McArdle, an economist, recently wrote that a substantial fraction of PhD programs really shouldn’t exist. McGowan goes further by pointing out that there is little demand for science and technology PhDs. I can only agree. Meanwhile, the gap between tenure-track positions and the number of new PhDs keeps on increasing. In countries like Japan, we are facing a crisis:

Of the 1,350 people awarded doctorates in natural sciences in 2010, just over half (746) had full-time posts lined up by the time they graduated. But only 162 were in [academia], of the rest, 250 took industry positions, 256 went into education and 38 got government jobs.

In the US, things are getting worse and worse:

In 1973, 55% of US doctorates in the biological sciences secured tenure-track positions within six years of completing their PhDs, and only 2% were in a postdoc or other untenured academic position. By 2006, only 15% were in tenured positions six years after graduating (…)

Even in a field like computer science, 55% all new PhDs go into industry, rarely landing a research position while less than 7% end up with a tenure-track position in a doctoral-granting school.
Yet between 2000 and 2010, in the US, the number of STEM graduate students grew by 30%. Why?

McGowan writes:

It is not surprising that many STEM students believe there is a strong industry appetite for freshly minted Ph.D.’s given these numerous advertising and public relations campaigns by many businesses.

McGowan provided little support for this assertion. I went looking and found nothing in the last 10 years. In fact, I ended up reading several articles warning students against graduate work…

I am left with something of a mystery… why are we producing more and more PhD students?

My thoughts:

  • To a student, the PhD job market is an abstract entity. The financial incentives provided by the government (e.g., scholarships) are not.
  • Even though the job prospects of a science PhD might be bad, maybe the alternatives look even worse. Yes, you can make a good living as an engineer or an accountant in industry… but maybe students imagine that the life of a PhD, no matter how difficult, is somehow more exciting?
  • It might be all about status signalling. A science PhD might be a way to tell the world that you are smart. Even if employers are unimpressed, maybe your friends or prospective mates could be impressed?
  • Schools and professors are misleading students. If your local dean says that there is a shortage of science PhD, are you going to check the data for yourself?

What are your thoughts?

75 Comments

  1. I hope, for my sake, that at least 50% of what is written here is false. Though, I have my doubts. :S

    Comment by Michele Filannino — 7/1/2014 @ 14:17

  2. Many data scientists start out with PhDs in science, and find happy employment in industry. Academia isn’t the only path.

    Comment by Hilary Mason — 7/1/2014 @ 14:19

  3. @Michele

    How did you perceive the job market where you started your PhD? How do you perceive it now?

    Comment by Daniel Lemire — 7/1/2014 @ 14:20

  4. @Hilary

    I encourage you to read McGowan’s article. His entire argument is about PhDs in industry.

    Comment by Daniel Lemire — 7/1/2014 @ 14:21

  5. @Daniel

    I’m still in the middle of my PhD so I can’t say it for sure. My overall impression is that my employment opportunity have grown a lot in terms of number offers. In terms of quality I really don’t know.

    I’m confident about my future. Although, even a PhD doesn’t give me the certainty about where I will work after.

    Comment by Michele Filannino — 7/1/2014 @ 14:32

  6. I get very upset when people talk about academics and higher education in terms of employment and job prospects. This has been a sick side-effect of the industrialization of the academia after WW2. I would hope deep down that there are some who still go into a PhD for the “right reason”, which you seem to strangely ignore: a love for higher learning. Learning for the sake of learning, not for the sake of jobs or status.

    But if you must search for economic reasons: I hear that getting a western PhD (or easier, a Masters) is an easy route towards permanent residence and citizenship in Canada/US. Is the 30% growth between 2000 and 2010 atypical given economic (number that can afford) and demographic (number of right age) trends in China, India, and other big foreign student producers?

    Comment by Artem Kaznatcheev — 7/1/2014 @ 14:51

  7. It would be interesting to compare the ratio of foreign-born to native-born PhD students. For a bright college graduate from a developing country, a PhD studentship in the US, UK or Canada is a way to get an immigration toehold and life experience in a country where he/she might like to immigrate.

    This obviously doesn’t apply to native-born students, so they will do advanced degrees for other reasons.

    Comment by Jack — 7/1/2014 @ 15:04

  8. This may sound a little idealistic, but perhaps students enter into Ph.D. programmes mainly because they are interested in their subject, want to study it further, and would like to contribute to human knowledge. That is certainly why I did my Ph.D. and if you had told me beforehand that the job market was bad then it would not have swayed me.

    The problem, if there is one, is not with Ph.D. programmes themselves, but rather with those who encourage people to do such programmes for any reason other than for the sake of the Ph.D. itself.

    Comment by Matt Leifer — 7/1/2014 @ 15:13

  9. @Hilary True, but making the transition is not necessarily easy. Even with a background in Physics and CS and doing research that would fit like a glove in most definitions of “Data Science”.

    Comment by Bruno Goncalves — 7/1/2014 @ 15:21

  10. part of it is simply due to inertia of already being in an academic environment. students commit 4 years to do their undergraduate degrees. instead of pursuing the unknown (ie. searching for a job, moving potentially, becoming more independent, etc.) students stick with what they already know–academia, albeit in a different, more serious form.

