The problem with on-the-job training is that the quality of it is highly dependent on the workplace and the coworkers around you and the job in question.
Just look at programming and software engineering. It takes very little to call yourself a programmer, and on-the-job training may or may not help at all.
The problem is likely that most people are willing to learn on the job, but then take that learning and go get a better job. A large portion of this problem is companies not being willing to increase pay to match the value you add by learning that skill, but it does lead to the doubt about bothering to train.
Job training by employers doesn't work great, because the worker can get training at an employer that gives training and then switch to one who pays more because that employer doesn't have to do training.
Though possibly that could be fixed with proper regulation.
The problem with training is that you hire 10 people. 3 of them pan out. Then they leave for 100% more salary from a company that doesn't need to make up for hiring those other 7 in the first place.
In my experience companies that rely on armies of junior programmers to do things manually tend to end with horrible code bases and a lot of technical debt. Few abstractions, lots of repetition, little automation and so on.
There's plenty of training available once you're in a company, because most companies understand that there's a large cost to losing an employee who's proven that they are intelligent, have a good work ethic, and are easy to work with.
There's very little appetite for training new workers until they've proven those same things.
In essence, on the job training might be in short supply, but on the job retraining is not.
The problem is that there's few companies who can afford to invest in training. Only big established companies can do it (and they do, extensively). Small companies and startups don't have the time or funding to do it.
Not only is it expensive, but training people in transferable skills can all too often result in them quitting right after their training is complete and going to work for a company that invests in paying its employees more money instead of training.
On-the-job training emerged because there was no way for workers to gain certain skills they needed without working under a corporation.
Sometimes the limiter was capital expenses....You couldn't go to your toolshed and train on how to work in a steel mill, so instead you got trained on the job.
Sometimes the limiter was social.... such as with lawyers---they receive on the job training because you can't just graduate and start taking cases for "practice".
On-the-job training will remain alive and well for people who are hiring into specialized positions (e.g. database scaling specialists, engineers for financial firms, anything to do with hardware, etc.).
Is there some institutional barrier to training new hires? It seems like the inevitable result if you need an increasingly niche skill. The situation doesn't seem like it'll get better on its own.
There is a huge value in on-the-job training, but that value comes from a direct alignment between the work and the learning.
Time delays in the University model (for one thing) make it hard to produce what is needed. For instance if you want to "learn deep learning", you could get a PhD, and maybe 9 years later you will be an expert, but 9 years is a very long time and the field may be tapped out and deep learning PhD's will be driving cabs.
The phasing out of on-the-job training is an atrocity, IMO, especially since what tends to replace it is training at a university, which ought (imo) be a place where people can get some education that isn't oriented primarily toward making them better workers.
It's especially ironic since often the university is a public institution, and the tech world (and the business world generally) tends to be so down on the public sphere and are big boosters of private enterprise. But they're not above taking a giant, giant handout in the form of worker training and then pissing and moaning when they think it isn't coming fast enough.
There's no such thing as someone you're not going to have to train. Every single company does something different. To suggest that you're not going to have months of train up isn't feasible. I wish this mindset would go away. Training is, and should be an integral part of any workplace. It's what makes your whole team better, it's what makes everyone have a similar style, and shows people the expectations. This goes for internal code process as well as external code and tools. Embedded especially is going to be hard because it's a specialized skill set that's far harder than people realize especially once you get into space constraints.
There is definitely an issue of matching skills to jobs and geography. That said, many companies that used to consider training and employee development no longer do so. So, even in locations where there is a workforce that might be trained for a job situation, companies aren't training...
I think the lack of training is probably worldwide. I'm in the UK and have seen it in most companies here (although can only provide anecdotal evidence).
Training is expensive and increased productivity is hard to measure - coupled together you see training as a cost with no tangible benefit.
They then just expect their employees to learn everything in their own time (which from your employers perspective, is free). There's a kind of horrifying logic to it really.
The issue is that a company that (all else equal) doesn't spend on training has more budget for salaries. Employees who already have skills generally value salary over additional training.
Just look at programming and software engineering. It takes very little to call yourself a programmer, and on-the-job training may or may not help at all.
reply