People really want those tenure spots at prestigious universities. Or those starting slots on NBA teams. Or CEO roles at exploding tech companies. Or the starring role in the summer blockbuster. Or that #1 slot on the NYTimes best-seller list. And such.

There will always be plum jobs out there that people are willing to compete and even suffer to get. And in situations where there's a ton of demand and not much supply. Well. Things get really unbalanced in favor of the supplier. I'm not sure what targeting academia in particular proves except that, yes, it also happens there. It's yet another situation in life where if you really, really want one of these jobs then you need to learn to realistically assess yourself, the market, and what paths you have to get where you want to go. If there's no clear path to the position you want, you either need to switch gears or prepare for some serious struggle that might never bear fruit. Thoughtfulness and being realistic with oneself are key to staying out of exploitative situations. That's the thing to avoid.

The average person with a college degree has options the drug dealer in Levitt's example probably doesn't have. As do your average Hacker News readers, I'm sure.

Also: I'm an adjunct professor at two universities. But it's a great side-gig and not at all a core part of my income, which makes it great, to be honest. There's a lot of university-related stuff I don't have any interest in. But being an adjunct means I've got real-world experiences that are completely fresh that I can use in the classroom. It's the kind of information that a full-time professor would have a difficult time keeping on top of, I would think.

In fact, I'd argue that business management is closer to a "tournament-like profession" - guys on the bottom of the management scale make nothing compared to even a mediocre Engineer, and get no respect. Folks a the top of that scale laugh at your puny half a million a year.

(Note, I wouldn't call business management a tournament-like profession... just that the difference between the bottom and the top is larger, I think, than it is in academia.)

The other problem here is that okay, you get your PhD in hope of getting a teaching gig. sure, okay, let's accept that. If you join a gang as a foot-solder, you do it to get to the top. Let's accept that, too, for this discussion.

The problem is that if you are a failed gangster, your /best case/ is to end up nearly unemployable and living with your mom.

If you are a "failed" PhD? which is to say, you get your PhD and you fail to get a teaching gig? There are all sorts of other jobs who would love to have you. Yeah, you might not get paid all that much more than others who stopped with a bachelor degree, but more than half the time, that's more money (even if it's not more stability) than a professor can expect to make.

Really, I think teaching higher education is a 'prestige job' more than a 'tournament job'

* Most researchers do dream of making great discoveries, or proving world changing theorems. In this sense I guess I too want prestige. But it has to be because of actual achievement, not just perceived achieved. E.g. fame for the sake of fame.

Who the heck wants to be a University President? I'm frankly surprised John Hennessy finds being Stanford president an interesting use of his time, but maybe I'm just naive.

You've actually tried that? Even in a STEM field? So you are speaking from first hand knowledge?

There is some evidence that outside academics or industrial 'research' really aimed at just publicity or 'luster', a Ph.D. will do somewhat less for a resume than a good felony conviction. Why? Because people without a Ph.D., that is, nearly everyone else, just does not want a Ph.D. around. The people without a Ph.D. commonly make lots of excuses, e.g., a Ph.D. is too bright to be successful in business, cares only about far out pure theory, hates anything practical, is no good at anything practical, always wants to go for big risk, low payoff blue sky stuff, just want to publish papers and make it back into academics, won't work on our practical problems, won't be happy here, etc. There is also the "failed" term; who wants to hire a 'failure'?

In an organization, a lone Ph.D. is in a very poor position, basically feared and/or hated by everyone else. Not good.

I've never shied away from things that are practical. My Ph.D. is essentially in some applied math, that is, for me, the practical and money making kind.

But what you mention is a common remark on Ph.D. degree holders and sometimes true.

I know computer science is a special case, but I haven't completed my PhD yet and I am regularly contacted by recruiters (through my web page or via LinkedIn). Most of the time it's for big data, finance, stealth start-ups, or for Google. I turn them down each time, but there are a lot of them…

Me? no, have you? I don't have a degree at all. I've worked with a few PhDs, though, and I have several family members in the 'failed academic' class who still putter about teaching a class here or there but mostly work in industry.

