house corporate training.
> There are "job hoppers" out there who seek a new environment and a pay raise every 9 months
These exist in all industries because there is too little growth for young employees who stay at a company for any time, even if loved and show promise.
Too true. So many companies, even younger ones, hold onto some stupid concept of paying your dues. Even the "great" companies give the employees who contribute 10x as much a 20% raise instead of a 10% raise...and that's AFTER a bunch of political bullshit to make sure you get recognized. Seriously? That's why even a tiny bit of effort to go out of your way to reward great people can have outsized returns...because nobody does it!
At most places there is just no concept of rewarding people according to how much they contribute. That's why talented people have to move all the time, because of stupid institutions that think a moderate increase in compensation is adequate for a massive increase in value. Ugh.
My current employer is actually really good about this, but so many places fail so completely they make me gnash my teeth in rage.
I would imagine that the internet, Facebook and constant connectivity to friends and colleagues has helped perpetuate this phenomena.
20 years ago you would occasionally here about someone who got a raise or promotion, relatively speaking. Nowadays you are connected and aware of what all your friends are doing and how they are moving on up in life. That easy ability to compare yourself with others:
e.g. "I graduated with Bob and I know I'm as good as him, if not better. Why aren't are moving up as quickly as him? Maybe I should demand a raise or look elsewhere for work if they don't appreciate me like Bob's employer appreciates him"
We today are also far more aware of when friends and colleagues change jobs so people perceive that changing jobs is a frequent occurrence and therefore change jobs more often, which then furthers the perception of job changing frequency in others. It's a virtuous or vicious cycle depending on what side you're on.
No, it's a real effect: entry level positions start at lower purchasing power then they did in the past, many requiring FREE unpaid internships. This coupled with anemic raises that do not match up with those in the past, mean companies do not reward the young to stay. This occurred before the 08 crash as well.
Yes, it is more visible what's going on in associates lives, however, real wages for young workers are declining in absolute purchasing power.
The lack of high responsibility positions opening up from lack of top of the workforce people retiring is not helping this issue any.
Totally agree with you. I wasn't saying that it wasn't a real effect. I was just suggesting that awareness may also be contributing to this changing social norm in addition to the issues you pointed out.
Exactly. In fact, I look at companies as Candidate hoppers. In the US, it is employment at will which means they can fire me anytime (except a few circumstances which could land them in legal trouble). So nothing wrong being a job hopper in today's market. But until you stay at one company, give them your best. Do not be married. May be if there were pension plans anymore. May be if their really was something known as job security. It is a myth.
Upvote for the term "candidate hoppers". That word should enter the vernacular.
The one thing I particularly like about the startup world is this rare focus on hiring the best and creating an environment that shows you value employees. One of the easiest ways to determine whether a company is a candidate hopper is to see how the quality of life is at the company. If it is low, it means that they aren't trying hard enough in general to try to keep people around.
@ malandrew. thanks and yes. You can easily spot a candidate hopper by how they treat their employees, consultants etc. It is so sad that many of the giant corporations that I have worked for so far are in that category. Yes you might have a great boss or co-workers but overall, the culture of the company reflects and disappoints.
Don't think requiring 1-2 days/week of education is going to work. People who aren't interested in pursuing new technical education/re-education will not only not going to benefit from it, but it will also be a waste of the employer's money (2X actually, for the time lost and for the education). They will just either sleep during class if physical presence is required or just find ways to cheat if done online. On the other hand, those who do want to further their skills will not stop even if they are already working overtime, have a family with kids, or perhaps even personal side projects.
The proof of the education shouldn't be passing some test or getting a certificate, it should be that you'll be required to use that knowledge in your job. If you "completed" the education and aren't able to do the work that requires it, then you didn't really understand the material.
At that point, it's up to your employer how they want to proceed (re-education, termination, etc).
