The Ph.D Bust: America's Awful Market for Young Scientists | Hacker News Reader

The Ph.D Bust: America's Awful Market for Young Scientists

theatlantic.com

143 pointsjseliger119 comments

I am surprised that so many people are relating PhD's to the money you'll make. I hear (very frequently) that "the amount of time you spend doing your PhD will put you far behind the 4 years of having a job" or "if you want to get rich, go into ...", or "management is where all the money is; you need an MBA for that".

Uh... I'm doing a PhD because I like scientific research and can't stand the thought of being a process engineer. If I'm going to spend 8 hours a day working on something for the rest of my life, it better be something I enjoy! While it is possible to do R&D with a bachelor's degree, your chances of being able to do exactly the kind of research you want goes up significantly with a PhD, plus you've already had the experience of doing it for 4 years.

Also, I'm not sure where this "not much money" concept comes from anyway. I'll start at an engineering salary, right out of grad school. I guess that's not "Hacker News wealthy", but that's plenty more than I need to even live comfortably!

Good for you. You'll be in good company. These folks aren't HN-rich either: fighter pilot, fireman, teacher, social worker...

[Retracted; I was conveying my point incorrectly.]

I don't see that. I think that someone who complained about $250k would be laughed at, and even mocked openly. Especially keep in mind that we have international readers.

The relative compensation in many companies between engineers and useless managers is another issue... but anyone who complains about poverty at $250,000 can eat it.

your chances of being able to do exactly the kind of research you want goes up significantly with a PhD, plus you've already had the experience of doing it for 4 years.

I'm not so sure that's a valid generalization.

I'm getting a PhD in computer science, but any corporate R&D department I know of that would hire a fresh graduate would only let you do R in service of their particular D needs, i.e., you have too little autonomy.

I definitely think I can find jobs that I love and that are lucrative (I have some in mind), but I won't be doing research as, say, a tenured professor would define it.

I'm not even sure that all tenured professors are doing research as they would define it.

Absolutely not. Look at the publication record of any tenured professor at a non R1 university and you'll typically find that it's dropped to almost nil.

True, I really just meant "at R1 universities." Which isn't fair to all those other tenured professors at non-R1 universities (and I may possibly even be one, someday).

Of course, I have heard of R1 tenured professors who claim to spend 50% of their effort procuring funding (though I think 10% is a much more representative estimate).

I've heard believable claims that "here professors don't write research papers, professors write grant proposals" - quite a few of them advance research and get paper authorship by leading a lab and procuring funding for the team that does the actual research and writes the papers.

And I'm not dissing them in any way - getting a grant requires the senior/experienced guys working on it. Getting the funding is a must-have (since without it researchers and students would be dismissed and the research would not exist); but spending your own time researching is a nice-to-have thing that you can do if there's any spare time left after teaching and bureaucracy.

Also, the research that people are writing grants for and doing is dictated by what is fundable, which is not necessarily exactly what they'd choose to be doing were there no constraints.

I know tenured professors who don't apply for grants, and therefore don't have any money to pay students, but they still get paid a salary by the department. They can do whatever they want (research, or not), modulo teaching responsibilities. Some profs actually do very good research, without students. It's a reasonable way to go, once you have tenure.

Problem is, you can't take that route to get tenure in the first place, unless you're an absolute genius.

Most places are redefining tenure though, so the only people that still have any meaningful resources are those that have extramural funding. Anymore, tenure means you get a closet for an office and a phone. And for new hires, if you don't have grant support, even with tenure, you can still lose your job.

Writing a grant proposal is about as hard as writing a paper. So if they apply for three grants and also write three papers in a year, then yeah - that's 50%.

Getting a grant proposal accepted is much more difficult than getting a paper accepted.

I don't think that's a valid generalization. It depends enormously on the funding agency and the place you're submitting the paper, respectively.

You're not going to be able to do a good grant proposal unless you have a good track record of research, with multiple papers. So in that sense, a grant proposal supersedes a single paper.

But, given that you've already done all those papers, a grant proposal is just writing, not new research, so it should actually be massively less work than doing all the original research that goes into a paper.

I see this negative claim on HN quite a bit. R&D CS labs where D is in service of R do exist. I work at such a lab. We expect fresh PhDs to lead their own research project. Spend some time perusing the ACM DL in your area and try to find labs that have high publication-to-researcher ratios (many of these labs are quite small so it does no good just to do a raw pub count).

You ever hear of Horn from Horn & Johnson fame (Matrix Analysis)? Guy had a 9-5 for years and researched at nights in his home office. That would be my ideal setup, I think.

Sure, but, unless you're a rockstar graduate student, you do R in service of your advisor's grant needs. Unless you're a rockstar post-doc, you do R in service of your PI's grant needs. Unless you're a rockstar PI, you do R in service of your own grant needs. Wherever you end up, someone is writing the checks. That someone gets to determine what you research.

I'd guess that computer science is very different from experimental physical science disciplines: In experimental, you have to have a research advisor with grant money if you want to have vacuum pumps/oscilloscopes/glassware/liquid nitrogen/etc., even if you plan to TA your way through grad school.

As long as you don't do a Ph.D. with the assumption that you will get a permanent academic job, i.e., faculty at a university.

I don't really think it's the salary of a permanent academic job that people complain about, it's the fact that in most fields, the number of such jobs is far surpassed by the number of Ph.D.'s produced. When you realize that "not that much money" means "postdoc kind of not-that-much-money", then the outlook changes.

4 years?

For mine (chemical engineering in an American university). I hear it varies quite a bit depending on where you are and what you're doing it in.

Do like writing grant proposals? Because in the biological sciences that is what you will be doing, and you will be competing against other smart folks which results in a 6% chance of a successful proposal. Then you can do some research while drafting journal articles, teaching undergraduates, and working for tenure by serving on a bunch of committees.