I don't think they could be classified as failures by any reasonable measure, well, except for the failure to make tenure.

I mean, if you measure success by job stability, then yeah, academics are super rich, and working in industry is a failure... but that's not how I see things.

More to the point, if it was really that big of a deal? you could just leave it off your resume. It's certainly better than a felony in that regard.

Yes. Been there. Done that. Have the T-shirt.

> More to the point, if it was really that big of a deal? you could just leave it off your resume. It's certainly better than a felony in that regard.

A time gap on a resume is about as bad as a felony conviction, really, for looking for a job, worse. A period as a prof, in a slot that would require a Ph.D., does say "I've got a Ph.D.". If leave off the prof slot, then that is another gap on the resume. Such gaps are resume fraud and into the circular file, that is, as bad as a felony conviction. Actually, a felony conviction is likely easier to hide during a job application than a Ph.D. No, I don't have a felony conviction. Indeed have never been arrested or convicted of anything except minor traffic violations.

What I said is true: A Ph.D. in a STEM field (maybe no one would resent a Ph.D. in art history) outside of academics or some industrial research lab concentrating on publicity or luster is a job killer.

At one point, with a Ph.D., I sent 1000 resume copies and got back only silence or laughs. Earlier in my career, before my Ph.D., once I sent a few resumes and in two weeks went on 7 interviews and got 5 offers.

Yes, one of my dissertation advisors is President of one of Stanford, MIT, CMU, Cal-Tech, or Georgia Tech; no help. The Chair of my oral exam is a Member, US National Academy of Engineering. No help. Math SAT 768. GRE Math 800. No help.

One reason I'm an entrepreneur is that due to my Ph.D., in engineering from some applied math from a famous, world class research university, I'm unemployable at anything better than stocking shelves at Wal-Mart. Literally. No joke. Period.

'Data mining?' 'Big data'. 'Machine learning'? I have some of the very best background in such topics. Still, just unemployable. Period. Computing? One of the best backgrounds, long at IBM Research where I won an award for my software, algorithm, and architecture contribution to IBM's leading artificial intelligence language. Published in computer science, mathematical statistics, and artificial intelligence. Long background in practical software. Right: Not in C++, Ruby, or Python. So, unemployable.

So I'm doing a startup. There my Ph.D. is crucial -- the core of the work is some secret sauce I did from some original research I did that very much needed my Ph.D.

Net, it's much easier for me to be the CEO of my own company, and likely get a few million in venture funding, and be wealthy if the company does much of anything, than just to get a job that will let me buy a house. Literally. In case my business is successful, mayhe Ph.D. will have paid off, but not for a job. "Paid off" needs some grains of salt since the delays cost me house, wife, family, etc. Right: Thanksgiving was eating alone. No, I won't be sending any children to college, much less a Ph.D.

Actually early in my career a Ph.D. seemed like it would help for non-academic jobs, but that was a boom time around DC with work heavily for the US DoD. Once the Cold War was over, that work went away.

A good Ph.D. in a STEM field sets one apart in a way that means that they might be able to get a job stocking shelves at Wal-Mart. Sorry 'bout that.

A hypothesis: I think you walked into industry interviews acting as if you were hot shit and didn't successfully demonstrate to employers that you could be one of them. I.e., "I know the leading artificial intelligence language, I don't need to waste my time with Python."

(Please don't be offended if this is wrong, it's only a guess based on the impression given by one HN comment.)

For others in this situation, I wrote some tips which I've been told are helpful: http://www.chrisstucchio.com/blog/2012/leaving_academia.html

There is a book, now old, E. O. Thorpe, "Beat the Market" where at the end he said that an argument was from measure theory. So, in graduate school, I made sure to get a good background in probability and stochastic processes based on measure theory, from a star student of E. Cinlar, long the main guy in financial math at Princeton. Yup, I was the only guy in the class who was able to show that there are no countably infinite sigma algebras!

