5 ms·

Where have all the graduate students gone? (2002)

34 points by physcab 18y ago

timcederman 18y ago
The title should include the note that the article is from 2002. I'm sure that given the current economy it's not so much of a problem now.
- physcab 18y ago
  Thanks. Fixed it! By the way, even though it is outdated, I still see some of the problems prevalent today. Does anyone know where to find current statistics? I tried browsing through the U.S Census data but that was almost hopeless.
  - RK 18y ago
    You can find US physics data here: http://www.aip.org/statistics/ http://www.aip.org/statistics/
tsally 18y ago
Not sure how relevant the problem is today, but most analysis of this situation is all wrong. It's K-12 math and science education that's the real cause. Developer A: "Our compiled code is slow!" Developer B: "It must be a problem in our implementation, we should fix that." Developer A: "But our C compiler is absolutely horrible. In fact, all it does is compile the C code to Python and then pass that to a Python compiler." Developer B: "Well why does it do that?" Developer A: "We don't have funding for an actual C compiler. State law requires us to use the Python compiler for everything. [1]" Developer B: "It's too late to fix the compiler. We're already in the project development phase. We can't fix the fundamentals, just optimize the code!" Developer A: "Ok fine, how about we change compilers for the next project?" Developer B: "Nah, that'd take too much work." [1] Note: accurate reflection of the logic used in curriculum planning for most K-12 city schools.
- blogimus 18y ago
  As far as your statement on K-12 being a problem, this topic has been discussed widely (read is a key concern) in the ACM. Here's a resource: http://csta.acm.org/ http://csta.acm.org/ (Computer Science Teachers Association)
  - tsally 18y ago
    Yeah, I've been pretty happy with the ACM's response in regards to Computer Science. Math on the other hand, is a completely different beast. At least with CS there is an easy profession like game programming that students can relate and aspire to. No similar profession exists in the mind of a kid for Math that has the same appeal.
    - blogimus 18y ago
      That's interesting, I find myself in the opposite environment. I've worked with many more mathematicians than CS grads. It might just be the environments I'm in.
    - jimbokun 18y ago
      Math could conceivably be better preparation for game programming than Computer Science.
      - tsally 18y ago
        It actually is. But not in the mind of a kid. :)
Tangurena 18y ago
One reason not mentioned in this article (but mentioned in others like at The Chronicle of Higher Education) is that changes in immigration and visas post-911 have strongly discouraged students from coming to the US. Many of the students who would have come, are prevented by ICE. Consequently, many don't even try for student visas in the US and seek higher education in other countries (for example, UK and Australia). Another reason not mentioned is the soaring cost of education in the US. Rising far faster than inflation, higher education has become unaffordable for many. Post-baccalauriate degrees are even further out of reach. And as for domestic students, the salary difference between a bachelors and masters (at least in electrical engineering, which was my first bachelors) would never pay for the time and money needed to get the degree. When I got my BS back in the 80s, the difference between BSEE and MSEE was $5k/year salary differential. Between the money and time lost getting the degree, presuming zero cost of money, going for the masters would have paid off after about 30 years in the workforce. People in engineering are likewise worried about where the next generation of engineers are going to come from. Students have been voting with their feet and wallets to get away from short-term professions like engineering (the half-life of an engineering career is 7 years, which isn't long enough to pay off your student loans before you have to find a different career).
- electromagnetic 18y ago
  > "Consequently, many don't even try for student visas in the US and seek higher education in other countries (for example, UK and Australia)." My wife's currently in college here in Canada and so are many of her friends, apparently all the colleges here are tied to some university in the US offering discounts and stuff to international students.
- lliiffee 18y ago
  The cost of education isn't much of an issue. PhD programs in the sciences almost always cover tuition and provide a stipend in exchange for teaching or research work.
  - blogimus 18y ago
    But there is the opportunity cost of missing real income when you stay out of the work force.
cia_plant 18y ago
