My original comment was about the PhD specifically, which is necessarily research focused.

But even so, I’d argue that it is prudent as a blanket statement about grad school generally. Wages across entire classes of roles can stagnate relatively quickly – more quickly than you can get a graduate degree – and if it happens, you’re still stuck with the grad school bill.

Entering a program which costs you directly means minimally taking significantly extra risk, and it should be presented as such.