No help.

Once I wrote a letter to Fisher Black (of Black-Scholes) at Goldman Sachs and got back a nice answer saying that he saw no opportunities for applied math on Wall Street.

I live 70 miles north of Wall Street and did interview at Morgan Stanley. Mostly they wanted me just for software, but I tried to indicate an interest in mathematically driven automated trading, but they had no such interests.

Once a guy who recruited for Goldman Sachs took me to lunch, but there was no interest.

I learned about James Simons only much later.

Heck, my Ph.D. dissertation was in stochastic optimal control -- still no interest.

Huh. I've found that I do much better at interviews when I try to act as "hot shit" as I possibly can. Of course, it's possible that me "turning it up to 11" just brings me into the acceptable range because of where I start. but still, I can not comprehend a level of arrogance that would get you shot down in a job interview.

It must depend on the company. Where I work, we have a "no assholes" policy and acting like one would make it very very hard to get past the interviewers.

Moreover they took me in spite of my Ph.D., and I didn't even know any programming languages they used. There is definitely more to this than the number of trendy bullet points on your resume.

Everyone says that. Even people who end up hiring me over my shorter and less loud compatriots. Maybe I just don't make it to "asshole"? but there is a huge advantage to being tall and loud over being short and quiet.

It depends on the field. For STEM jobs, a PhD may not justify the time investment, but it's better than nothing. In non-STEM fields, a PhD can actually be a disadvantage.

Wait, what? I mean, you can argue that they are really bad and maybe you are right; I disagree, but maybe I am wrong.

But how is a blank spot fraud? Hell, I leave off most of my short jobs, I've done a fair bit of contracting, so there's a lot of that, and I've run businesses on the side for most of my life, so there is a fair bit of overlap, but even so, I don't see blank spots as dishonest (even if they may look bad)

>Net, it's much easier for me to be the CEO of my own company, and likely get a few million in venture funding

Hey, good luck. But until it's in the bag, guessing what is in the mind of VC is... not something I'd bet money on. Of course, I have gotten exactly zero dollars of VC, so what do I know? But my impression was that they were even more irrational than your average hiring manager, which is saying something.

Of course, you don't always need VC to run a company.

>A good Ph.D. in a STEM field sets one apart in a way that means that they might be able to get a job stocking shelves at Wal-Mart. Sorry 'bout that.

Huh. See, last time I tried to get a job of that type, I got turned down because I had serious experience (even though my level of official education was appropriate.) - It was actually easier for me to get another well-paid SysAdmin job than to get a job at wal-mart or the coffee shop, even at the depth of the dot-com crash in 2001.

(It was... actually pretty funny when I applied at the coffee shop. It was very clear that the hiring manager thought I was not remotely cool enough to work there.)

Interestingly, most of the PhDs I've worked with were physics majors.

I think that physics students aren't always "monogamously" in love with physics, so much as we're intellectual omnivores / opportunists. I was in love with math, electronics, and programming. But I ended up at a college with no engineering school, and the people who seemed to have the most fun playing with toys were my physics profs.

But as a result, my ability to roll circuits and sling code with the best of 'em was still marketable in 1993. Today, I see a trend towards hiring people with more formal skills in their job areas, such as CS and EE degrees for programming and electronics design. And I see more emphasis on tools (IDE, CAD, etc.) than on knowledge of fundamentals.

As an aside, it's worth noting that the public image of physics is the Theoretician, but most physics students are Experimentalists.

That, and they tend to be bright sorts who are able to break down problems into something that can be reduced to code. My impression is that the standards as far as intelligence goes are higher in the physics department than in the CS department, but that might just be that where I work, I don't get to see the best output of the CS department.