In short, there is very little research in the above Ph.D. in academia career.

You make a good point with people relating PhD's with the amount of money you will make. PhD's not only have a cost of paying for university but also opportunity cost is also existent. Also I completely agree with your comment on the idea that you can live comfortably, not rich on a PhD.

The missing half of this article is the failure on the part of academic departments to even acknowledge this problem publicly in their recruitment process.

I was an astrophysics Ph.D student at a fairly well-known program until last year when I left with an M.S. to go full time for the web development shop I started. It was a good decision on my part, for sure, but what really struck me is that after I broke the news to my advisor, he was quite open with me about how poor the job market is (not that I wasn't already aware). He and many other faculty members in the department sung a very different tune to me and my classmates during the recruitment process, as did my undergraduate mentors. This is an endemic problem throughout astronomy (at least) and probably many other disciplines in the physical sciences as well.

Even the students often seem to have a sort of Stockholm syndrome about the problem. I still hear from lots of my former classmates that a.) they're well aware of the extreme shortage of jobs in astronomy, b.) they're not seriously expecting to get an academic job and are aware that there are virtually no non-academic jobs doing astronomy and c.) they'll figure out how to get a job "in industry" (i.e. what the rest of us call "having a job") when they finish.

Many students in these programs seen to labor under the assumption that if academia doesn't pan out, their programming skills or quantitative knowledge will make them good candidates for a software or finance job. This is not really as true as they think, since as most HN readers are aware, good developer jobs entail knowing about a lot more than just a programming language, and the sort of programming and quantitative analysis you do in Ph.D research is really pretty far from what those of us in the private sector do with our programming skills.

Nonetheless, a lot of these same classmates would go full-out with their encouragement of prospective students when they came to visit. In my last year at grad school, I remember going on a long rant at the prospectives about how bad an idea a Ph.D in astrophysics is, and the looks of horror on my classmates' faces.

I agree deeply. After I chose not to do a PhD I was told all the downsides from everyone. Before that it was all Roses to try and get me to do my PhD with everyone.

This goes well with Philip Greenspun's "Women in Science:" http://philip.greenspun.com/careers/women-in-science, though I don't think his argument about medical school is a good one.

(See my view on the perils and opportunity costs of med school here: http://jseliger.wordpress.com/2012/10/20/why-you-should-beco... ; Greenspun hosted this: http://philip.greenspun.com/careers/why-i-gave-up-practicing... in 2011.)

There's certainly a big misconception (in the U.S. and elsewhere)that study and research produces innovations and innovators. However, academia really fosters a type of conservative innovation. You pick your battles, apply for grants that you think you can win (and hope for the best).

For the most part, you don't swing for the fences. You learn how to integrate yourself into a certain type of system, and you hang on until you graduate. You operate in the zone where the marginal return on labor is usually miniscule.

In my opinion, Academia is pretty poor at fostering and developing any type of disruptive innovation. This kind of innovation is generally what makes headlines and gets funding.

Disclaimer: I am by no means the first person to have this thought. Here's a blog post from Forbes: http://www.forbes.com/sites/ericaswallow/2012/04/19/innovato...

Two points not made here:

1) It's quite common to study for PhD while getting paid for it instead of paying for it - and in that case, it should be treted as a fun and enriching (though not very high paying) job for a couple of years instead of "delaying 4 years". Getting into a huge debt for it is a whole different thing, though.

2) It feels that this problem is partly USA-specific. Sure, other places may have noticed similar tendencies but not as sharply; and science is very, very global - I'm seeing a lot of great researchers moving between countries/universities every 5-6 years based on where the major research projects in their area are happening & funded.

In general, a solution would be to try to decouple research from teaching instead of mixing them all under PhD/professor positions - the future tendency seems to be with much less people needed in teaching (due to changes in society, student funding/loan systems, MOOC's, etc) and thus more PhD's shifting towards research.

Opportunity cost. PhD stipends pay less than unskilled labor, after you subtract "tuition"

Tuition should be covered by grant funding (RA) or the department (TA). If a university "accepts" you in a science/engineering PhD program, but is not offering tuition and salary, they're really saying they can't take you. Note that at many universities, students will cost a grant more than postdocs for this reason.

I think his point was that if you don't consider your salary as your take home pay plus the tuition that is paid on your behalf, then it is a very meager income.

If you consider the sheer number of hours I tend to spend on my dissertation work compared to the stipend I get, I'm paid criminally less than minimum wage. But you're not given the leeway to focus "exclusively" on a single research contribution in a focused manner in other higher-paying options.

Even if you're not taking out a loan in the monetary sense, you're taking out a loan from your long term earnings, one that has little chance of being repaid, to put your mental assets/skills to a non-remunerable task, and get a certification that you did so. The problem is that the option to continue doing this (tenure-track academic jobs) are limited and (naturally) highly contested.

However, you make a good point--there's an underlying and insidious opportunity cost that is often unknowingly sacrificed: that of atrophying skill sets. It's easy for a PhD to be a hugely insular experience, if you let it, and if you take the easy way out and don't stretch your engineering skills (speaking in terms of CS here, since that's what I know), you're in for a rude awakening if you determine that academia is not for you. If you're not careful, you'll get good at writing papers, but might actually get /worse/ at writing portable, readable, and maintainable code. And as brilliant as your papers may be, if you can't ship good code, you're going to have trouble in industry.

The good thing, again at least for CS students like me, is that the "fun and enriching" environment of academia means a lot of opportunity for starting companies, creating libraries/frameworks, working on side projects, and doing contract work, so there's no reason you have to atrophy. Which is something that, sadly, the visions of the tenure-track academic job are engineered to beat out of you.

