Sunday, March 22, 2009

Should You Go to Grad School?

The Chronicle of Higher Education has a provocative article that strongly recommends not going to graduate school in the humanities. The last paragraph is particularly striking:
It's hard to tell young people that universities recognize that their idealism and energy — and lack of information — are an exploitable resource. For universities, the impact of graduate programs on the lives of those students is an acceptable externality, like dumping toxins into a river. If you cannot find a tenure-track position, your university will no longer court you; it will pretend you do not exist and will act as if your unemployability is entirely your fault. It will make you feel ashamed, and you will probably just disappear, convinced it's right rather than that the game was rigged from the beginning.

While the article is specifically about graduate school in the humanities, I feel that some of its points are also somewhat valid for computer science. The gist of the author's argument is: incoming students are not aware that the chances of getting a faculty job are tiny; further, even when you do get that prized faculty job, the job is not that good. I personally think being a professor is great job, but I can see how some could argue against that, especially considering how hard it is to get the job.

Let’s start with numbers. The number of people who graduate from “top 10” computer science programs every year is approximately 250. Conversely, the number of faculty positions that get filled at “top 50” research universities is about 25. That’s a ratio of about 10%, which doesn’t sound so bad (certainly not as bad as in the humanities), but there are two things that make the situation actually bad: (1) Notice that I took graduates from “top 10” programs and placed them in “top 50” programs, so this is not quite a fair comparison. The ratio becomes more like 4% if you count all graduates from “top 50” programs. At CMU, when we advertise a single faculty opening, we get approximately 500 applicants. That’s a success ratio of 0.2%. (2) The people who enroll in “top 10” computer science programs have already beaten the odds more than once. To be accepted into one of these highly ranked programs, you have to have excelled in an excellent college; to be accepted to an excellent college, you have to have excelled in high school, and so on. These are truly amazing individuals. To a large extent, they have 4.0 GPAs from college, perfect scores in the GREs, have managed to impress their professors to the point where their recommendation letters say things like “best student we’ve had in the last five years,” and by the time they graduate from college they have already published a few academic papers. All of this just to be placed in a situation where their chances of success are much less than 10%!

At this point you start wondering if being an NFL player is easier than getting a faculty job. I don’t actually know whether this is the case, but I can say one thing: a starting professor salary is about $120,000/year, and by the time you have become insanely famous or won the Turing Award, you’re making maybe twice or thrice that amount. The minimum salary for NFL players is about $300,000/year (and that's for like the rookie backup backup kicker), and if you become insanely successful, you can be making $30 million per year or more.

Ok, enough with the grim numbers. After all, things have worked out pretty well for me. Let me now give some reasons why the situation is not as bad as in the humanities and argue why going to graduate school in computer science is not that bad of a decision.

First, with a PhD in computer science, you can get a job at one of many great research labs or “researchy” companies like Google, and in many ways these jobs are better than being professor -- they certainly pay more with time. This means that the chances of getting a “good job” after getting a PhD in computer science are much higher than 10%. Second, the job of a researcher or a professor is pretty awesome: for all practical purposes, you have no boss! Also, according to many surveys, being a scientist is one of the most “prestigious” occupations. Third, I think graduate school is extremely enjoyable: you have about 5 years to work on WHATEVER you want, with very few responsibilities whatsoever. You don’t have a set 9-5 schedule (i.e. you can stay at home for days or even entire weeks), and you get to travel throughout the world -- as a graduate student, I went for free to Mexico, Hawaii, Austria, the Netherlands, Poland, Panama, Switzerland, more than 30 places inside the continental United States, and about 5 cities in Canada.

In the end, I think getting a PhD in computer science can be a good idea provided you actually enjoy doing research. But, (a) you should not do it for the money, and (b) you should be aware of how hard it is to get a faculty job afterwards.

Thoughts?

56 comments:

  1. Luis, awesome article... Phd program rocks. That is why I am always smiling :). Its a pity though that they do not get tenure.

    ReplyDelete
  2. I think graduate school is very enjoyable but I disagree with your third argument: I don't think that all graduate students have the freedom to work on anything they want or that most of them can get travel so extensively (30 sounds like a huge number to me). I have to admit though that I'm biased by being in the on of the "top 10" CS programs. :P

    One more reason why grad school is so nice is that it can give you the opportunity to meet and work with some incredibly smart and talented people. :-)

    ReplyDelete
  3. "that most of them can get travel so extensively (30 sounds like a huge number to me)"
    Luis was a celebrity grad student. Poor souls like us cannot go to so many places.

    ReplyDelete
  4. Oh, you'll travel more with time. There were some grad students that traveled more than me.