I need to get some code written and typed too quickly, but basically I'm correct. So, need to fill in dates on a resume. So, if have years in grad school and some more years in a Ph.D.-only slot as a prof, then how to account for that time on the resume? If put in the truth, then won't get hired. If just leave a gap, then won't get hired. If fill in the gap with the jobs before and/or after the gap or something else, then that's resume fraud. Likely if just omit the Ph.D. in the section on "education", then that's resume fraud much the same as claiming a degree don't actually have.

> But until it's in the bag, guessing what is in the mind of VC is... not something I'd bet money on.

It took me a while, but eventually I got a reasonably good view of what VCs want.

For how I learned to do projects for the US DoD and in grad school, the VCs know essentially nothing: E.g., at one time I was programming in the group that did the orbit determination for the Navy's version of the later USAF GPS (my work was in processing data for passive sonar), and I heard the stories for how the Navy's system was developed. Heavily the plan was a dressed up version of something on the back of an envelope. The work got funded, and has been a great success, from a proposal just on paper. Good for the Navy and US national security.

But VCs essentially won't invest in projects described just on paper if only because they don't even want to read the paper -- it may be possible to get some VC seed funding just from paper, but that seems to be recent. NSF and DARPA will read project designs on paper, but the VCs won't.

What the VCs want for a Series A round is essentially in one word, traction. Or traction significant and growing rapidly.

At one point Menlo Ventures wrote me that the level of traction they want to see is at least 100,000 uniques a month, that is, for a Web site project 100,000 unique visitors. Okay, consider 100,000 uniques, on average: Considering what my site does, 100,000 uniques would mean about 500,000 visits; each visit sees 6 Web pages, with 4 ads per page; and I get paid, in effect, $2 per 1000 ads displayed (drawing from the Mary Meeker data at KPCB). Then my monthly revenue would be

500,000 * 6 * 4 * 2 / 1000 = 24,000

dollars at which time no thanks seed or series A and welcome some new computers and a new car! If the site keeps growing, and once it gets to 100,000 uniques it should, start hiring and grow just organically.

> Of course, you don't always need VC to run a company.

Yup. My first server might be an 8 core AMD processor at 4.0 GHz with maybe 32 GB of ECC main memory and 12 TB of hard disk in a mid-tower case for about $1500 and software, Windows Server and SQL Server, from the Microsoft BizSpark program. From my software timings, that server might be enough for the $24,000 a month in revenue at which time I can get a $100 wire rack from Sam's Club, put maybe 12 such servers on it.

My software is written for such scalability now -- e.g., so far I'm using SQL Server essentially only as a key-value store so that each SQL Server table (there are only a few tables in the SQL Server schema) can be in its own running instance of SQL Server with no communications among the instances; for the various servers, there can be a lot of parallelism via sharding, etc.

E.g., I wrote my own session state store; the timings indicate that on a single core running at 4.0 GHz I should be able to do the work needed to send 2000 Web pages a second; such a store is easy to parallelize (partition the collection of possible keys and have one server for each partition); and a few racks of such servers could do enough for session state to serve the world and build a big company.

For a seed round, that now can be as high as a few million dollars and no longer limited just to, say, $200,000, the VCs want to see the running software and "play with it".

So, soon, maybe later today, more likely before next Sunday, my software should be ready to be played with over the Internet.

This software development has taken a while as I learned enough about essentially system administration and Microsoft's .NET Framework to get good with such infrastructure. For the .NET Framework, I found, downloaded, indexed, and studied about 5000 Web pages from Microsoft's MSDN Web site. Yes, my code is awash in comments with tree names on my file system of such Web pages. So, when I'm reviewing code and have a question, e.g., a TCP/IP timeout parameter, one keystroke in my favorite editor pops up the relevant MSDN Web page. In the end, I have found the .NET Framework to be nice, but the 5000 pages took a while.

I actually studied all those 5000 pages; at 10 pages a day, that's 500 days just for the MSDN Web pages, and in the end I spent more than that.

