The reality is pretty simple, though: a lot of people making decisions like this will look at their career options when they graduate and have to face the real

The reality is pretty simple, though: a lot of people making decisions like this will look at their career options when they graduate and have to face the real world, struggle for a while, then realize that they're more than qualified to make twice the money as a junior programmer and make the switch then, recouping their costs easily unless they're really stubborn. Half the people that I went to school with in physics, math, econ, bio, did exactly that. They're doing just fine now because TBH, those fields are vastly more intellectually challenging to learn at a research level than being a well qualified working programmer is (doing proper CS at a grad level is a different story, on par with those other fields, but you don't need to be a research-class CS PhD to be an above average programmer), and most of them already had to learn to program to get through their courses anyways, so it's an easy transition. It's the people in completely non-technical majors that have the most trouble, because they don't know math, they don't understand proofs or algorithms, or even basic problem solving, let alone how to design systems, and worse, they've never really had to learn to effectively use software except in extremely limited ways.