    ReplyDelete
  5. I have been to a lecture by professor (now the associate department head) James Hoe regarding whether or not it is worth it for ECE students to go to grad school. Basically, his conclusion was that getting a masters (if you can) is definitely worth the time and effort.

    However, he told us that getting a PhD is a completely different story because it is extremely difficult. Therefore, when applying for grad schools, apply as if you want to get a PhD (because colleges like it that way) but always leave the option of dropping out of the PhD program with just a masters degree open.

    He also emphasized how people trying to get a PhD should not try simply because 1) they think they will earn more money that way (they dont!!) 2) they think they are so smart (for them, few years of grad school will teach how many smarter people there are out there)

    ReplyDelete
  6. "2) they think they are so smart (for them, few years of grad school will teach how many smarter people there are out there)"
    I only agree with this however you need to change years to months.

    ReplyDelete
  7. Just noticed a significant typo in my previous comment. The sentence ``I'm biased by being in the on of the "top 10" CS programs'' was suppose to be ``I'm biased by NOT being in one of the "top 10" CS programs''. :D

    @ bradley.yoo: people should come thinking they are smart! Even more: they should be ready to fight (courses, quals) to prove it!

    Looking back I would say that a positive attitude is one of the most valuable thing to have during the grad school. :-)

    ReplyDelete
  8. Varun Aggarwala - I should have meant months not years..
    RME - There are people out there who think that they should become grad students only because they think they are so smart. I'm pretty sure few of them are ready to fight (courses, quals) to prove it. However, many of them think that competition in grad school will be easy as undergrad. It was really a statement meant for them.

    ReplyDelete
  9. I have to agree with the CoHE article. As a finishing PhD student in a 'top 10' program now on the job market, it sure does feel like I wasted a good portion of my life.

    ReplyDelete
  10. When I decided to go to CMU for my PhD, I didn't realize their were great jobs for computer scientists outside of faculty positions and research labs. Actually, back when I started my PhD in 1993, I'm not sure there were! But that's changed dramatically in the intervening years, and I can't imagine trading the job I have now for a faculty position, even at a top-tier school.

    My PhD has served me well--both the credential and the education / accomplishment that it documents. I can't turn back time and know what else I would have done with my professional life has I not pursued it, but it's hard to imagine I would have done much better than I did.

    I think the critical take-away is that going for a PhD in order to get a faculty position or a position in a traditional research lab is a crap shoot, and probably bad career move. I'm lucky that, even though I didn't realize there would be other, even better (for me) options, I found them nonetheless. I'd hope that students today would go in with their eyes open and make informed decisions about their professional development.

    ReplyDelete
  11. There is no such thing as a free lunch!

    ReplyDelete
  12. I think you make a good argument for entering a PhD program but not for going to grad school. There isn't as much benefit in going to grad school if your goal is a MS degree. An MS doesn't open many doors that a BS wouldn't open, and it doesn't open the doors that require a PhD.

    I'd also say that an MS doesn't offer the personal growth that a PhD does. The experience of completing a PhD is valuable even if you never do anything with it: doing original research, developing the self-discipline to work on a long-term project with minimal supervision, sticking with a narrow topic past the point of being sick of it, etc.

    ReplyDelete
  13. I totally agree with you John. I've never understood the value of getting a masters degree (except if you do it in a different area than your undergrad major).

    ReplyDelete
  14. The comparison to NFL players is unfair. The average career of an NFL player lasts 3-4 years, whereas for a professor it's 30-40 years.
    Also, if you can't get a faculty job, your cs ph.d. will still get you a good job at research/IT/programming whereas if you didn't get into the NFL your investment in your football skill is probably totally wasted (there are some other professional/semi-pro leagues but there is a huge gap in salary, prestige and other conditions)

    ReplyDelete
  15. The average career of an NFL player lasts 3-4 years, whereas for a professor it's 30-40 years.

    If you do get to play in the NFL, odds are you make more money during those few years than you do as a professor your whole life. Also, you're comparing average versus max. The max length of a career in the NFL is more than 10 years.

    That said, I sort of agree with what you're saying.

    ReplyDelete
  16. Great article. More data is always better!

    Here is the obligatory link to the Taulbee survey:
    http://www.cra.org/statistics/

    One thing, median starting salary across the US is actually around $92,000 (for 9 months) according to the survey (well, was, about 2 years ago, new survey should be published soon).

    In there I also noticed that only about 9% of last year's new PhDs found tenure-track positions. This is a new record low, I believe. Still, they had <1% unemployment, which is much better than the humanities PhDs!