The obvious answer is that a good science student can make 6 figures today as a programmer or in finance, or they can go to grad school and work harder for much less money. My own reasons are actually different. I'm not particularly interested in money, and I love to teach and to do research. The academic environment just seemed stifling to me. As someone who values creative freedom enormously, the phrase "publish or perish" is very frightening.
- asciilifeform 18y ago
  Where, if at all, did you finally find creative freedom?
  - cia_plant 18y ago
    I haven't really found a good way to support myself while doing what I want, yet. Right now I live frugally and work part-time as a programmer. One reason that I hang out at news.yc is that I'm hoping to turn some of my ideas into some sort of startup, even though I don't really fit the profile of a startup founder (I'm not very driven or passionate, at least not when it comes to business.)
gamble 18y ago
"...there are far more post-docs working at universities than there are permanent jobs for them to move into, so a lot of people get stuck in limbo." The author answers her own question. Since the era of expansion in higher education and private research labs driven by the baby boom and cold war came to an end in the 1980's, the scientific job market hasn't been matched to the number of qualified applicants. If anything, there should be fewer grad students. For all the rhetoric about higher education as a bastion of left-wing politics, academics put up with an employment system that would make Ayn Rand smile. Aside from an ever-shrinking pool of tenured professors, most academics have low salaries, high workloads, little-to-no job security, and few employment prospects outside academia.
- davi 18y ago
  Yup. I came up with some back of the envelope numbers around this here: http://news.ycombinator.com/item?id=470181 http://news.ycombinator.com/item?id=470181
  - gamble 18y ago
    Interesting analysis. I suspect it's only going to get worse in coming years, with the economic crisis taking a big chunk out of endowments and professors' retirement savings alike.
dkarl 18y ago
I dropped out of a top-ten math PhD program because I looked around and thought, "Most of these guys aren't going to get jobs, and I don't feel like one of the exceptions. I'm not going to spend five years in graduate school to end up a failure." Of course, by "job" I meant a tenure-track research position in a livable city, and by "failure" I meant... well, basically, by "failure" I meant my father, who taught at a small, bad public university in a rural wasteland. I'm pretty sure my mom hated him for every hour and minute she had to live in that town. If I picked up a wife or girlfriend in the hip, thriving city where I lived, would she follow me to Bumfuck, USA? Would I let her if she wanted to? I constantly obsessed over the fear of ending up teaching hopeless kids at a double-directional school in a little town at the end of the universe, watching my kids grow up alienated from their provincial peers and underserved by the school system, burdened with the knowledge that it was my failure that forced my family to live there. So I quit and got a job as a programmer to keep myself afloat while I figured out what to do with my life. It turns out there are lots of challenges in computing, and I've worked alongside a couple of math PhDs on interesting problems. I think grad school would have been a lot more fun if I had realized that failing to get a hot tenure-track position wouldn't ruin my entire life.
giardini 18y ago
Many U.S. citizen know that it's not worthwhile to get a Ph.D. and in some instances, a Master's degree. See Philip Greenspun's "Career Guide for Engineers and Computer Scientists" at http://philip.greenspun.com/careers/ http://philip.greenspun.com/careers/
Dilpil 18y ago
"Being a scientist is still a comparatively low-stress, high-pay career" Compared to jobs done by those without college educations, sure. Compared to the many career options open to anyone who is in the position to become a scientist, not at all. The author, like Obama and all the others demanding that (other) young people become scientists and engineers, are living in a fantasy world where a) science is an amazing career path an b) there is a shortage of scientists and engineers.
varjag 18y ago
"Learning physics is actually not so different a process as learning a foreign language." Uh, no.
asciilifeform 18y ago
"Adjusted for IQ, quantitative skills, and working hours, jobs in science are the lowest paid in the United States." (http://philip.greenspun.com/careers/women-in-science http://philip.greenspun.com/careers/women-in-science) Despite agreeing with everything in the article, I am still devoting my life to somehow getting into graduate school - simply because my life goal is to eventually spend all of my prime waking hours working on beautiful, elegant, entirely non-commercial concepts, and there is still a non-zero number of academic positions which allow this, in contrast with industry. Like choosing gladiatorial combat in place of meekly kneeling for beheading, it is an issue of having a sporting chance vs. no chance at all.