> If you consider the sheer number of hours I tend to spend on my dissertation work compared to the stipend I get, I'm paid criminally less than minimum wage.

That may be so. Are you familiar with the concept of a wage being equivalent to the marginal product of labor (or value of last hour worked)? :).

I'm with you. I'm working on my dissertation at the moment.

Depends on location and situation; here in EU the take-home pay (if the professor has a grant/project to take in PhD students) tends to be quite livable - less than an equivalent commercial IT job, but more than unskilled labor and more than, say, a humanities master would make in their area.

It will no doubt be shocking news to Mr. Weissmann to find out that factors such as a desire to learn more or to make new contributions to human knowledge can motivate aspiring PhD candidates. It's also rather ironic to see such analyses promoted by someone who chose a career in the notoriously unprofitable field of journalism.

It will no doubt be shocking news to Mr. Weissmann to find out that factors such as a desire to learn more or to make new contributions to human knowledge can motivate aspiring PhD candidates.

"A desire to learn" and "Make new contributions to human knowledge" are laudable goals that I admire. But, even leaving aside the important question of whether grad school as it is currently structured is a good way of pursuing either, it's hard to make new contributions to human knowledge when you're having trouble making enough money to support yourself, and it's frustrating to see your new contribution made when your buddy is working for Google and writing software that millions of people use every day—and getting paid well for it.

It seems not unreasonable to me to balance life / career goals with learning and wealth. The opportunity costs of grad school are incredibly steep. If you are the rare person who doesn't care at all about material possessions or the physical quality of life, then by all means go to grad school. But if you go expecting a tenure-track job at the end—which most people seem to—then you're making a mistake. I am not at all opposed to someone who simply says, "I don't care at all about income."

I've been to graduate school and disagree with your comment about expecting a tenure-track job. When you get to know them, PhD students are not poor deluded souls slaving away in the misguided expectation of a plush, secure position at the end. They are rather intelligent, motivated individuals, who choose to work on projects that are more interesting/important than optimal ad placement or crafty financial swaps, and who are prepared to sacrifice some Caribbean cruises and latest model cars in return.

Yeah, when I was in grad school, I and my friends said the same thing. However, when reality sinks in that choosing to work on such interesting stuff means living year-by-year on soft money at a postdoc salary, and that in many fields you have little advantage transitioning to a nonacademic job over someone with a MS, it doesn't seem like such a good deal anymore.

Also, I'm not sure you're calibrated correctly. It's not about sacrificing "latest model cars", more about sacrificing a reasonable chance of retirement.

It's a few years in your 20s working on an interesting and very difficult project. Don't worry, you'll still be able to retire at thrice that age.

I have too (see here: http://jseliger.wordpress.com/2012/05/22/what-you-should-kno... for some comments, mostly job-related), and I think most grad students—at least during the first half of their experience—do think they'll be the exception.

They are rather intelligent, motivated individuals, who choose to work on projects that are more interesting/important than optimal ad placement or crafty financial swaps, and who are prepared to sacrifice some Caribbean cruises and latest model cars in return.

I think we'll have to agree to disagree. Most seem to be pointlessly delaying adulthood. Note that there are exceptions.

Without data the rest of this discussion might be pointless, but the prevalence of articles warning against PhDs seems to me to point in an important direction.

Ah, right, because doing anything financially sub-optimal is delaying adulthood.

Can you cite a sentence in which I said that?

Can you give me a concrete example of delaying adulthood? Or explain to me how going to graduate school delays adulthood?

You can work on more worthwhile things than ad targeting in industry and still get paid a full salary.

However, it is true that a disproportionate amount of jobs will follow the money.

it's frustrating to see your new contribution made when your buddy is working for Google and writing software that millions of people use every day—and getting paid well for it.

I don't think the average grad-student could work at Google instead. My personal experience is that Google keeps having its recruiters call me periodically, and then rejecting me for even internships after interviewing.

I don't even get the questions wrong anymore, so I feel like I just somehow suck, despite being a CS grad-student at a well-regarded institution.

Love your comment!!! I have met people who gave me weird looks because I want to do a PhD. And she said "what is so good at doing a PhD. A professor can't make more money than me as a small restaurant owner." She just doesn't get it..It is not about the money...sigh

The problem is the number of PhD's doing "nothing" after education is going up. "You don't need a PhD to do nothing, man. Take a look at my cousin: he's a dropout, don't do shit."

Awareness of the opportunity cost and likely outcomes has been hitting me hard lately as I consider dropping out of my top-5 science (not CS) program 2 years in. I've gotten to a comfortable place with python/numpy/matplotlib and am wondering how hard it would be to break into web development as somebody who has only programmed in a scientific setting. Besides Python, I also know enough R and statistics to build linear regression models, do significance testing, make pretty plots with ggplot2, and other things you learn in a first year grad data analysis class.

Those of you in the gallery: Have you or somebody you know successfully made the transition from (non-computer) science into a tech career? In the current economy, what are the chances you'd hire a highly motivated science dropout with programming competence and basic stats knowledge? Would something like Dev Bootcamp be worthwhile for somebody like me?

I hate buzzwords like "data scientist", but I think that might be my best angle. What I'd really like is to apprentice in a Django shop, or possibly wear a little of both hats as a dev and a "data scientist".

"Have you or somebody you know successfully made the transition from (non-computer) science into a tech career? In the current economy, what are the chances you'd hire a highly motivated science dropout with programming competence and basic stats knowledge?"

I did it, have worked with others who have done it, and would now hire people who did. The important part is your ability to program. There are tons of PhD students (even in CS), who can't (or won't) write code. There are even more who write horrible, unmaintainable code. You have to be better than those folks.