Heck, I did the research for my Ph.D. dissertation independently in my first summer in graduate school, about 12 weeks, did a second pass later in about 2 months, wrote and ran some illustrative software in about 2 months, and did the final writing and typing in about six weeks. Those 5000 Web pages from MSDN mean that at times I have been known to scream to the point of a sore throat at what I regard as some really poor technical writing at MSDN.

In the end, once understood, the .NET Framework is nice and not very difficult to use, but going from the first day with .NET, through 5000 Web pages and maybe a cubic foot of books to some good facility can be a bit much.

My whole project is now a surprisingly large number of lines of code (maybe I will add them up tomorrow), but the work unique to my project was easy once the system administration independent of my project was done. E.g., SQL Server installation was an unanesthetized root-canal procedure (once wiped out my boot partition). My software installation got wiped out by a virus in an Akamai download manager I used to get a PDF file from an Asus Web site. There were more such problems, all independent of my work.

For my project, I created and derived the basic core applied math in a few days and typed it into Knuth's TeX in a few more days; the software unique to my project has been easy for me; but the routine software part has been grim.

The last thing I did, as of last night, was finish the code to have the code of a Web page communicate with a server, in my architecture, a Compute Server, for the core work. So, it's a simple TCP/IP socket application. .NET streams over TCP/IP might have been easier; maybe I rewrote some of that functionality.

I avoided such streams because the MSDN documentation seemed too sparse to give me good confidence in what I really might encounter and have to consider, and I didn't want to write and run test cases to clarify just what such streams would and wouldn't do. For what I did write, based just on the basics of TCP/IP sockets, "the workhorse of the Internet", that I've had experience with for a long time, there was little question about how to write the software or what it would do. So, it was easier just to stand on basic TCP/IP than to make sense of Microsoft's documentation of the streams they wrote to use TCP/IP.

I put the data to be sent to/from the server code for a Web page and one of my compute servers in an instance of a class, serialize the instance to a byte array, and send the byte array via TCP/IP.

> It was actually easier for me to get another well-paid SysAdmin job than to get a job at wal-mart or the coffee shop, even at the depth of the dot-com crash in 2001.

Wow, but I can believe it.

That is another problem that can be solved by money and wise hiring decisions; I mean, I wouldn't touch a windows installation with a ten-foot pole, but there are plenty of other SysAdmins who specialize in that sort of thing.

Why did you choose a windows stack instead of a Linux stack?

My reasons are likely very different from yours, as I chose systems administration as my specialty. I chose to go into UNIX (BSD before Linux) because it seemed so much easier to reproduce problems on UNIX. When I solved problems on Windows, it seems it was a process of randomly changing things until it worked; on Linux or BSD, I could almost always point to an actual problem, and explain why. I felt like I was actually learning things. I ended up staying with Linux, in part, because it seems that Linux sysadmins get better pay and respect compared to our compatriots managing windows installations.

Always a big question. I went from IBM mainframes to Prime supermini computers, to more IBM mainframes, to PC/DOS, and OS/2. Then to Windows was natural enough.

Some of the work by Microsoft is good.

Microsoft has a lot of customers doing important computing which means that it's possible to make their stuff work. There are a lot of programmers posting problem fixes on-line.

I don't know enough about Linux and Windows to make a good comparison. In the end I trusted Microsoft's big customers to make Microsoft make their more serious software actually work. Yes, once I get some revenue, for questions I will get a Microsoft 'account executive', expect him to make things 'right', and call for expert paid support when things don't work.

If you put in a job or experience you didn't have, or lie about the dates, sure.

>Likely if just omit the Ph.D. in the section on "education", then that's resume fraud much the same as claiming a degree don't actually have.

This is the part I take issue with. Me? I leave off the "education" section entirely, as a high school diploma from a not very good high school with a poor GPA isn't going to help me any. It takes up space in that valuable first two pages, and doesn't convey anything that helps me.