    Comment by 7up — 7/1/2014 @ 15:35

  11. It’s simple: the immigration system rewards advanced degrees.

    While Americans often pursue Ph.Ds because of a desire to pursue academia or pure research, most people coming to the US for educations are looking for a way to stay in the US, and there are H1-B slots specifically for those with advanced degrees from US universities.

    Further, the life of a graduate student on stipend is remarkably similar in lifestyle comforts to working in their native countries, and they will be rewarded with better jobs in their native countries even if they do not stay in the US.

    That is what fuels the graduate student STEM system.

    Comment by Eric — 7/1/2014 @ 15:42

  12. “…are you going to check the data for yourself?”

    Of all the people in the world, I would hope members of the group “science PhD applicants” would be among the most likely to answer “Yes!” to this question.

    Comment by Bill the Lizard — 7/1/2014 @ 15:43

  13. Better question: why are there so many news outlets willing to pay a hack like McArdle to write nonsense that confirms their biases?

    Comment by Matt — 7/1/2014 @ 15:49

  14. Your premise seems to be (to a certain extent at least): People enroll in PhD programs for the sole purpose of furthering their careers, whether they want to stay in academia or not. Why not pursue a PhD for the experience itself? Research in academia is a different world that offers exciting possibilities to explore new fields, travel, and meet interesting people around the world. Or just the sake of curiosity and the promise of creating something novel.
    (Disclosure: I will be starting my PhD in a month, and these are my motives, maybe I’m a bit idealistic)

    Comment by Jurgen — 7/1/2014 @ 15:58

  15. If you think those with science PhDs are in bad shape, consider the even worse prospects of graduate students in the humanities. It strikes me that we ought to be orienting much of our research training — much, not all — towards individuals who intend to take those skills and create products and services to support the private and social sectors. That, to my mind, makes much more sense than orienting them to the academy, where most have precious little chance of landing a tenure track appointment, and stand an excellent chance of landing themselves in supply teacher hell.

    Comment by John — 7/1/2014 @ 16:09

  16. As a recent STEM grad, I think the biggest part of the issue isn’t the status or potential for job opportunities in industry. Rather, I think the biggest culprit are the schools themselves. They benefit the most from having enrollment increase in both undergraduate and graduate routes. Many of students in undergrad are fooled into thinking that it’s not just a favorable way to continue their education — we’re told that it will be the only way we get jobs.

    In my last two semesters as an undergraduate student, my university would send me letters advertising how graduate degrees were sought after and how amazing of an experience it would be. My professors even went as far as encouraging everyone (even the students that struggled) to apply and go to grad school.

    Comment by Sol — 7/1/2014 @ 16:16

  17. @Jurgen @Artem @Matt

    The PhD degree does nothing for your “knowledge”, it is only useful for status and jobs. Simply put, anyone who goes through the motion and gets a PhD does so for less idealistic motives.

    There is an easy test: if anyone tells me that they are pursuing the PhD for “knowledge” and not “jobs”, then let them agree to do all the work, but stop short of filling with the schools in the thesis at the end…

    Wait? Nobody would do that? Right. Exactly. People will insist of getting the degree.

    Comment by Daniel Lemire — 7/1/2014 @ 16:16

  18. The job market for PhDs did not enter into my decision to do one. I love the intellectual freedom of academia, and I want to use my skills to improve people’s lives. I chose a PhD for those reasons, and 2 years in, I’m very glad I did. I’m certainly not going to struggle to get a job afterwards, and I had very deep job satisfaction at the moment. Win all round.

    Comment by Richard Smith — 7/1/2014 @ 16:22

  19. “…it is only useful for status and jobs” – I find this statement kind of sad. At least in Europe no one enters a PhD for the prospect of better jobs, status yes (probably even more so than in the US).

    I would argue, though, that it is also for the experience. Academia is a different world than industry: Conferences, Scientific writing and presentation, Visiting Research, etc. and of course all the contras as well: Grant applications, paper deadlines, reviews

    I would not leave this aspect out of the equation and simplify it to “status and jobs”

    Comment by Jurgen — 7/1/2014 @ 16:25

  20. @Daniel

    Yes because getting the degree (at the end!) gives you a lot more opportunities to continue in research (to learn and create “knowledge”).

    In this sense, a phd is for getting “phd jobs” even for the most idealist person, but not necessarily for getting jobs with higher salaries.

    Comment by Gokberk — 7/1/2014 @ 16:33

  21. @Gokberk

    Yes because getting the degree (at the end!) gives you a lot more opportunities to continue in research. In this sense, a phd is for getting “phd jobs” even for the most idealist person, but not necessarily for getting jobs with higher salaries.

    I do not think I ever alluded to the possibility that people pursued a PhD for better salaries.

    Simply put, people want to get a PhD to get desirable jobs.

    A desirable job is not necessarily a job with high pay… For example, being a novelist is a dream job for many… but not because of the income…

    Comment by Daniel Lemire — 7/1/2014 @ 16:37

  22. @Jurgen

    I find this statement kind of sad. At least in Europe no one enters a PhD for the prospect of better jobs, status yes.

    So it is sad to pursue a PhD for the prospect of better jobs but not for status?

    Comment by Daniel Lemire — 7/1/2014 @ 16:41

  23. No, it is sad to to ONLY see these two reasons. If you read further, I’d consider the experience you gain whilst pursuing a PhD . Others have also named “contribution to human knowledge”, which I think you have disregarded in your first reply.