    ReplyDelete
  17. $92k for 9 months is about $120k for 12 :)

    ReplyDelete
  18. NSF will only fund you up to 11 though... :)

    ReplyDelete
  19. I am nearing graduation with a Ph.D., and I am finding for the first time how difficult it is to get faculty positions. Had I known about the 4% success rate (not being from a top ten school, that applies to me), I might have chosen a different (and more profitable) way to throw away my 20's.

    Ultimately, I can only blame myself for not asking the right questions at the start of the program. But unlike the careers of Astronaut or Navy Seal, I don't think most people think of university professor as being something all that unattainable *by Ph.D. graduates*. I think my faulty assumption that the hard part was getting the Ph.D., as opposed to getting hired after the Ph.D., is understandable and widespread, and Ph.D. programs should do more to educate new grad students about the opportunities and be frank about the prospects.

    I'm not sure if I agree with the article that claims this is intentionally done to get cheap labor. It may be selection bias because departments are run by the 4%. But intentional bias or no, it is a moral imperative for any advisor to tell his or her students the real prospects of academic careers. I thank Luis for having the stones to bring this up.

    ReplyDelete
  20. I think we are missing a point here. Grad school (at least in computer science) is what can be called a 'rock star' profession. I think most students do not enter a CS PhD program aiming for a life marginally better than an average CS PhD- working in a research lab and without much acclaim beyond a narrow set of people. When they enter the PhD program, they aspire to do something which may be the next big algorithm/technology/paradigm.

    Contrast that with the set of aspirations when somebody enters a medical graduate program or even undergraduate engineering. Their aim (even at best) is only marginally better than that of an average outcome. How much more money/acclaim can a doctor get by only being a doctor (not starting a company, hospital or doing research). Grad school in that sense, is in IMO not very different from the set of people who aspire to be say actors. There are a lot of failures, the average outcome is not very appealing, but most people do not aim for that. Same goes for other (at least) some research professions. You have Feynman/Perelman and then there is a physicist/mathematician mostly spending his life in obscurity. And compare that with the difference two dentists (even at extremes) have in their lives.

    So, I think unreasonably high expectations, high failure rate (on initial expectations) is only natural in grad school. And that is only good- people would not be trying so hard, if it was not for the high aspirations. In the end society benefits from the results of the entire pool. And IMO the same distribution/properties hold true for other related things like papers/inventions/books etc.

    ReplyDelete
  21. That's true, but culling the "rock star" field sooner would help everybody. A big problem is that profs at worse schools have just as strong incentives to take students as profs at good schools. In other words, they have strong incentives to ruin their advisees' lives. :) One partial solution would be to say that only the top 10 schools can grant PhDs.

    ReplyDelete
  22. I wonder if this means grad schools should kick out more students than they do (referring to your earlier post on how few schools kick out undergrads). Tough as they are, maybe those schools are being nice in the long run and saving you 5 years?

    Also, do these CS numbers change much when you consider postdocs or others who get a faculty position a couple years after graduation?

    ReplyDelete
  23. Seems that the numbers should be somewhat more favorable than 10%: Many of those 250 graduates go to industry and are not interested in faculty positions, so the others are not competing with them. Of course it would be hard to measure how many take an industry job after balking at the academic job search.

    ReplyDelete
  24. An asst. prof in CS makes 92k a year? That sounds fairly high. I was under the impression that asst. profs made less than that...

    ReplyDelete
  25. Great post Luis, thanks for sharing. I think that one important factor (to some) that you neglected to consider is the percentage of time that you need to devout to your studies as a PhD student. Yes, its great that you can technically do whatever research that interests you, but you still have to dedicate a significant amount of your time on a daily basis to it and keep that up for several years. If research is all that you are really interested in, that is an awesome state. However if you're like me and have a very eclectic nature and wide span of interests, then its not so great. I like to compete in running and triathlons, work on open source projects, participate in discussions about religion and politics, and have a social life with people outside of school and work. In general, I strive to keep a well balanced life of intellectual, physical, and emotional wellness.

    I've received my bachelor's and master's degrees in computer engineering from top 10 schools (Purdue, UT Austin). Before I started my graduate work I had the full intent of going on to complete my PhD and even entertained the idea of becoming a professor. What I found as a grad student was that my life became completely imbalanced. I was not physically or emotionally satisfied with my life, even though I was completely fulfilled intellectually. Mine is a long and bloody story, but zero financial assistance, working full-time while still a student, trying to continue my other hobbies and activities, and an anxiety disorder diagnosis forced me to end it prematurely after I finished my master's requirements for the sake of my own well-being.