Surely there must be some people who are in the middle, who would gladly pursue training if it were paid for and if time were made available for it by their employer. Consider, for instance, those people who have significant family commitments who cannot work more than fifty hours each week without suffering some hardship (maybe they have children or elderly parents or both). Just because you don't have free time outside of work to learn new things does not mean that you wouldn't want to learn new things if you had more time.
People who aren't interested in pursuing new technical education/re-education will not only not going to benefit from it, but it will also be a waste of the employer's money (2X actually, for the time lost and for the education). They will just either sleep during class if physical presence is required or just find ways to cheat if done online.
It wouldn't have to take the form of classes. It could be some sort of relevant side project or open-source work.
On the other hand, those who do want to further their skills will not stop even if they are already working overtime, have a family with kids, or perhaps even personal side projects.
Sure, but they don't grow nearly as fast... and for many, not fast enough to meet the demands of the marketplace.
First, hyperspecialization. Instead of a "programming" career we have back-end, front-end, embedded, machine learning/data science, and many other silos to the career, and people are expected to be "at level" (i.e. as good as a typical person of their age and experience) in the specific silo for which they're applying. We have a zillion active languages and database technologies and web frameworks. People who choose the right specialty are in high demand. People who don't, or who fail to specialize, are more likely to lose out. When you see a two-sided scarcity problem, that's usually what's happening: a mismatch between sought skills and what is available on the market.
This is good and bad. Divergence is when exploration happens (people move away from Java and C++ to Python, Ocaml, Scala, Ruby, Haskell... and find out what works well and what doesn't) and convergence is when they move back, hopefully to better mainstays (in PL, Scala and Clojure are the front-runners; it's obvious that dysfunctional programming in the style of Java-esque AbstractVisitorHandlerFactoryFactory patterns is over) than those that were in the mainstream before. People do eventually get sick of having to change tech stacks every time they take a new job. Convergence sets in at some point, and it will, soon. We're late in a divergent phase. Both divergent and convergent phases are necessary and healthy. When we move into a convergent phase, it's going to be easier for typical programmers to find matching jobs (a reduced "curse of dimensionality" as there are fewer dimensions of variation) but some specialties are going to dry up.
Second, short job tenures. Another factor for which both sides are to blame. There are "job hoppers" out there who seek a new environment and a pay raise every 9 months, and there are companies that fire rashly, or that do such a bad job of mentoring their employees that anyone sane will leave in 6 months. Job fluidity is good to a point, but it's passed the optimal point. Now it's in a vicious cycle: an environment of short job tenures discourages companies from investing in their people, which leads to high turnover. Some startups welcome turnover with the self-satisfied and cultish belief that those who leave "weren't good enough" but that just shows a lack of introspection to a degree that's irresponsible. Extremely high turnover is just bad.
Third and most importantly, we need to handle the education problem. We will see technological convergence and increasing job tenures for macroeconomic reasons... to a point. However, technological change is going to remain rapid and continuing education is going to stop being a luxury (for programmers and in general) in no more than 10 years. People will need to access educational resources continually in order to direct their career growth, and they'll need some way to prove they've mastered skills necessary for the transitions they need to make.
The best solution for this problem, for a forward-thinking company, is to drop to 3-4 days of metered work (for most jobs) and require people to pursue some kind of education in the other 1-2 days/week. That reduces unemployment of the "not enough work to go around" sort, and it reduces the friction caused by mismatches in experience. That said, I've seen "20% time" type programs, as wonderful as they are in intention, fail more often than they succeed. Google's 20%T is only used by about 10% of engineers because there's absolutely no anti-retaliation insurance (i.e. managers can punish employees in performance reviews for taking 20%T and employees have no recourse, which means, in effect, that employees only really have it if the manager wants them to have it). Building a robust 20%T program is really not easy.
Tuition reimbursement is a good policy, but people rarely use it because it requires them to plan in ways that most professionals simply can't. How many people can predict that they won't face a badly-timed drop-everything production crisis or other hourage spike over a 15-week semester? Very few. The solution is going to have to be more asynchronous than the typical university schooling model, but also more open and self-directed than most in-house corporate training.