Also, do not be deceived by "data science": it's mostly a bullshit term, and translates roughly to "programmer who knows basic statistics", rather than "scientist who knows some programming". Nobody wants to hire you if you can't implement your theories in a production context.

The bottom line is that if you're a good coder, nobody cares how you wasted your youth.

Also, do not be deceived by "data science": it's mostly a bullshit term, and translates roughly to "programmer who knows basic statistics"

This is utter rubbish; I really wish people would keep quiet about things they know nothing about. I suggest that you have never actually discussed a domain with a data scientist if you think it's "basic statistics".

In our dev. shop, we have a lot of great programmers, but none of them can touch our data scientist when it comes to working out what our tens of millions of users are actually doing and what their salient attributes are.

As for the data scientist needing to 'implement their theories', that's what the developers are for. The data scientist does the analysis, then works with the developers to implement systems that incorporate the results. Neither group is capable of the other's work.

"I suggest that you have never actually discussed a domain with a data scientist if you think it's 'basic statistics'."

Utter rubbish, perhaps. But since I've actually done the job, I do happen to know something about the subject. It's a marketing term, not a term of art.

The vast majority of "data science" performed at web companies boils down to knowledge of summary statistics and probability theory, a smattering of basic statistical models, and (most importantly) the ability to write code. There's not much that would challenge an advanced undergraduate, let alone a doctoral-level statistician.

Hey jurassic, we (repustate.com) might have something for you. Send us an email if you'd like to hear more.

I've known several people who have transitioned into the tech industry without a CS degree. Particularly good examples that I can think of got their degrees in physics or some type of engineering; I even knew a guy who got his degree in political science. Really, anyone who is interested and motivated enough can learn programming, analysis, and good Unix practices.

I did it. I did a Master's (never had any intention of a PhD) in engineering, mostly focused on OR decision making applications in geography. So I knew maps and could write Matlab and Python scripts and I'd played around with some C/C++. I progressed into a tech career starting with a place where the mapping & geography knowledge was a huge asset. My CS experience was limited to 2 undergrad courses and reading a few books. It's taken a few jobs since then and about 5 years but I'm definitely purely a software developer now. Two things that I think helped were having a technical bachelor's degree (even though it's not CS) and having some projects I could talk about that that meant I knew stuff about the company's products that CS students wouldn't tend to know, like mapping projections and norms in GIS. When they asked me questions about stuff like sorting algorithms, I could explain some assumptions I would make based on the kind of work they do, not just give the generic 1st year CS course answer. Dev Bootcamp or similar might be worthwhile to give you some buzzwords for your resume and some projects you can show off. If you can get those on your own or your experience buzzwords match what you see in job postings, then it's probably not worthwhile. I wouldn't put the bootcamp on my resume, I'd put the skills and projects it got me down as personal interest projects.

The problem here is merely a misunderstanding of career statistics. The statistics for success in academia are thoroughly against any one Ph.D. The competition is too high, the rewards are too low, and the competitors are too desperate. This is not a place where you wish to compete - unless you have some kind of overwhelming advantage - aka really, really brilliant - you know who you are.

For everyone else - look straight at industry - where the competition is a hell of a lot lower, the salaries much higher, and the competition is really not that bright. If you are getting a Ph.D - and you know you're not a genius - be looking at getting a cushy engineering job at a large firm.

The benefits are great, the salaries are great, the job security is great and you can always jump back into academia. It's also the closest analogue to academia as you'll get, next to research labs.

Don't ever compete on the same terms as the people around you. Don't play a game where you will lose - a game where you are probably outclassed. Go look to where you can charge monopoly pricing and extract as much as you possibly can.

Competing for limited resources is a losers game - hell, competition SUCKS BALLS (just ask any Chinese manufacturers you know :) - monopoly is where it's at (just ask AAPL/GOOG/MSFT/TSLA).

Separately - people should go out and learn some basic statistics, microeconomics (macro is pretty fucking useless), game theory and psychology. These will help you to avoid getting stuck in what can appropriately be called real life versions of the "Hunger Games" (http://en.wikipedia.org/wiki/The_Hunger_Games).

Reminds me of Tony Hsieh talking about poker strategies, and the conclusion of the movie WarGames:

> Through reading poker books and practicing by playing, I spent a lot of time learning about the best strategy to play once I was actually sitting down at a table. My big "ah-ha!" moment came when I finally learned that the game started even before I sat down in a seat.

In a poker room at a casino, there are usually many different choices of tables. Each table has different stakes, different players, and different dynamics that change as the players come and go, and as players get excited, upset, or tired.

I learned that the most important decision I could make was which table to sit at. This included knowing when to change tables. I learned from a book that an experienced player can make ten times as much money sitting at a table with nine mediocre players who are tired and have a lot of chips compared with sitting at a table with nine really good players who are focused and don't have that many chips in front of them.

In business, one of the most important decisions for an entrepreneur or a CEO to make is what business to be in. It doesn't matter how flawlessly a business is executed if it's the wrong business or if it's in too small a market.

Imagine if you were the most efficient manufacturer of seven-fingered gloves. You offer the best selection, the best service, and the best prices for seven-fingered gloves--but if there isn't a big enough market for what you sell, you won't get very far.

Or, if you decide to start a business that competes directly against really experienced competitors such as Wal-Mart by playing the same game they play (for example, trying to sell the same goods at lower prices), then chances are that you will go out of business.

In a poker room, I could only choose which table I wanted to sit at. But in business, I realized that I didn't have to sit at an existing table. I could define my own, or make the one that I was already at even bigger. (Or, just like in a poker room, I could always choose to change tables.)

I realized that, whatever the vision was for any business, there was always a bigger vision that could make the table bigger.

Source: http://www.huffingtonpost.com/tony-hsieh/tony-hsieh-zappos-c...