34 points by physcab 18y ago

The title should include the note that the article is from 2002. I'm sure that given the current economy it's not so much of a problem now.

physcab 18y ago
Thanks. Fixed it! By the way, even though it is outdated, I still see some of the problems prevalent today. Does anyone know where to find current statistics? I tried browsing through the U.S Census data but that was almost hopeless.
- RK 18y ago
  You can find US physics data here: http://www.aip.org/statistics/ http://www.aip.org/statistics/

tsally 18y ago

Not sure how relevant the problem is today, but most analysis of this situation is all wrong. It's K-12 math and science education that's the real cause. Developer A: "Our compiled code is slow!" Developer B: "It must be a problem in our implementation, we should fix that." Developer A: "But our C compiler is absolutely horrible. In fact, all it does is compile the C code to Python and then pass that to a Python compiler." Developer B: "Well why does it do that?" Developer A: "We don't have funding for an actual C compiler. State law requires us to use the Python compiler for everything. [1]" Developer B: "It's too late to fix the compiler. We're already in the project development phase. We can't fix the fundamentals, just optimize the code!" Developer A: "Ok fine, how about we change compilers for the next project?" Developer B: "Nah, that'd take too much work." [1] Note: accurate reflection of the logic used in curriculum planning for most K-12 city schools.

blogimus 18y ago
As far as your statement on K-12 being a problem, this topic has been discussed widely (read is a key concern) in the ACM. Here's a resource: http://csta.acm.org/ http://csta.acm.org/ (Computer Science Teachers Association)
- tsally 18y ago
  Yeah, I've been pretty happy with the ACM's response in regards to Computer Science. Math on the other hand, is a completely different beast. At least with CS there is an easy profession like game programming that students can relate and aspire to. No similar profession exists in the mind of a kid for Math that has the same appeal.
  - blogimus 18y ago
    That's interesting, I find myself in the opposite environment. I've worked with many more mathematicians than CS grads. It might just be the environments I'm in.
  - jimbokun 18y ago
    Math could conceivably be better preparation for game programming than Computer Science.
    - tsally 18y ago
      It actually is. But not in the mind of a kid. :)

Tangurena 18y ago

One reason not mentioned in this article (but mentioned in others like at The Chronicle of Higher Education) is that changes in immigration and visas post-911 have strongly discouraged students from coming to the US. Many of the students who would have come, are prevented by ICE. Consequently, many don't even try for student visas in the US and seek higher education in other countries (for example, UK and Australia). Another reason not mentioned is the soaring cost of education in the US. Rising far faster than inflation, higher education has become unaffordable for many. Post-baccalauriate degrees are even further out of reach. And as for domestic students, the salary difference between a bachelors and masters (at least in electrical engineering, which was my first bachelors) would never pay for the time and money needed to get the degree. When I got my BS back in the 80s, the difference between BSEE and MSEE was $5k/year salary differential. Between the money and time lost getting the degree, presuming zero cost of money, going for the masters would have paid off after about 30 years in the workforce. People in engineering are likewise worried about where the next generation of engineers are going to come from. Students have been voting with their feet and wallets to get away from short-term professions like engineering (the half-life of an engineering career is 7 years, which isn't long enough to pay off your student loans before you have to find a different career).

electromagnetic 18y ago
> "Consequently, many don't even try for student visas in the US and seek higher education in other countries (for example, UK and Australia)." My wife's currently in college here in Canada and so are many of her friends, apparently all the colleges here are tied to some university in the US offering discounts and stuff to international students.
lliiffee 18y ago
The cost of education isn't much of an issue. PhD programs in the sciences almost always cover tuition and provide a stipend in exchange for teaching or research work.
- blogimus 18y ago
  But there is the opportunity cost of missing real income when you stay out of the work force.

cia_plant 18y ago

The obvious answer is that a good science student can make 6 figures today as a programmer or in finance, or they can go to grad school and work harder for much less money. My own reasons are actually different. I'm not particularly interested in money, and I love to teach and to do research. The academic environment just seemed stifling to me. As someone who values creative freedom enormously, the phrase "publish or perish" is very frightening.