    But I agree that pursuing a PhD only for status is even sadder and unfortunately more often the case that I’d like to admit (at least in central Europe)

    Comment by Jurgen — 7/1/2014 @ 16:45

  24. @Jurgen

    Others have also named “contribution to human knowledge”, which I think you have disregarded in your first reply.

    Yes, people want to have research jobs. I acknowledge this.

    Comment by Daniel Lemire — 7/1/2014 @ 16:53

  25. @Daniel

    Higher salary was just an example. Seeking a phd for better public status or pride is not essentially different for this context in my view.

    My point is that a phd is statistically a good investment for anyone who truly wants to do (public) research, even though it is not necessary nor sufficient.

    Comment by Gokberk — 7/1/2014 @ 16:55

  26. @Gokberk

    My point is that a phd is statistically a good investment

    Is it? What is your definition of a “good investment”? Please see the following charts:

    http://www.theatlantic.com/business/archive/2013/02/the-phd-bust-americas-awful-market-for-young-scientists-in-7-charts/273339/

    Comment by Daniel Lemire — 7/1/2014 @ 16:59

  27. Science is fun.

    There will always be far more people trying to be researchers than anyone is willing to feed. Getting a PhD is the easiest chance to do research while getting paid for it, so that’s where people go.

    Some people are content with boring, meaningless, productive, relatively stable jobs, while others are willing to take risks for a chance of getting paid for doing something they enjoy. Some might try to earn their living by playing guitar, while others may do research, travel around the world, or practice sports.

    Comment by Jouni — 7/1/2014 @ 17:01

  28. @Daniel

    Good investment.. for “anyone who truly wants to do (public) research”

    I couldn’t see a chart comparing “research job” opportunities with or without a phd?

    Comment by Gokberk — 7/1/2014 @ 17:05

  29. Are you a PhD?

    Comment by Sun — 7/1/2014 @ 17:08

  30. @Gokberk

    Good investment… for “anyone who truly wants to do (public) research”

    What is your criteria with respect to the job market to say a phd is statistically a good investment? How bad should the job market be for you to change your assessment?

    Comment by Daniel Lemire — 7/1/2014 @ 17:17

  31. @Sun

    Yes. I have a PhD, a tenured (full) professorship, a higher-middle-class salary and a reasonably successful career as a researcher.

    Comment by Daniel Lemire — 7/1/2014 @ 17:21

  32. @Daniel

    I would change my assessment only if there is data proving
    p(GetAResearchJob|WithAPhD)<=p(GetAResearchJob|WithoutAPhD)!

    Of course, this exaggeration is just the for the sake of my argument. Otherwise it is disappointing to see that job market is not large enough for the phds.

    Comment by Gokberk — 7/1/2014 @ 17:35

  33. I confess I also started my PhD on the “idealistic” premise that I wanted to continue studying and contribute to human knowledge and development.

    Being in my final year, that idealistic view seems very naive in retrospect and I am certain I let myself get deceived by this benevolent picture the greater public seems to have of scientific research.
    Once you’re inside the research community the level
    of politics and general hand waving seems nauseating
    at first.

    Personal rant aside, at least in the UK research
    institutes get many of their PhD students’ stipends
    paid for by a number of research councils.
    When said students graduate within a certain number
    of years the degree-granting institute will increase their
    chances of receiving an equal level or higher
    level of funding for PhD students.

    When it comes to publicly funded PhD’s there seems
    to be some level of government funding “use it or
    lose it” sort of stuff going on.

    Comment by Georg — 7/1/2014 @ 18:03

  34. For me it is quite simple. I have not met a single PhD (in my field) with troubles to find a job, quite on the contrary. All of my ex-colleagues seem happily employed. Thus I have no problem welcoming new students to do a PhD. It seems hard to find “the job of your dreams”, but finding _a_ job seems quite doable (e.g. banks do welcome phds).

    That being said, I most certainly suffer from sampling bias (i.e. where could I meet the unemployed PhDs ?).

    Comment by rodrigob — 7/1/2014 @ 18:15

  35. @rodrigob

    finding _a_ job seems quite doable (…) That being said, I most certainly suffer from sampling bias (i.e. where could I meet the unemployed PhDs ?).

    There was no mention of unemployment in this post or thread. What we are discussing are difficult job markets. In this respect, the signs are everywhere… go to any campus, you will find plenty of PhDs who could not secure a job and have to settle for postdocs… even the very brightest. This is even true today in fields where it was unthinkable before such as computer science.

    Comment by Daniel Lemire — 7/1/2014 @ 19:07

  36. Fortunately, it still makes sense to do PhD in Computer Science. Computer Science, as we know full well, is neither about computers nor about science :-) So, computer science PhD students are doing just fine.

    PS: backup people who say that broken immigration law creates more PhD students than it’s perhaps necessary.

    Comment by Leonid Boytsov — 7/1/2014 @ 19:42

  37. PS: I also think that we have more PhDs because universities want to keep grant and tuition money abundant.

    Comment by Leonid Boytsov — 7/1/2014 @ 19:44

  38. Professors need sweatshop labor. There is really nothing stopping them in building an empire, all they have to do is raise a paltry amount of money, and then dangle the prospects of “You can get a sexy research job like mine” in front of an idealistic 22-year old and … bingo! new sweatshop worker !!!

