Basically, I think of PhDs as mostly falling into one of three categories:
1. Lifestyle PhDs. These include math, literature and the humanities, theoretical physics, history, many social sciences, and the arts. These are PhDs you do because you really, really, really love just sitting and thinking about stuff. You work on you own interests, at your own pace. If you want to be a poor bohemian scholar who lives a pure "life of the mind," these PhDs are for you. I totally respect people who intentionally choose this lifestyle; I'd be pretty happy doing it myself, I think. Don't expect to get a job in your field when you graduate, though.
2. Lab science PhDs. These include biology, chemistry, neuroscience, electrical engineering, etc. These are PhDs you do because you're either a suicidal fool or an incomprehensible sociopath. They mainly involve utterly brutal hours slaving away in a laboratory on someone else's project for your entire late 20s, followed by years of postdoc hell for your early 30s, with a low percentage chance of a tenure-track faculty position. To find out what these PhD programs are like, read this blog post. If you are considering getting a lab science PhD, please immediately hit yourself in the face with a brick. Now you know what it's like.
(Note: People have been pointing out that EE isn't as bad as the other lab sciences, with somewhat more autonomy and better job prospects. That's consistent with my observations. But econ still beats it by a mile...)I think that's a pretty bogus reading on lab science Ph.D.s, even if it is connected to the infamous (and rightly so) Guido/Carreira letter. Yes, there's a low percentage of Ph.D.s who make it into the tenure track, but that's not the goal of most/many chemistry Ph.D.s. I agree with the characterization of brutal hours, but I think "slaving away" is probably a metaphor too far.
The problem is, in my opinion, that Professor Smith conflates the jobs scenario in biology (and also neuroscience) in which supply is clearly being fueled by (past) NIH funding increases with other fields (chemistry, electrical engineering), where limited economic growth and other structural changes probably have a role to play in the relatively difficult industrial jobs environment.
He goes on to say that economics Ph.Ds. are where it's at, because you're more or less guaranteed a job. Uh, if true, good for them.
UPDATE: There's something missing in my brief analysis, which is trading the correct words for the hyperbolic "suicidal fool or an incomprehensible sociopath". Here's how I would phrase it:
Chemistry Ph.D.s are Ph.D.s that you do if you love the actual doing of science more than you love most other things, including money, status, joy and family.Readers, care to add your comments?
UPDATE 2: Corrected to note that Professor Smith is, well, a professor. Added the word "industrial."