But still, you /did/ get your BS, right? Putting that on your resume is no lie. if you put your BS on your resume, then have a five-year gap before your next job, I don't see how that's fraud, I mean if it's true that the PhD really isn't relevant to the job, for all the boss cares, you could have been making paper airplanes for that time, or reading about Napoleon.

From what I've seen? resume gaps are /terrible/ if they are current; e.g. if you have been unemployed for the last five years? yeah, you're gonna have to work for someone like me for a few years.

But if you were unemployed for five years, but were employed in the field for five years after that, and now you are looking for a new job? the big blank spot from way back when is not a big deal anymore, because the more recent five years shows that you are able to do the job.

In 1993-4-5, sent 1000+ resume copies, just simple, perfectly formatted, factually written, perfect writing and spelling, heard back only a very few times, got only a few interviews and no offers. I did nothing, but Nothing wrong. Nothing. I sent resumes to tech headhunters -- nothing.

All I wanted was a job, just a job, only a job. I would have swept floors. I just wanted a job. Did I mention that all I wanted was just a job? I didn't want to be unemployed. I did want to be able to pay for food, clothing, shelter, etc. I didn't want to dig into my savings and inheritance. I just, just, just wanted a job. But, I was 100%, absolutely, positively, totally, permanently unemployable at anything above minimum wage. That was 20 years ago. Now I am doing a startup. I've been a serious burden to my brother's family. All I wanted was a job. I could never get one.

My conclusion: Even in a STEM field, a Ph.D. is resume poison, worse than a felony conviction. You are just not wanted. It's as if you are hated.

Again, the problem was nothing about my personality, etc. E.g., before my Ph.D. in a two week period I sent a few resume copies, went on 7 interviews, and got 5 offers. No problem. After my Ph.D.? Stuck in academics or something very close and otherwise hated.

You expect the fact that you got an 800 on the GRE math to impress someone? Congratulations, you are good at junior high level math.

I have a PhD in CS and have never had trouble getting jobs before, during, or after grad school. 90% of my labmates went into industry rather than academia and they all had job offers before they even defended.

You seem to expect people you spam your resume to to hire you based on your name dropping. If you're as talented as you seem to think you are, you should be able to learn python and/or ruby in a fraction of the time it takes you to send out 1000 resumes.

Your problem isn't your PhD it's your attitude.

One of my contributions was architecture, algorithms, and code for rule-subroutines. The core of the work was to understand the PL/I stack of dynamic descendancy and to exploit it to keep the state of the user's rule subroutine calls that had yet to return while calling a run-time package with the RETE algorithm and then return to the user's code with it's state intact. It was a fast idea one afternoon and then a long night typing. When I got some sleep and back to the office, our programmer trying to finish the implementation of rule subroutines was done that day instead of some weeks, and the result was much cleaner.

In addition, I cooked up some new ideas for anomaly detection in data from the envisioned server farms and networks. It's an ambitious approach to behavioral monitoring (try to say what is normal and declare anything else as an anomaly and potential problem to be investigated -- as in the old Herve Debar post at SANS, such monitoring has had false alarm rate too high, and I solved that problem) where a user gets a knob to turn to select false alarm rate exactly in advance.

The work was ambitious when I did it because it can use quite a lot of data; now the cost of so much data is so low that it is from trivial down to no concern.

As applied probability, it's a statistical hypothesis test that is both multivariate (use more information and, thus, get higher detection rate of real problems) and distribution-free (make no assumptions about probability distributions) -- maybe still the only example of a test with these two properties.