    Comment by systemBuilder — 8/1/2014 @ 6:13

  39. In my graduate school, a top-10 school in computer science, I studied theory and I was the ONLY (non-jewish) american among 50 PhD students. The only one.

    Comment by systemBuilder — 8/1/2014 @ 6:17

  40. I usually find your posts very interesting and mind bending but I blanked out after the first two words of this one. If the context is brought from that person then I don’t bother. Call me lazy but I hate spending time disproving flawed assumptions.

    Comment by Igor — 8/1/2014 @ 6:50

  41. From what I can see, it looks like there has been a commodification of advanced degrees by the universities, and this seems to be related to the amount of money that they get from governments for creating all these doctors.

    I would also like to point out that at the present economic situation is bad for young people (and older ones as well) who are looking for a job.

    So maybe there is a combination of universities pushing and young people looking at the lack of possibilities and thinking that an advanced degree will get them something useful at the end.

    In any case, I have started telling young people not to do a Ph.D. unless they were passionate about it. They usually look at me like I sprouted a second head.

    Comment by Sylvie Noel — 8/1/2014 @ 8:26

  42. @Sylvie

    In any case, I have started telling young people not to do a Ph.D. unless they were passionate about it. They usually look at me like I sprouted a second head.

    It is easy to be passionate about landing a tenured professorship in a great school. It is easy to dismiss the consequences of the loss of many years on your lifetime income… who cares about your lifetime income if you are to secure a great job later? Who cares about delaying your entire life (retirement, house purchase, kids, promotions…) by 5 years when great things await you?

    Comment by Daniel Lemire — 8/1/2014 @ 8:50

  43. useless phds have low income prospects… if you have usefyll phd i think industry aka google fb … will appreciate it. thi is just one piece of education bubble fueled by governments.

    Comment by ivan — 8/1/2014 @ 9:02

  44. I really don’t see the problem. A PhD usually gets you an interesting high-paying job even if it is not in academia. See official statistics: http://www.bls.gov/emp/ep_chart_001.htm

    It seems utterly reasonable that in steady-state only 10-20% of PhD’s land tenure-track academic positions, while others end up in industry, government, education, management, or many other jobs that do benefit from the type of independent research that one learns how to do in grad school.

    This is good; not bad. It will be a very bad sign if the only thing that a new PhD can do is to replace his/her advisor and churn out more PhDs.

    Comment by noam — 8/1/2014 @ 9:07

  45. @Sylvie @Daniel

    This hits the nail on the head: I think most people will agree that working as a tenured professor at a top school is a desirable job in many ways.

    Part of the problem is probably that incoming PhD students are not aware of the length (and high quality) of the queue of candidates for those desirable positions.

    I wonder though how universities were impacted if, say, enrollment for PhD programs dropped significantly.

    Comment by Georg — 8/1/2014 @ 9:55

  46. @noam

    It seems utterly reasonable that in steady-state only 10-20% of PhD’s land tenure-track academic positions, while others end up in industry, government, education, management, or many other jobs

    How would you feel if you were told that you can no longer occupy an academic post. Would you feel like it is reasonable? Or would you feel like you are throwing away an expertise that you acquired at a high cost?

    A PhD usually gets you an interesting high-paying job (…)

    Does the PhD get you the job?

    If I were to erase your PhD from your resume and throw you on the streets, I am confident that I would find you, two years later, earning a solid salary. What is cause and effect?

    I don’t think that participating in the Olympics is what causes you to have an athletic body… I think that having an athletic bodies is a condition to enter the Olympics.

    Similarly, I don’t think that PhD programs causes people to have ambition and intelligence… I think that people who enter PhD programs tend to be more ambitious and intelligent.

    Comment by Daniel Lemire — 8/1/2014 @ 11:03

  47. @daniel

    “How would you feel if you were told that you can no longer occupy an academic post. Would you feel like it is reasonable? Or would you feel like you are throwing away an expertise that you acquired at a high cost?”

    Even though academic posts may be preferable to others (for many, including myself, but not for all), I certainly never felt that getting some other kind of job would somehow “throw away my expertise” or imply that my PhD was wasted in any way. Industry jobs for PhDs are usually more interesting than those for non-PhDs; high school teachers with a PhD are usually better than the average teacher, etc.

    Also, I hardly consider my PhD to have been acquired “at high cost” — quite the contrary: it was one of the best periods of my life. I would certainly recommend graduate school even if you are 100% guaranteed never to get an academic post.

    Comment by noam — 8/1/2014 @ 11:59

  48. Industry jobs for PhDs are usually more interesting than those for non-PhDs

    Ah. But where is the evidence for that? That’s McGowan’s point: there is little evidence of any desire from industry to recruit PhDs.

    Even though academic posts may be preferable to others (for many, including myself, but not for all), I certainly never felt that getting some other kind of job would somehow “throw away my expertise” or imply that my PhD was wasted in any way.

    So you are telling us that you would be upset to lose your academic post? Having to apply for a software engineering position, say? As you’d be passing Java programming tests in HR departments, you would not feel in the least that your expertise writing research papers is being neglected?