    To this day I still entertain the notion of going back and working on my PhD, but I am afraid of throwing my life out of balance once again and that's what continues to keep me at bay.

    ReplyDelete
  26. Let's turn the question around: what does or should motivate someone to pursue a PhD in computer science? If the answer is the prospect of a tenure-track faculty position or a comparable job in an industry research lab, then that person should realize they're facing long odds. If the answer is the opportunity to immerse yourself in computer science research for 4 to 15 years and then find a software job that emphasizes problem solving, I think the odds are much better.

    Luis, that's pretty much what you say at the end of your post. In an era where so much data is widely available, the scarcity of faculty positions shouldn't be big news to entering students--particularly to computer scientists who should have skills to find it and the initiative to seek it out. Caveat discipulus!

    ReplyDelete
  27. Any advice for a student that has excelled academically and would like to go to grad school, but did not attend a top university?

    ReplyDelete
  28. Apply :) You probably want to try doing some research -- that will help your application a lot.

    ReplyDelete
  29. Hi Luis,

    I was wondering where you got these numbers from (i.e., 250 PhD graduates from top 10 schools and 25 faculty positions filled at top 50 school). I don't think the second number is correct.

    I looked up the 2006/2007 Taulbee survey (available at http://www.cra.org/statistics/survey/0607.pdf). They list 288 PhD graduates for the top 12 schools, thus assuming 250 PhD graduates from the top 10 schools seems reasonable. However, table 18b in that report says that the top 36 schools alone had already filled 58 tenure-track positions. If you extrapolate this linearly this leads to approx. 80 tenure-track positions filled by the top 50 schools.

    Thus, now the ratio of 250 graduates from the top 10 schools to 80 tenure-track positions at the top 50 schools doesn't look so bad anymore (approx. 32%). Furthermore, note that this is only for tenure-track positions and is not even considering academic research or teaching positions.

    ReplyDelete
  30. Another thing about computer science in general I like is that it helps in real life thought process or may be vice versa. [:o]
    e.g. Your using "Human Computation" via your blog [;)]
    +Research is Adventurous....(Sometimes)

    ReplyDelete
  31. Sven,

    My numbers are mostly a projection for this year, and they are lower than 2006 because many universities are not hiring.

    ReplyDelete
  32. I even find the original numbers given by Luis (25 positions for top 50 schools) doubtful.
    Faculty positions are very less and there are some awesome phds from > top 10 ranked universities. So the competition for top 50 schools is fierce.
    But this raises another question...
    Is it better to be a faculty at >20 ranked university or work in a research lab??

    ReplyDelete
  33. My question was in respect of funding. If the faculties at top 10 institutes face a funding crunch than what about people at even lower ranked universities.

    ReplyDelete
  34. I have a question which is tangentially related (and I apologize for the deviation), but this question has been bothering me a lot lately.

    Most people here are mentioning the "top 10" programs, and Luis points out that they often come from the "top 10" undergraduate programs. I want to graduate school (for the right reasons, I think), but I do not go to a very good school for computer science. That doesn't change the fact that I work my ass off, have a paper published, strong letters of rec, and the Barry Goldwater scholarship. But my question is will I ever be able to work out of the deficit caused by the name of the school on my transcript? Or would an application to the top 10 programs be a waste of time and money? If I cant make it in to a top 10 program, is it still worth going? What about the top 20? At what point is it not worth it since my chance of being employed at some type of research institution (university, lab, company) is too low?

    I'm sure for most of you, this is a non-issue, but it would suck if I ruined some of my dreams at the age of 16.

    ReplyDelete
  35. If you're only 16, you have ample time to work on an application for a good program. Keep on doing research :)

    ReplyDelete
  36. I'm sorry, I should have qualified that last statement. I screwed up at the age of 16, so I didnt get into a strong undergraduate program and I am now wondering if that fact is something that I cannot overcome in the eyes of the top 10 departments when I apply for graduate school admission.

    ReplyDelete
  37. The answer is: you can still get in, but it's slightly harder.

    ReplyDelete
  38. Does someone with a 4/5 GPA from a top 10 school have a shot at a PhD anywhere?

    ReplyDelete
  39. If you've done some research, yes.

    ReplyDelete
  40. Luis,

    Excellent post. Sadly, my PhD is (a) interdisciplinary (environmental studies) and (b) not in computer science. Tenure-stream positions are getting 160 applications for 1 position. Scary thought. This is an example for Canadian universities.

    Funny enough, I was going to do my PhD at CMU but my (former) PhD advisor moved to Canada and thus I did it here. And now that the PhD is well behind me, I'm starting to question whether I should have done it in the end!

