Which means job that used to be either low-demand, low-paid, or (usually) both. Otherwise demand has to keep rising, because without barriers, an increase in pay will cause a delayed increase in supply (as people are attracted to the field) which then meets the demand, or at least evens it out, reducing pay in turn, unless demand keeps increasing.
My own 2$ ameteur micro-economic theory there :-)
Anyway, the point is this; Is hard to predict which jobs will pay well when you start on that path. well paid jobs now might be lower paid by the time you and every other chaser-of-the-dollar goes into the same field, increasing supply, pushing down pay. The opposite can happen too.
Builder/labourer might not be seen as too lucrative in the past, yet in some times it certainly has been. And I wonder how many BAs actually fail to get high paying jobs?...
Yes, there may be micro-fluctuations where every company is suddenly scrambling to hire the latest flavor of the month (e.g. data analyst is now data scientist). These are hard to predict. But you know who will reliably earn more than the data scientist in three years? The data scientist's boss.
> But you know who will reliably earn more than the data scientist in three years? The data scientist's boss.
I'm not 100% sure of this. There will be a boss in the chain that earns more, but actually I suspect that this isn't always true, given one assumption - the boss, as a job requirement, has to have a good knowledge of DS (maybe an ex/senior DS?), in which case this barrier ensures management will get higher pay - otherwise they'd just stay/go into in DS.
A well-paid manager saying, "But I could have made more developing Minecraft!" might as well complain, "But I could have made more if I won the lottery!"
If high compensation is the goal, being a manager is simply a much more straightforward and reliable path than being a programmer.