    Comment by Daniel Lemire — 8/1/2014 @ 12:17

  49. As my undergrad advisor, Robert Fano said, the good jobs in Academia are at the top-20 schools. If we assume that each faculty has 45 posts, that’s only 900 positions. The average academic career ends in death (the academic retires only 2 years before death), so if we assume that the lifespan is 77, then we get an expected career length of 77 – 2 – 6 – 22 = 37 years. Therefore, there are 24 jobs at these top-20 schools available, every year. And in my field, roughly 1000 graduate. Your chances are 2.4%. Far far more beneficial to buy a lottery ticket, and those are A TAX ON STUPID.

    P.S. I’m one of the lucky few who DID get a job in academia, but I COULD NOT afford to keep the job in academia, at a top-2 school (equiv to top-20% in USA) in Canada.

    Comment by systemBuilder — 8/1/2014 @ 13:10

  50. Whoops, 47 year career, i.e. 19 jobs per year, so there is less than 1 job per year per school, in the top-20 schools.

    Comment by systemBuilder — 8/1/2014 @ 13:12

  51. @Daniel Lemire

    So what’s your purpose of getting a PhD? A Job for Money? That’s what this blog implies according to my understanding.

    I am not saying it is an improper purpose, in fact, it is a very good motivation to get a PhD for today’s CS industry.

    At least for CS, the PhD research experience is needed to do challenge work in industry. Companies like Google and Facebook are facing with large scale computing/data problems that academic people could not imagine before. It is not only solving the problems but also developing software with good understanding of the problem requires a person with CS research experience. A master or bachelor with well-trained coding skills can’t even understand the origin and context of the problem without spending long time to study the basic knowledge.

    Not only those problems faced by Google and Facebook, there are more and more challenging problems coming out with the rapidly increasing computing requirements, like utilizing heterogeneous architecture, statically analyzing program bugs, writing better graphics/video game engines … Yes, there are still many basic jobs like writing webpages, composing mature libraries to an app, writing boring tests, doing routing fixes and features. But they are not for CS PhD. CS PhDs in industry may also do such work, but with more challenging background problems, just like firstly comping up an idea/solution and then writing boring code to demonstrate it as a PhD student. The essence of a PhD, either in academia or in industry, is in the first problem solving stage, though the second realization stage is inevitable.

    Comment by Anonymous — 8/1/2014 @ 13:18

  52. I get even a little bit angry when I read texts like this that are based on the assumption that each human being is a cold-blooded profit maximizer. I am a physics PhD student and I know that the choice of going to PhD school probably doesn’t maximize my future income. I even know that probably I’m not able to get a permanent job in the academia. But you know what? I don’t give a damn! I’m in graduate school because I like learning physics, I like the atmosphere of an academic workplace and to be honest, I also like the social status that being a science phD gives me. Even if I will never work on anything physics related again, I will still not consider this time wasted, because I got to learn and think about the mysteries of the nature that most people will never know anything about.

    After I finish graduate school, I might stay in the academia or I might do something entirely different like find a company, try to become an author or move to Africa to do volunteer work. None of these things are rational for a cold-blooded profit maximizer, but they make sense to me, since I get to experience interesting things, do things that I enjoy doing and, yes, gain some social status associated with these professions.

    Being born in the Western world in the end of the 20th century, I already consider myself rich enough so that I don’t have to plan my life around maximizing my income (most if which I won’t use anyway because I naturally tend to live frugally). Hell, if everything else fails, I can live on social welfare for a while or move to my parents house. I wouldn’t even consider either of these things a major setback since my life goals and aspirations are completely unrelated to the state of my bank account.

    Comment by Anonymous — 8/1/2014 @ 13:31

  53. @Daniel Lemire

    “Ah. But where is the evidence for that? That’s McGowan’s point: there is little evidence of any desire from industry to recruit PhDs.”

    Take a look at papers in many top CS conferences, many of them are published by industry people to describing the challenging problems solved in industry. There are also many papers that are firstly published by academia, then become commercial products that need PhD level developers to really understand and improve it with their problem-solving skills.
    I am in system research, some examples: GFS, MapReduce, Xen, Dynamo, Facebook’s memcached improvement, MS Azure storage, Google F1, Spanner, EMC’s and NetApp’s many storage papers, VMWare’s VM migration …

    There are also many others, which are not published as papers at all, in the industry. I know some companies such as (I’d not say their names) some GPU companies, SoC companies, router/switch companies, they’ve done many challenging work “silently” without revealing the principles and details.

    Comment by Sun — 8/1/2014 @ 13:46

  54. @Sun

    Of course. There is a lot of R&D going on in industry. And it can be more fun than academic research. Certainly, it can be more relevant.

    Take a look at papers in many top CS conferences, many of them are published by industry people to describing the challenging problems solved in industry.

    Comment by Daniel Lemire — 8/1/2014 @ 14:09

  55. Hey anonymous! This is a huge-huge mistake. There are a lot of masters/bachelors who can do research. And in fact many masters got a formal training in research (it is just shorter than that of a PhD). There are really a lot of clever people without a good formal education, which are nevertheless are very productive and research-oriented.