asciilifeform 18y ago
Where, if at all, did you finally find creative freedom?
- cia_plant 18y ago
  I haven't really found a good way to support myself while doing what I want, yet. Right now I live frugally and work part-time as a programmer. One reason that I hang out at news.yc is that I'm hoping to turn some of my ideas into some sort of startup, even though I don't really fit the profile of a startup founder (I'm not very driven or passionate, at least not when it comes to business.)

gamble 18y ago

"...there are far more post-docs working at universities than there are permanent jobs for them to move into, so a lot of people get stuck in limbo." The author answers her own question. Since the era of expansion in higher education and private research labs driven by the baby boom and cold war came to an end in the 1980's, the scientific job market hasn't been matched to the number of qualified applicants. If anything, there should be fewer grad students. For all the rhetoric about higher education as a bastion of left-wing politics, academics put up with an employment system that would make Ayn Rand smile. Aside from an ever-shrinking pool of tenured professors, most academics have low salaries, high workloads, little-to-no job security, and few employment prospects outside academia.

davi 18y ago
Yup. I came up with some back of the envelope numbers around this here: http://news.ycombinator.com/item?id=470181 http://news.ycombinator.com/item?id=470181
- gamble 18y ago
  Interesting analysis. I suspect it's only going to get worse in coming years, with the economic crisis taking a big chunk out of endowments and professors' retirement savings alike.

dkarl 18y ago

I dropped out of a top-ten math PhD program because I looked around and thought, "Most of these guys aren't going to get jobs, and I don't feel like one of the exceptions. I'm not going to spend five years in graduate school to end up a failure." Of course, by "job" I meant a tenure-track research position in a livable city, and by "failure" I meant... well, basically, by "failure" I meant my father, who taught at a small, bad public university in a rural wasteland. I'm pretty sure my mom hated him for every hour and minute she had to live in that town. If I picked up a wife or girlfriend in the hip, thriving city where I lived, would she follow me to Bumfuck, USA? Would I let her if she wanted to? I constantly obsessed over the fear of ending up teaching hopeless kids at a double-directional school in a little town at the end of the universe, watching my kids grow up alienated from their provincial peers and underserved by the school system, burdened with the knowledge that it was my failure that forced my family to live there. So I quit and got a job as a programmer to keep myself afloat while I figured out what to do with my life. It turns out there are lots of challenges in computing, and I've worked alongside a couple of math PhDs on interesting problems. I think grad school would have been a lot more fun if I had realized that failing to get a hot tenure-track position wouldn't ruin my entire life.

giardini 18y ago

Many U.S. citizen know that it's not worthwhile to get a Ph.D. and in some instances, a Master's degree. See Philip Greenspun's "Career Guide for Engineers and Computer Scientists" at http://philip.greenspun.com/careers/ http://philip.greenspun.com/careers/

Dilpil 18y ago

"Being a scientist is still a comparatively low-stress, high-pay career" Compared to jobs done by those without college educations, sure. Compared to the many career options open to anyone who is in the position to become a scientist, not at all. The author, like Obama and all the others demanding that (other) young people become scientists and engineers, are living in a fantasy world where a) science is an amazing career path an b) there is a shortage of scientists and engineers.

varjag 18y ago

"Learning physics is actually not so different a process as learning a foreign language." Uh, no.

asciilifeform 18y ago

"Adjusted for IQ, quantitative skills, and working hours, jobs in science are the lowest paid in the United States." (http://philip.greenspun.com/careers/women-in-science http://philip.greenspun.com/careers/women-in-science) Despite agreeing with everything in the article, I am still devoting my life to somehow getting into graduate school - simply because my life goal is to eventually spend all of my prime waking hours working on beautiful, elegant, entirely non-commercial concepts, and there is still a non-zero number of academic positions which allow this, in contrast with industry. Like choosing gladiatorial combat in place of meekly kneeling for beheading, it is an issue of having a sporting chance vs. no chance at all.