> Instead, Falken and David direct the computer to play tic-tac-toe against itself. This results in a long string of draws, forcing the computer to learn the concept of futility. Joshua obtains the missile code but before launching, it cycles through all the nuclear war scenarios it has devised, finding they too all result in stalemates. The computer concludes that nuclear warfare is "a strange game"; having discovered the concept of Mutually Assured Destruction ("WINNER: NONE"), therefore "the only winning move is not to play."

Source: http://en.wikipedia.org/wiki/WarGames

In summary:

Competition is M.A.D. (http://en.wikipedia.org/wiki/Mutual_assured_destruction), the only winning move is not to play!

And long live monopoly!

This is a pretty awesome post that I hope near- and new Ph.D.s think about, especially because at that point in your life, you are probably immersed in a 'macho' culture where thinking about anything but a life in academia is seen as wimping out. Don't pay attention to that. It's nonsense. Pay attention to confluence's post.

The poker example is a great one. Years ago when I first started to learn to play a friend of mine told a similar story more succinctly.

When you sit down at a table if you can't pick out the sucker, it's you.

Pick your table, pick your game, do whatever you have to to slant the odds in your favor. If you're the sucker get up and move. Don't let your pride get in the way. Back when I started playing poker I could sit at tables in LV and do pretty well. As poker became popular LV became a much harder place to play. The quality (and recklessness of the player went up), so I quit playing.

The last time I played was on a business trip in Phoenix. After a couple of hands a woman already at the table and I realized we were the top ones at the table and just stayed out of each others way while drunk businessmen and golfers emptied their wallets :)

Ah but you see - there's too much competition for short, sharp comments!

I try and own the high value long comments. Although I relapse ever so often and move back towards the short/sharp side - winning a high volume competition is very alluring.

It's probably why reddit/stackoverflow/wikipedia/HN exist.

Nate Silver has a similar point[1], and claims that he has been successful by choosing fields that were not already saturated with excellent players. Baseball prediction was waiting to be cracked wide open with modern statistical techniques. It would now be much, much harder to make meaningful gains, you'd be competing with a host of competent firms who have years of experience and have long since claimed all the low-hanging fruit. The competition in online and even professional poker matches used to be much looser, until enough people started getting serious about it that the field narrowed to exclude merely above-average players.

To the OPs point, If you want to compete in a well-established academic field, you have to be really freaking good. Want to be an experimental physicist? You see the competition out there: they write books and appear on TV shows, or at least have tenure and get grant money. Are you really one of those guys or gals? On the other hand, there's interesting stuff out there that is just waiting for someone with a modicum of intellect and desire to revolutionize it. The challenge lies in finding (or creating) such a field and exploiting it at the right time. I think this can be done in academia or in business alike, business is just much more flexible and forgiving of mistakes.

[1] "The Signal and the Noise"

> Baseball prediction was waiting to be cracked wide open with modern statistical techniques.

Interesting point. How about Bitcoin? Am I too late to that game, or is it too hard to model?

As a currency people trade, you can always make money. Short it when you think it'll go down, buy when you think it'll go up. If you're right, you win.

Couldn't agree more! I'm an economics Ph.D. dropout who focused primarily on microeconomics and game theory (because like you said, current macro theory [see: DSGE] is completely bogus) and I think your points about finding monopolies is a great one. I left my Ph.D. early because it was a losing game as I wanted to be learning more computer science since it was a necessary skill to obtain in addition to a microeconomic/econometric education to master the "data science"-esk skills I wanted to in order to ensure I properly aligned myself for the future.

Although I haven't made the move to industry yet, I did recently get myself a nice little research gig doing big data work in a lab at one of those universities in Cambridge.

Hey, I'm an undergrad considering a phd in Econ. Could I email you some questions about your experience?

I'd also love to hear about the big data work that you do.

depending on your area of specialization, microeconomics and game theory are no less bogus than DSGEs. MWG, for example, does not mention real world data once...

As far as desirable traits, a job can be: 1. Interesting 2. Secure 3. Lucrative

You get to pick two at most, and that's if you're lucky--most people only get to pick one, or none.

Jobs in academia are certainly no exception to this rule.

OT, but I find it amazing that the triangle of desirable traits applies in so many places. There's the obvious "faster, cheaper, better. pick any two" and yesterday I heard another one "sane, smart and beautiful, pick two", which I'm sure has a corollary like "handsome, sensitive and rich, pick two". It's as if Brook's "no silver bullet" can be applied to just about everything our life (or it's just that easy to pick three desirable traits that are often mutually exclusive of one another)

Yes, and then there are the Venn diagrams that describe the disappointing result of any combination of two desirable traits. Here's one humorous example:

http://1.bp.blogspot.com/-7L9xAjgoZxo/ToydjJ7Hb8I/AAAAAAAABG...

Am I the only one who flared in anger that "Smart" and "Nice" combined to be the "negative trait" of Nerd? And that according to this we cannot be "Handsome"?

Inane crap like this helps absolutely nobody, and it perpetuates stereotypes that I think on HN we would be trying to avoid.

Why you would post this is beyond me.

yeah, that fascinates me too. there are even some actual theorems, like brewer's cap theorem [http://en.wikipedia.org/wiki/CAP_theorem] and arrow's voting paradox [http://en.wikipedia.org/wiki/Arrows_impossibility_theorem], that take the "pick any two" form.

That "amazing triangle" idea is just nonsense.

For example, computers have gotten faster, better, AND cheaper over time.

There definitely are women who are sane, smart, and beautiful. Are you kidding me? Just walk around any research university where they're doing graduate-level research and you will see lots of them.

As for the other gender triangle, pg is definitely handsome, sensitive and rich, so there's your disproof.