    Comment by Leonid Boytsov — 8/1/2014 @ 14:21

  56. It is an economic waste to train for a profession for 5-7 years (research or academia in XYZ), and then fail to find employment in that profession. Canada does it best, they over-educate their engineers so thoroughly (26% of adults over 25 have college degrees) that only about half of bachelor-degreed engineers EVER work in their field in of study, in Canada, even at the BS level. See http://www.canada.com/Education+statistics+glance/8580700/story.html

    It is far more productive from a societal perspective to just let those people pursue careers with BS degrees. Also, about 90% of what gets published in academia is not an “advance” in a field per se, it’s simply “practice papers” that the thesis advisor justifies in the name of “research education” for the particular students involved. And this is wasteful and pollutes the academic research stream.

    Comment by systemBuilder — 8/1/2014 @ 14:56

  57. So when I scrapped my original plan to do my diploma and then stay at university right about after studying for around 2 years (but still finishing what amounts to be a Master’s in CS) and am a happily employed person – how do I fit into this picture then?
    Is it good that I didn’t go for a PhD? I would’ve done it for the learning experience but got a little annoyed by pure theory, so maybe it’s for the best :)

    Comment by fa — 8/1/2014 @ 15:20

  58. @daniel: in response to my claim that “Industry jobs for PhDs are usually more interesting than those for non-PhDs” you ask “Ah. But where is the evidence for that?”.

    I thought that this was clear to anyone in our field, but here’s the “official” point of view of the US bureau of labor statistics. Just compare their description of “What Computer and Information Research Scientists Do” (a job that they say requires a PhD) at http://www.bls.gov/ooh/Computer-and-Information-Technology/Computer-and-information-research-scientists.htm#tab-2 to their description of “What Software Developers Do” (a job that they say requires a BSc) at http://www.bls.gov/ooh/computer-and-information-technology/software-developers.htm#tab-2 . These descriptions point out very clear differences in the type of work, differences that seem to be generally acknowledged.

    Comment by noam — 8/1/2014 @ 15:35

  59. Noam, I back you up. I also got (though based on anecdotal evidence) an impression that, e.g., big companies separate jobs for PhDs and non-PhDs. At least in computer science. The problem is: these PhD specific jobs are rather hard to get.

    Comment by Leonid Boytsov — 8/1/2014 @ 15:58

  60. @noam

    Let us take a top 10 employer for computer science graduates: Google. Out of the 55% of all PhDs who go to industry, I would say that some of the “luckiest” and best are at Google.

    Google does recruit people directly at the PhD level. Here is a typical job:

    https://www.google.com/about/jobs/search/#!t=jo&jid=3239002&

    Look under “responsibilities”. How does it differ from any other engineering job (without a PhD)? The answer is that it does not.

    Effectively, at least judging by their job ads, there is no difference when you enter Google with a PhD, or not. You do the same things.

    Yes, Google has a few star researchers (http://research.google.com/researchers.html). About 650 worldwide… compared to over 20k engineers.

    My point is that the vast majority of CS PhD graduates who go into industry have the same jobs as BSc graduates. In fact, he may very well be working for his friend, a BSc graduate, who decided to forgo the PhD.

    Yes, there is a lot of R&D in industry, but a PhD is not required to do R&D. And since you are unlikely to end up with a job that matches your PhD topic, it is not clear that your PhD experience helps a lot compared to, say, 5 years of industry experience.

    My point is not that people who work in industry with PhDs are miserable or unlucky. In fact, they often have better jobs than people who land academic jobs… a point I have made repeatedly on this blog… But the PhD does not open lots of doors in industry.

    Comment by Daniel Lemire — 8/1/2014 @ 16:16

  61. @noam

    About comment #60, I checked with a Googler (Philippe Beaudoin) who has a PhD. He says: “This is absolutely true. You can quote me, I say as much whenever I’m invited to give a recruitement talk to PhD candidates. (I mean, don’t quote the numbers, I don’t know exactly how many we have in each group.)”

    I am sure that difference people within Google or elsewhere in industry would have different views… but there is truth in what I write.

    Again. I should stress that I don’t think Philippe is unhappy. He seems to absolutely love his job. He does not appear to regret academia. I don’t think I have a better job than he does, and I am pretty sure he does not envy me. He probably makes more money than I do. He works on cool stuff.

    But the whole idea that if you have a PhD and you go work in industry (e.g., for Google) you are going to be a scientist and be apart from the guys (and gals) you merely have a BSc… I don’t think that there is a lot of truth in that.

    Yes, there are research labs in industry where people are paid to do research full time and it requires a PhD… Microsoft Research comes to mind… but that’s not the typical industry job for someone with a PhD.

    Comment by Daniel Lemire — 8/1/2014 @ 19:42

  62. A PhD does not open many doors in industries because the USA gutted its industrial research complex in 1990. Really and truly, gutted it, forever. There are very, very few pure research jobs in industry any more, the best you can find is an “advanced development” job now.

    The PhD does not open doors in industry, but a degree from a top-10 school certainly helps a great deal.

    I have worked at a Canadian university, as a consultant, at Qualcomm Research , and now, at Google in production. I work next to PhDs from top-10 schools (and top-3 schools from abroad). And also, some BS and MS people, too, but not very many.

