I don't think many in the 21st century would expect a bachelor's or associates degree in english or history to provide many opportunities for additional income.
NY Times: "Does the College Major Matter? Not Really" (http://thechoice.blogs.nytimes.com/2013/04/29/does-the-colle...)
WSJ: "Your College Major Is a Minor Issue, Employers Say" (http://blogs.wsj.com/atwork/2013/04/10/your-college-major-is...)
Seems to me like there's a top tier of 3 highly technical majors (EE, CS, ME), and then a big cluster of 9 pretty decent majors (10 of the 12 at this stage might reasonably be called "STEM"). Communications is in the decent tier for women but not men. And then there are a bunch of weaker majors.
The only thing that's even remotely surprising to me is that Chemistry isn't a little higher.
I meant that for a major that's generally considered to be hard and lead to a lucrative career (it's part of "STEM" after all), it's not that different from Pol Sci or Business (and it does worse than Accounting or Economics, both of which are non-STEM), which most people probably wouldn't expect.
I would expect (but do not have data to support) that those who do accounting/econ who take a more math-oriented approach do better than those who take a less math-oriented approach.
STEM as a whole is basically the entire top half of that chart. Even math, which you call "pretty weak", is (taking the average) as close to CS as it is to communications, and much closer to the top cluster than it is to the bottom cluster.
So I think "study STEM" is perfectly good advice, particularly given that there will be shifts in the value of various degrees based on subtle economic changes.
Unfortunately, that doesn't really help out the 50% or so of the population that really doesn't have the abilities to hack that kind of material.