Can I be really cynical here? I might rephrase this "Too many kids quit science because they're wise as well as smart".
I've been on this hobby horse for a while here on HN, and I'm worried people are getting tired of me. But seriously, think about it. I don't have attrition rate numbers for typical undergraduate STEM degrees, but I do know that undergrads don't drop out of Economics or Poly Sci or Literature to major in CS, Engineering, Physics, and so forth. Where I do have data, look at attrition rates for graduate schools, especially at the elite level. A typical attrition rate for an elite Law or Medical school is below 1%. A typical attrition rate for an elite Engineering or Science PhD program is 35-50%! MS programs aren't s studied, but nation wide, attrition rates are about 33% (I'd have to dig though my HN comments to find the cite for that one, and it was behind a paywall).
So, considering how difficult all this is, outcomes must be much better for those STEM grads, right? Well, no, not really. CS is a good major at the undergrad level, and grads do pretty well. But again, at the elite level, top software developers, even from elite CS schools with grad degrees, just don't pull it in like dermatologists or radiologists. I read a pretty heartbreaking story about a woman who was denied the right to use sick leave for maternity leave from her grad program. My cousin, a radiologist, worked part time (20 hour weeks) for a while when she had a kid, and she still pulled down more than my salary as a developer. I know another woman who is an emergency medicine physician who did the same thing. They are so, so, so much better off not going into science. A RAND study found that the American aversion to STEM graduate degrees in favor of professional degrees is economically rational and shouldn't be a head scratcher. The head scratcher is why people still act like the only thing we need to do to get more people into science is "make it cool", or "improve middle school math", and so forth.
We have got to stop acting like there's something surprising here. Science gets kids who are very smart and utterly devoted but perhaps lacking in wisdom, and who just can't believe that they might be in the bottom half of a cohort with high grades in STEM coursework and test scores above 95%ile. Or, alternatively, they are students who don't have these other options (US med and law schools are not nearly as accessible to international students as engineering and science PhD programs are). So if the "professions" are cut off to you because you aren't a US citizen, your decision to go into a graduate STEM program (especially if it fast tracks you to US residency) may be rational in a way that wouldn't hold for someone who already possesses US citizenship (who typically shun science PhD programs).
I've been on this hobby horse for a while here on HN, and I'm worried people are getting tired of me. But seriously, think about it. I don't have attrition rate numbers for typical undergraduate STEM degrees, but I do know that undergrads don't drop out of Economics or Poly Sci or Literature to major in CS, Engineering, Physics, and so forth. Where I do have data, look at attrition rates for graduate schools, especially at the elite level. A typical attrition rate for an elite Law or Medical school is below 1%. A typical attrition rate for an elite Engineering or Science PhD program is 35-50%! MS programs aren't s studied, but nation wide, attrition rates are about 33% (I'd have to dig though my HN comments to find the cite for that one, and it was behind a paywall).
So, considering how difficult all this is, outcomes must be much better for those STEM grads, right? Well, no, not really. CS is a good major at the undergrad level, and grads do pretty well. But again, at the elite level, top software developers, even from elite CS schools with grad degrees, just don't pull it in like dermatologists or radiologists. I read a pretty heartbreaking story about a woman who was denied the right to use sick leave for maternity leave from her grad program. My cousin, a radiologist, worked part time (20 hour weeks) for a while when she had a kid, and she still pulled down more than my salary as a developer. I know another woman who is an emergency medicine physician who did the same thing. They are so, so, so much better off not going into science. A RAND study found that the American aversion to STEM graduate degrees in favor of professional degrees is economically rational and shouldn't be a head scratcher. The head scratcher is why people still act like the only thing we need to do to get more people into science is "make it cool", or "improve middle school math", and so forth.
We have got to stop acting like there's something surprising here. Science gets kids who are very smart and utterly devoted but perhaps lacking in wisdom, and who just can't believe that they might be in the bottom half of a cohort with high grades in STEM coursework and test scores above 95%ile. Or, alternatively, they are students who don't have these other options (US med and law schools are not nearly as accessible to international students as engineering and science PhD programs are). So if the "professions" are cut off to you because you aren't a US citizen, your decision to go into a graduate STEM program (especially if it fast tracks you to US residency) may be rational in a way that wouldn't hold for someone who already possesses US citizenship (who typically shun science PhD programs).