    Comment by systemBuilder — 8/1/2014 @ 21:01

  63. I think that Google is an exception in that many of its software engineers have PhDs. These are obviously some of the most desired software development positions in the world, and a PhD does help you land such a position, mostly because of what you’ve (hopefully) learned during your PhD. Of course there are many other routes to these jobs (say, found a start-up), but for many people the PhD route is safer, more flexible, “fun-er”, faster, and provides more possibilities.

    Comment by noam — 9/1/2014 @ 1:31

  64. This has been an interesting discussion. I could echo the comments of other about the harder to quantify benefits of getting a Phd, but there are two other points I wanted to make. Mostly in relation to the idea of people with Phds taking jobs that do not require a Phd.

    1) Not every CS Phd with a job in software development is there because they could not get a research position. I have seen many people get burned out by the research process—it can be very political, and a series of poor results can make your work feel meaningless. After graduation they take a development job because they have learned that is what they prefer.

    2) Just because a job does not require a Phd, does not mean a person could have obtained that job without a Phd. If, like me, you did your undergraduate at a no-name school, top companies are probably not interested in hiring you. While my job prospects after my BSc were not bad, the work did not seem particularly interesting. Now, after getting a Phd, I have a reasonable chance at getting a research position at a company that would not have even considered me for an entry-level development position 6 years ago.

    Comment by William Hartmann — 10/1/2014 @ 3:48

  65. @William

    Not every CS Phd with a job in software development is there because they could not get a research position. (…) After graduation they take a development job because they have learned that is what they prefer.

    Of course. I personally like to code more than I like to do research.

    I think that the PhDs that go to industry and become programmers are probably having more fun, earning at least as much money, enjoying great job security…

    I certainly do not discourage people from taking programming jobs in industry.

    While my job prospects after my BSc were not bad, the work did not seem particularly interesting. Now, after getting a Phd, I have a reasonable chance at getting a research position at a company that would not have even considered me for an entry-level development position 6 years ago.

    It is not at all a fair comparison. You want to compare yourself right now with an alternate version of you who went to do other things… in 6 years in industry… you could have written a book… you could have ran a successful blog… you could have launched a big open source project… you could have started a company or two… you could have been promoted 2 times…

    I don’t know what your research expertise is… but you could find that if you try to go to industry, they may not be interested. This is the point that McGowan makes. If you are lucky and whatever you worked on happens to be red hot at the moment… then good for you… but if you are just another PhD who did regular work that is not talked about in the New York Times… well…

    Comment by Daniel Lemire — 10/1/2014 @ 9:57

  66. I am currently a prof at a well-known research university, and I can vouch that I would have loved my PhD experience even if I hadn’t gotten the (fantastic) job I have now.

    My PhD years are some of the best years of my life because of the flexibility that grad school offered. I was able to travel around the world, read and debate and date plentifully, enjoy 3-day weekends and 6-week vacations (while working superhard the rest of the time), take classes on languages and film in addition to computer science, all while doing world-class work on my own project (which, by the way, was exactly what I wanted to do). Socially, academe was unparalleled, much better suited to someone of a broadly intellectual bent than the corporate sector.

    Even if the job I got after my PhD were the same as what I’d get with a BS, it wouldn’t matter an iota. I’d still have enjoyed 6 years of carefree, intellectual, bohemian life. Fortunately, I ended up with a great academic job, which guarantees a grad-school-lite life with excellent job security for the next n decades.

    Comment by CS professor — 11/1/2014 @ 0:37

  67. @CS professor

    Socially, academe was unparalleled, much better suited to someone of a broadly intellectual bent than the corporate sector.

    Though I don’t outright disagree…. Your statement assumes that whoever does not follow the PhD is going to focus his life around a monotonous corporate job. They are not going to travel, read and write books, start new exciting projects…

    You are also discarding the loss of income… which can translate into accrued freedom. For example, someone who starts saving for retirement 5 (or 9) years sooner can, logically, retire 5 (or 9) years sooner. And even if you love your job, the extra lifetime income can be used to take extra vacations, attend lots of fancy conferences or pay for a much larger house, better cars…

    Comment by Daniel Lemire — 11/1/2014 @ 15:17

  68. @CS Professor,

    It’s quite common for “the 2%” who win “the academic lottery” (and it really is a lottery if you only knew what I knew) to talk about their fabulous lives as if nothing could be better. Well, nothing can be better when you’ve won a lottery, as long as you don’t squander your winnings.

    But the sad fact is, social networks count as much as research achievements. The sad fact is, fads count more than facts in academia. I can support this with my experiences, both as an “out of fashion” academic, and also later when my field came into fashion, the way it generated instant interviews at top-10 research universities in America. I didn’t change. Fads changed.

    @Daniel Lemire

    Starting saving 5 (or 9) years sooner does NOT allow you to retire 5 (or 9) years sooner, because your lifespan is more or less fixed by genetics, and so you need a heck of a lot more money saved if you will retire 5 (or 9) years sooner.

    On the flip side, by pursuing a PhD you are compressing your career into a shorter number of years. So if you spend 15% of your working years doing a PhD (which is the average), you will need WAY MORE than 15% more income during the remaining years to catch up to yourself. The problem is, not only do your savings & investments have less time to grow, but your tax rates are way higher if you shorten your career in return for a higher salary.