    ReplyDelete
  41. Luis,

    Thank you for posting this! Since finishing my undergrad I have been contemplating changing my career to 'start over' by entering a PhD program. I am concerned because I have been out of school for about 4 years and had finished with a A-/B+ avg at a NY state university. What is your opinion from knowing students in a similar situation as I am in what their success level is with regard to re-entering the academic role after 'time off'?

    ReplyDelete
  42. I know of many people who enter grad school after taking some time off. To get into a top grad school, though, you will most likely need some research experience.

    ReplyDelete
  43. Interesting point. Just google "profzi scheme", it seems that the author of PhD Comics has a similar point.

    ReplyDelete
  44. But getting an academic position should not be considered the declared goal of going to grad school. As noted in this week's Science, "we must promote the movement of scientists into many occupations and environments
    if our end goal is to effectively apply science and its values to solving global problems."

    ReplyDelete
  45. I have a master degree in humanities, and next year about now I will also have my PhD (as I am currently in my last year of grad school. But I have no real hope in finding employment within the academic world when I am done, so I have decided to go back to school to finish my (third!!!) degree in teaching. Hoping this will give me something to fall back on. However, the furutre for high school teachers is as bleak as jobb oppertunities within academics. I have spent the better part of a decade educating myself, but to what end???? To an underpaid job which I am clearly overqualified for???

    Why do the universites allow us to attend grad school if they dont want anything to do with us when we are done? And why dosent society at large want to take advantage of the education and knowledge we have??? I really don't get it...

    Someone might say that I have myself to blame because I didn't go to medical school, or become a mathmatician or something like that. However, I have a slight difficulty with blood, and I am no numbers person. I would have been a dead end.

    We are told that we should follow our dreams, find our passions, and live life to the fullest in the pursut of all of the above. Well I did, the problem is that now they tell me it is not enough.

    ReplyDelete
  46. Is "Top-10" school/department more of a qualitative descriptor than an actual set of 10? For example in CS, right away we know which departments will be part of such a list (MIT, CMU, etc). But what about in the 7-10 part of the list? Penn and UMd CS are not in the top 10 (as of last year's US News rank) but I consider both to be of top-10 quality.

    ReplyDelete
  47. >>> "If the answer is the opportunity to immerse yourself in computer science research for 4 to 15 years and then find a software job that emphasizes problem solving, I think the odds are much better."

    This is not a very inspiring mindset. It is easy for those who succeeded to say - "Oh, do it just for the sake of it, it's okay if after 5 years you come back to a job that you could have gotten just with an undergraduate degree".

    I want to do grad school, but feel that outcome would mean a total waste of time... If I knew I can't get a research position, I would never do it. I can very well think on my own and continue to learn & solve problems, outside of a grad program.

    ReplyDelete
  48. Greg, if you have a Goldwater scholarship, consider that the criteria for that application and many grad schools is essentially the same. Except for Computer Science, the Goldwater has a much lower accept rate (only 7 nation-wide this year)
    [full disclosure: i'm one of the seven]

    ReplyDelete
  49. Thank you for the encouragement!

    ReplyDelete
  50. It is great topic of debate & in my point of view every one must go to graduate school. I know some people does not agree with me, but exception is everywhere.I think grad school is fantastic if you know what you want to do and a grad degree will help you get there.

    -----
    Praxis study guide

    ReplyDelete
  51. I would advise going to graduate school as an end in itself, not for any future career. The future career is only guaranteed in a few select fields.

    I enjoy graduate school because I get many of the perks of being an academic professor without any of the burdens! Free travel, totally flexible schedule, feel like you are contributing to knowledge, get treated as an expert in your area.

    The big rule is, ONLY go if you get a free ride and a decent stipend; unless you are going for a professional degree.

    ReplyDelete
  52. I feel lucky can read this useful news. Now I find something what i want to know.
    Thank you for this great information.

    Kenali dan Kunjungi Objek Wisata di Pandeglang | Blog SEO | cah bagoes | oes tsetnoc | blogger

    ReplyDelete
  53. Biggest Praxis II Test Blunders Revealed, Praxis II Test & Praxis Test Prep Debate Spawns Rage, How You Too Can Pass Your Praxis II Test!

    ReplyDelete
  54. I don’t know If I said it already, but this so good stuff keep up the good work. I read a lot of blogs on a daily basis and for the most part Belajar seo blogspot just wanted to make a quick comment to say I’m glad I found your blog. Thanks.

    Best regards
    Seo Motivation Oes tsetnoc

    ReplyDelete