The "no silver bullet" thing is talking about something totally different. And it's an observation specific to software engineering (or, possibly, other specific domains, considered individually). There is a silver bullet for vaccinating against polio, but not for classroom teaching techniques.

The point that breaks the triangle(rectangle?) is if you add whether or not she is interested on you.

Yes and no.

Yes, you're correct. No, she's not interested. :P

Add this point before yours: whether or not she is taken already.

If you find a smart, sane and beautiful woman doing graduate research, it is highly probable that she is already married or planning for marriage with her boyfriend/partner as soon as she is out of school. (Well, this is from my experiences in Science and Engineering departments. Might be different for other fields.)

   faster, better, AND cheaper *over time*.

If it happened over time, it wasn't done faster. Speed in terms of processor is synonymous with "better" (or at least one of the aspects of "better"). A better processor is a faster processor. Faster in the original expression refers to when you want something done/completed/delivered by.

It might be that three traits usually represent an arbitrary number of dimensions sufficiently well (factor analysis, PCA). Or that thinking about more than three gets very hard (number of distinct combinations), so we just resort to three. For that matter, two dimensions, or even just the poles of one, are enough for most people (2-party politics, a/theist, occidental/orient, etc.).

In general, it breaks down into "less time discount," "more of what other people want," "more of what I want". That's a simplification, but broadly speaks to the tensions in the system. I used to like this system because it helped identify the tradeoffs, but what I found most surprising about it was how often it didn't apply at all, in business/products/jobs but what was most in egregious violation of the principle were people. Some people just don't have a lot of upside to knowing; other people walk around with some nominal flaws, but for the most part are inspiring.

I doubt this.

Very talented people can easily accomplish all three, especially in tech.

I guess the lucrative and secure part are supposed to be in tension? EG, if something is lucrative, there will be lots of competition that fights and eats away the security? But that same lucrativeness causes security, eg, golden hand cuffs.

Yeah.

In an economy with a healthy labor system, someone skilled in a certain field (say, software engineering) can just get a new (sufficiently lucrative) job if they lose the one they have, so there's no _necessary_ tension between those two.

Most people don't find software development "interesting" at all. And let's face it, the kind that pays well is usually rather boring, while the kind that is interesting usually doesn't pay very well, or at all in the case of open source.

Lucrative and secure are in tension only if the work is also very interesting (or fulfilling/challenging/stimulating, etc.)

If a job is all 3 of interesting, lucrative, and secure, then everyone will try to get that job, and the competition will either drive the salary down, or drive the job security down, or both.

A more general version of this rule is that anything desirable (or "scarce" if you want to use an economic term) in life will soon be pursued by others to the point where it becomes no longer desirable. That's the essence of how markets work.

The trick to wealth is to know (through some combination of vision and luck) what is going to be desirable ahead of time and obtain it before everyone else knows they want it.

One thing this article leaves out is the sky high attrition rates in PhD programs. The grim statistics presented are for the people who actually finish. Overall attrition rates for PhDs in engineering, the best of the bunch, are about 35%. It only gets worse for science and humanities. Keep in mind that attrition rates for elite law or medical schools are generally less than one half of one percent. So it is no exaggeration to say people fail out of elite PhD programs at roughly 100 times the rate for elite professional schools.

It does frustrate me to watch congress base public policy on the notion of a shortage of scientists and engineers when the evidence clearly does not support that assertion.

I'm not sure that all of them fail out. When I got my masters, the prevailing wisdom seemed to be that you had a better chance of being accepted if you applied for a Ph.D. and then quit with a Masters, which is what I did. Also, many of the other posters here decided that the job prospects were terrible and quit.

You are right that "mastering out" is not the same thing as failing out, so I may be overcounting here. Universities do also use the trick of admitting aspiring PhD students as MS students first, so that it doesn't affect their numbers when they fail to enter the doctoral program later. So there is also a possibility of undercounting as well.

I need to think a little more about this, but you're certainly correct in pointing out that it is more complicated that what I wrote - especially in engineering or CS, where an MS makes sense as a terminal degree goal.

The article fails to compare Computer Science PhD's who are in high demand and job openings for them far exceed the supply of CS PhD grads.

That's a big point that should be mentioned. The problem is getting a PhD in a field that no one is interested in paying you for leaves you with only academia as an option (I know multiple PhD's who realized towards the end that they don't want to go into academia).

According to the American Physical Society the same is true for Physics PhDs. Sure we're tiny relative to the life sciences, but we've got 4% unemployment one year after graduation (data from 2010, 2012 isn't available yet).

http://www.aip.org/statistics/trends/reports/phdinitial.pdf

I'm extremely wary of this guy's data part of which come from his "own calculations". Bleh.

I agree. Here are some CS stats http://logos.cs.uic.edu/recruit/csstatistics.htm

I am extremely skeptical of this . . .

Of course, there are plenty of job openings that ask for PhDs, but most of the work does not truly require a PhD. Those jobs are filtering out the vast majority of capable candidates while targeting people who have already shown that they would prefer a different sort of work by doing the PhD in the first place. Is it a surprise that they can't fill the position?

Does that mean that CS PhDs can easily get jobs? Perhaps.

Does it mean they can easily get interesting jobs? I am more doubtful.

Does it mean that producing more PhDs is necessary or even useful? I doubt it.

Canada is handing out citizenship for anybody with a Ph.D right now, so are a lot of other countries. If you don't want to move write books or self publish ebooks, or write articles for industry journals or mailing lists and watch offers come in. This is what a girl I know with a doctorate in anthropology did: write endless articles for some academic publication, she had a job offer in Israel after the first few months because another graduate somewhere was doing research into what she was writing about.

That's an exaggeration for Canada.

As http://www.workpermit.com/canada/points_calculator.htm says, a PhD is +25 points and you need 67 to get in. It helps a lot, but far from a guarantee.

