It doesn't get said enough, but I believe that the naive belief that a math student will be fine is a reason for too many grad students, Carl got it right. They seem to think research is just a different way of continuing into 'Undergrad 2.0' and never give the risks a second thought.
Only the classmates that had their PhD topic and academia in mind a few months, if not years, before they applied, enjoyed their PhD - made a hell of a difference as they seemed happier with life choices, even if they did a job afterwards. That, and a naive undergrad belief that employers look at postgraduate degrees as 'better' than undegrad drives this oversupply of PhDs. I have classmates that hated research from day one and wound up finding jobs in unrelated fields like accountancy just to escape grad school - complete waste, when an MSc would have given more stimulating options (and in less time in one case). I've also met PhDs in the last moments of a PhD panting wondering if they will quit - I remember trying to explain the wrongs of quitting to some clown finishing off a physics PhD. She spoke to me like I was a 4 year old as she'd be minted in her view - as far as I know she completed and 1-2 years later she was still under-employed doing leaflet distribution jobs at minimum wage (her jobsearch skills seem horrifically bad to be fair).
Makes a hell of a difference knowing what you are doing with a PhD, while drifting into an MSc isn't as bad as long as the course is your kind of stuff.