The competence crisis

In Sweden at least people are talking about a competence crisis, referring to the fact that almost nobody is studying IT at university these days and that companies are consequently having huge problems finding people to hire. How bad will it be in a couple of years?

When a discussion has been a bit one-sided for a while, there must be some grumpy old man who jumps in and says the opposite. I don't think I'm extremely grumpy and not all that old either, but I'd like to toss a few words into the debate.

Please note that I know there are problems in the lack of skilled people so I am not saying it's not a problem. OK, now that's out of the way, over to my arguments.
  • Nope, I don't see off-shoring as the solution to the whole problem. Sure, off-shoring could be an interesting alternative in some cases, but it's hardly going to be the mainstream solution when the hype is over. And if the project is easy to specify in details, why should humans have to transform that specification into code instead of using another approach? Letting developers and domain experts work together, in one and the same room, is extremely efficient and something you don't want to or should give away easily in many projects.

  • If we look back, we can find several instances of forecasted crises that never happened. One example that springs to mind is the story about the expected amount of horse droppings on the streets of London. Quite some time ago a researcher did some extrapolation and stated there would be an enormous problem a few decades later... then came the car and the problem never arose. (OK, there were other problems of course and you might wish for the horse problem instead, but that misses my point here.) Something might radically change the scene. What will that be? We'll see...

  • Brook's law is well known and often referred to ("Adding manpower to a late software project makes it later"). Even so, strangely enough it's tried every so often, but that's a story for another day. Instead, I wanted to take Brook's law one step further. I quite often "joke" about how late projects should move half of the developers to other projects so that the late project can deliver; and sometimes going down by 50% isn't a big enough percentage. Let a very small, but extremely skilled, team work in peace and collaboration. Get the best developers you can find. Quality over quantity. Even if they cost you twice as much as the average developer, that's cheap if they are three times more efficient. Simple and obvious, isn't it?
To sum up my attempt at an "opposite" opinion, perhaps it's even "correct" that only a few are studying to become system developers. At least if those who are are the most interested and suitable. Admit it, you weren't expecting me to say that!