Not even that.. "A two-year university degree at the Bachelor's level AND at least 14 years of full-time study = 20 points"

So just doing a bachelor's will give you 80% of the PhD points..

“For more information on levels of prociciency click here.”

I can only speak to electrical engineering and computer science since that's my field. I've been in the working world for 7 years, and here's what I've found...

First of all, I don't know any good engineers who are unemployed, regardless of their level of EE/CS degree. If you have an EE/CS degree, and you're having trouble securing a paycheck, this might not be the field for you. Sorry.

That said, if "getting a job" is your goal, a BS is all you need. A 5-year BS/MS program is a great deal if your school offers it. Otherwise, I'd look for a job with tuition benefits. I was able to get a Master's part-time at night, and it was completely paid for by my first employer. In my experience, an MS is definitely worth getting, as it will give you a slight salary increase, possibly a better job title, and is probably just expected at good companies.

A PhD in engineering is for teaching or being a research engineer (i.e. at a government/corporate lab or a university). I honestly believe you should only get an engineering PhD if you think (1) you're really smart, and (2) you can study at a top-tier university. If you're not an elite engineer, I don't see the point of a PhD. Work experience will look better on a resume than a thesis with an advisor no one's heard of. And a research job won't pay more, but it will require a bigger brain.

I'm wondering if this same graph can be applied to non-PhD track students, meaning people coming out of STEM programs with a Bachelors, or are you basically getting nothing by even bothering coming out of STEM without a PhD/Masters?

"Employment at graduation" is not a particularly meaningful metric, because STEM PhDs who are not employed at graduation usually just means that they haven't gotten a post-doc. They will take a few months and find a job in the tech industry.

In addition : The end-date of a PhD is indeterminate (at least in the UK), since one doesn't know for sure whether the thesis defense will require any re-writing, etc. So it's difficult to have a job fixed up completely on the day the magic letter arrives.

I chose to travel around Europe for a few months after school and then bothered looking for a job, so I guess by this metric I was unemployed. In the end I left Kansas for a job in London, a much better choice for me than being employed at graduation.

Reading the article and then reading the comments. I start to feel very confused. Do we need a huge debate about this? Some of us just want a higher degree and we just want to go to graduate school to do research. For me, I also love the people I have met in graduate school. It may turn out to a wrong decision? But we will never know that because we could not go back and re-make those decisions. Even if somehow one day I knew for certain that going to grad school was a wrong decision, so what? It was something I wanted to try and I tried it. I think I would regret more if I didn't try something I really wanted...

The article is written from an American perspective where PhDs (5? 6? years) are typically several years longer than, say, in Britain (where they're 3/4 years). So, a question: do you think it's possible to have the best of both worlds? Do a shorter PhD for the intellectual benefit without sacrificing too much opportunity cost in industry?

I've chosen to do a PhD because I enjoy research and it gives me the time to learn a host of other skills too. I should have just turned 25 when I finish, all going well, with a whole bunch of stuff to add to my CV (not just papers, but the more intangible stuff too--going from a shy undergrad to someone who can present at international conferences, foster collaborations between different groups at different institutions, program relatively competently, use very delicate/expensive equipment responsibly etc.) Though it feels worth it to me it's also difficult to get perspective from "within" the PhD, so I'm asking anyone reading this from a more objective point of view: do you think it's all worth it? Or would it have been better to go into industry at the end of my undergrad? Obviously it varies from person to person, but I'm just trying to get a sense for where other people stand on this.

I'm doing a PhD in NZ (also 3-4y). It's nice to hear someone who has similar reasons to mine for doing a PhD. Like you, I'm hoping to get the best of both worlds.

For me, my (engineering) PhD is a chance to take on an entire project. I get to do system-level design, hardware, software, signal processing, user interface, the lot. The thought of spending four years in a job at the bottom of the ladder doing small, well-defined tasks that my manager told me to do didn't really appeal. Instead I get to manage a project, make real decisions, and learn to deal with the consequences of those decisions. I'm not doing it for the sake of any opportunities that having a PhD might bring so much as the opportunities presented by actually doing the PhD.

In Computer Science, at least, outside of teaching, a Ph.D and a masters for that matter is required for only a small percentage of positions, and here's some evidence: http://www.indeed.com/jobtrends?q=%22computer+science%22+and...

Relative graph to assure you that the dip at the end is just due to less data collected recently: http://www.indeed.com/jobtrends?q=%22computer+science%22+and...

I have only would have wanted to apply for one position in my life so far that required a Ph.D, and that company did not end up doing well. I also managed a Ph.D before (from China) and, in my opinion, he was no more productive than others on the team that had neither Masters or Ph.D. That doesn't prove anything, I know. But, although I have a lot of respect for education, I don't think it is worth it typically.

IMHO this numbers must be compared with employment rates of people who haven't got into a PhD. We may discover a similar (or even more significant) downfall. I think that it is not an awful market specifically for young scientists, its an awful market for everybody.

Most jobs, even technical ones, truly do not require a Ph. D. This is just the market sorting things out. Why have zero (or negative) income to get post baccalaureate education if you can go work somewhere (for money) and learn on the job?

maybe cause some things are more important than money?

Physical and intellectual freedom are worth more to me than money.

Who's more likely to have physical and intellectual freedom:

1) A recent PhD with $40,000 to his name

2) A Google engineer who graduated from undergrad at the same time and has $400,000 to her name

The first one.

agreed

How many PhDs have $40,000 to their name, as opposed to $4.00?

How many MTV or NYC Google engineers have $400k to their name after 5 years, as opposed to $4,000 and a bunch of money torched on rent? Google pays well, but not as well as you seem to think.