    As a canadian professor in my first year of work in British Columbia, my marginal tax rate for a $60k salary (US$41k due to the worthless canadian dollar) was 59.5%. I still have the tax software to prove it. There were bus drivers in Vancouver who were my age and earning my salary. That was very discouraging.

    Comment by systemBuilder — 11/1/2014 @ 15:31

  69. @systemBuilder

    If you are on a public sector pension plan, then your retirement age is determined by the number of years you work. If you start working 5 years sooner, you can retire 5 years soon. It is that simple.

    Admittedly, my statement is wrong if you are doing the savings yourself.

    Comment by Daniel Lemire — 11/1/2014 @ 16:01

  70. @ Daniel,

    so in Quebec you have a public sector pension? I should have taken that McGill job LOL … At UBC it was RRSPs and nothing else. At UIUC it was a savings plan guaranteeing 8% (for faculty), but hourly staff had a sweet public-sector plan based upon years of service. I’m pretty sure that defined-benefit pension plan been taken away for everyone now …

    Comment by systemBuilder — 11/1/2014 @ 17:28

  71. @Daniel Lemire (#67): How many weeks of vacation would I have gotten every year as a worker in an American corporation in my 20s? How often would I get projects over which I had the huge level of control that I did?

    As for your point about loss of income, underlying it is the usual fallacy that money buys you freedom or happiness. Beyond a point, it doesn’t. After paying for my cheap grad student apartment, I had a reasonably high disposable income, meaning I was able to read and date and eat out and travel the world (especially if you account for all the money earned during internships) to the extent I wanted to. Bigger houses and flashier cars were never my thing. A Mercedes, after the first few months, is just another car. My grad student apartment was nice enough to impress (or at least, not depress) my friends and lovers, and that was all that mattered.
    :-)

    At the current point in my life, I have a lot more money, but I am also a spread a lot thinner. I am not sure my quality of life is much better than before, and this is in spite of the fact that I do work I love.

    More generally, one huge reason I don’t ever want to work in the industry is that I like my flexible summers. I can work in the summer from Europe or Japan or China or India or wherever. I am willing to take a pay cut for this freedom. As a grad student, if you are strategic enough, you can arrange for such things too. As an advisor, I am very open to allowing grad students do such things. I wouldn’t mind if one of my grad students wants to write her dissertation from an island cabin in the Caribbeans, for example.

    I don’t particularly care to retire at any point, but if I did want to retire, I would do so in a country where the cost of living is much lower.

    @systemBuildee (#68): You are missing my point. I did win the academic lottery, but my point is that even if I hadn’t and gone off to work at Facebook or Google or wherever, I would still have enjoyed my 6 years of grad school. As you know, the “college experience” is a hallowed American institution. Grad school is, in a way, 5-6 more years of college without losing much as a result.

    Comment by CS professor — 11/1/2014 @ 19:55

  72. @CS professor

    I did win the academic lottery, but my point is that even if I hadn’t and gone off to work at Facebook or Google or wherever, I would still have enjoyed my 6 years of grad school.

    I don’t think that, at any point, anyone did allude to the fact that grad school could be unpleasant.

    Lots of people enjoy grad school. I certainly did.

    My original post was bluntly about what happens after the PhD, should you not be part of the elite that lands a job at an interesting university. At this point, we are talking about about 90% of all PhD graduates.

    What people say is, in effect, don’t worry about it… companies are hungry for new PhDs and they’ll offer you an exciting job (compared to what you would have gotten with a BSc).

    I think, like McGowan, that this is largely a myth. There are a few people who land industry research jobs where they can continue their research… but most do not. They effectively become engineers working side-by-side with people who completed nothing more than an undergraduate degree. Worse: their lack of industry experience might set them back at a rank well below people who opted out of the PhD. This means lesser salary, fewer vacations and so on.

    Granted, lots of people would not feel short-changed by any of this… but to imply, from a tenured and interesting academic position, that you know, somehow, that you would have felt fine… well… how do you know?

    I have chatted with a number of PhDs who failed to win the academic lottery and they don’t feel fine at all. They will not write blog posts about it, of course…

    Comment by Daniel Lemire — 12/1/2014 @ 14:29

  73. @systemBuilding

    in Quebec you have a public sector pension?

    I don’t know whether every prof. in Quebec does, but it is common, yes. Not just in Quebec.

    Comment by Daniel Lemire — 12/1/2014 @ 14:32

  74. @Daniel,

    Who doesn’t enjoy themselves in their 20′s?

    As somebody who worked from age 22-24 in CA at a semi-R&D job (Xerox Office Systems Division, which begat the Macintosh computer) before going back to grad school, I can tell you that a lot of people are at universities (like my sister) simply because they have a terrific fear or disdain for the real world more than anything else.

    Comment by systemBuilder — 12/1/2014 @ 15:53

  75. @systemBuilder

    I can tell you that a lot of people are at universities (like my sister) simply because they have a terrific fear or disdain for the real world

    Thorstein Veblen would have pointed out that it is a matter of status. Everything else being equal, the more immediately “useful” you are… the lower your status.

    You have to have a high status for people to pay you to do things that are not obviously useful…

    Of course, things have changed a bit since Veblen’s time…

    Comment by Daniel Lemire — 12/1/2014 @ 19:34

Sorry, the comment form is closed at this time.

« Blog's main page

Powered by WordPress