I'd say that (2) wins, if only because someone who sticks with Google for 5 years is probably one who made Real Googler and now has a fair amount of freedom. Google is pretty nice once you're above the Real Googler Line.

I don't want to veer into minutiae, but I think you underestimate how much someone can save in the amount of time it takes to get a PhD. Say, 5 years on the PhD, at $50k savings/year, gets you to $250k by itself. Add in investment earnings from that, and you top $300k. ($400k was a bit optimistic.)

But that's aside from the actual point. It's not even about what the Googler does after they've done 5 years. That kind of money is enough to do whatever the hell you want to kind of money. You want to read papers all day? You can do it full time and pay yourself a graduate student's stipend for life, without having to worry about pleasing the dictates of your advisor, academic fads, TAing, and internal politics. Even the lucky students who become professors still have to apply for funding and teach students, on top of producing popular research to get tenure: with route 2), you never have to apply for funding, and you can choose how much time you want to devote to teaching and what kind of teaching you want to do.

It's fair to quibble with the actual numbers, but the point is that money buys you a hell of a lot. The fact that someone might not save, or pathologically continue slaving away at the corporate behemoth they hate once they've saved enough, doesn't change that possibility.

But his point, which I do want to reiterate, is that the vast majority of software engineers don't have that kind of saving/investing power. Sure, you might earn $80k/year out of college, but then you pay 1/3 of your income in taxes and another 20% in rent on a one-bedroom apartment and pretty soon you find you've got about $3300/month to pay all your bills, buy food, make student-loan payments, and basically live your life. In my experience, it came to a total saving power of a little less than half what you described.

Now admittedly, $20k/year is a hell of a lot to be able to save for normal Americans! But at the end of five years it leaves you with $100k of savings, not $250k. In the Boston area where I was, you couldn't even put a down-payment on a house with that little money. It was basically just lots and lots of beer money.

On the other hand, if you know a magical company in a magical land, perhaps adorned with pastel ponies, at which I can make enough money with a low enough cost of living to retire after five or ten years to become a self-funded gentleman scientist, I'll certainly take that over graduate school. I just don't think it actually exists, as you can probably tell by the sarcasm.

The answer is 1), but anyways:

1) Most engineers do not work at Google, and 2) I have plenty of friends who do/did, and, guess what?: They still write code all day! Most of which does not do neat things like drive cars, but rather: increases AdWord click-through rates! Or: manages address book contacts. Fun!

I don't care how much free pizza and foosball you offer me, nothing can match the freedom of pursuing my own interests.

Could not have said it better myself.

Agreed, but, I don't see what a PhD has to do with any of these. My snarky, probably very ignorant, reply would be something like, if you want intellectual freedom, get a subscription to a library. Renegotiate your lowly Bachelor of Sciences' contract to work fewer hours, that'll get you physical freedom.

Seriously. IMO someone reading through, say, the TAOCP, Rudin's Real and Complex Analysis, Horowitz & Hill's TAOE, or any selection of good, difficult books like these, will have acquired undoubtedly more intellectual freedom afterwards than any PhD graduate. For starters, it's self-directed study, which is orders of magnitude harder, and more important to self-growth, than study directed by an advisor, or any bit of structure. It's also faster. It doesn't sink you in any significant debt, which is a plus on the order of physical freedom. It's much more eclectic, which is incredibly important, both in terms of the monopoly power you'll bear, and the scientific value your unique analogies between concepts will create. Hopefully.

Sure, there's no paper afterwards to show. My take is that, credentials or not, you'll still have to show your competence to a potential employer or investor, your exposition tailored to the particular needs you suspect they have.

It depends on your subject. Libraries will only take you so far if you need access to a million-dollar microscope or a petaflop supercomputer to get your work done.

So be careful when you say "[reading these books will give you] more intellectual freedom afterwards than any PhD graduate." That's simply not true. And I'm not even touching on the benefits of having a good mentor...

> So be careful when you say "[reading these books will give you] more intellectual freedom afterwards than any PhD graduate." That's simply not true.

You do have a point, but the way I see true "intellectual freedom" (lolcraft may have meant somth similar) is that let's say you spend your late 20s-early 30s slaving away at a post-doc only to have a chance to access a "petaflop supercomputer" and then you suddenly decide that you actually are really interested in classical Persian poetry and you'd like to devote some time to learning Persian, only to realize you do not have time for that because hey, somebody needs to write those papers for the grant money to come in and your chances for tenure to remain intact.

Replace "Persian poetry" with the study of Mathematical logic, the reading of the pre-Socratics or trying to make sense of the early-medieval migration patterns, as things stand right now both the people following PhDs and those too immersed in industry are not "intellectual free" because they have no time for these sorts of intellectual pursuits. We do need to find that middle-ground between extreme-science and industry again, we do need universal people like Leonardo and Democritus back.

I am getting a PhD (with a stipend) and learning on the job.

> Why have zero (or negative) income to get post baccalaureate education if you can go work somewhere (for money) and learn on the job?

Well, ok, so here's the thing: If you apply to grad school in one of the physical or life sciences and they don't offer you a TAship for your first year[1], then they are politely telling you to f*ck off. It's "generally understood" (which means that nobody knows this unless someone tells them) that any serious offer comes with a TAship that will just barely pay your rent in a crappy one-bedroom apartment with a little left over for food.

[1] And make wild promises about how everyone has a full RAship with ponies by his second year

https://news.ycombinator.com/item?id=2813519

practicing what was called "the mushroom theory of management." It was an old expression, used in many other corners of corporate America. The Eclipse Group's managers defined it as follows: "Put 'em in the dark, feed 'em shit, and watch 'em grow."

"For everyone else - look straight at industry" https://en.wikipedia.org/wiki/Mushroom_management

Reply on news.ycombinator.com