The paper was published in Information Sciences. I had some data from a cluster at Allstate and showed with that data that the false alarm rate calculation was correct. And with some additional simulated data, showed that the detector was darned sensitive (intuitively) to anything different. Basically could have a checkerboard with normal performance on the white squares and anomalies on the black squares, and the detector will figure this out, including if the checkerboard is in 20 dimensions. Due to such checkerboard examples, data in fewer dimensions would be useless; so, in this case, the multidimensional aspect was crucial. While such a checkerboard is unlikely in practice, the practical point is that the detector is adaptive to quite bizarre normal performance. Indeed, we should expect that the multidimensional geometric region of normal performance will, often in practice, be essentially a fractal -- my work does well approximating fractals. Now that there is so much data on real-time performance from high end server farms and networks, my work should be deployed, but I doubt that it has been.

Exactly. The culture is 'presitige maximization', and it's ruthlessly pursued. TLDR: Man is a political animal

That might be true for some fields but not all. I am looking for a teaching gig with my phd and coming up with nothing. At this point I would love to find one of these other jobs that would love to hire me.

what is your PhD in? I thought that all physics academics eventually ended up programming. I mean, i guess if you are an English major or an art major, that would seem a bit like plumbing, but eh, it pays the bills, and it's something that most smart people can figure out how to do.

Amusingly, from what I've seen, fresh physics majors have more programming experience than fresh computer science majors. And way more linux/large scale systems experience. Most of them even know how to use basic tools like ssh; something you can't say of many CS grads.

I've really got no real experience outside of the computer industry, so for all I know, everything everyone says about a PhD being a huge black mark is true in other industries, but it sure seems pretty unlikely.

Seems like you could do an experiment, much like the 'Mr. Kim'[1] guy did about his name and gender; leave the PhD off your resume and see what happens.

[1]http://whatwouldkingleonidasdo.tumblr.com/post/54989171152/h...

I think this is the crux of my point. Generally speaking? if you have a PhD you are pretty bright, and come from an upper-middle class family, and have a certain level of togetherness and work ethic.

You are gonna do okay.

And of course, many enjoy teaching.

It's as stupid as playing the lottery. In academia people should just rebel and ask for one of two things:

1. higher salaries (like significantly higher than the average income that someone with a similar qualification would get in industry!), OR

2. tenured positions with lots of "free" time to do research and the security that even if they pursue unpopular lines of research and get negative research they'd still get to keep their tenure.

People should not accept working for the "hope of much better status/income" in the future, they should aggressively and unreasonably demand IMMEDIATE compensation for their struggles, not work for expectation of a "better future". There are enough money and resources in the system for almost everybody to "have their cake and eat it too" and you shouldn't settle for anything less!

Now if people would start thinking this way, worldwide academic institutions would collapse, but I think this would be a good outcome as, imho, despite their contributions to mankind's progress, educational systems and academic institutions are mostly unreformable and can only be rebuild from scratch for the better.

Umm... no. Even for the very best people, at the very best places, salaries tend to top off below $150K. In places with high costs of living, this isn't a cushy life.

AFTER EDIT: Of course, I was specifically disagreeing with the specific dollar figure ("$150K") in the comment immediately above mine. The grandparent comment, referring to "unlimited money," is of course using the literary device of hyperbole, which academics in the disciplines I studied in my higher education know better than to take literally. An academic who writes a widely adopted textbook can make serious money in quite a few different disciplines, and academic employment in general offers more opportunity than most forms of employment to take on side jobs.

Edit: If you're measuring or living off of a salary in general, without vast outside income from investments or the like, you arn't even close to "unlimited money".

This year I decided to secretly do some consulting/contracting work on the side (for various reasons). It was a really difficult year as I worked like a dog but I was able to make an additional US$175K in income.

It's strange, I've felt more 'fulfilled' this year than simply spending that time on writing research publications. I feel like I've actually helped people and worked on stuff that matters...

Academia is broken IMHO: Too much focus on volume instead of quality, MOOCs and online courses are breaking teaching, most students don't care about the content, 50% of university staff are administrators.

I plan to make the most of it though (while it lasts).

I speculate (although I can't really know, most salaries are private) that there are less than 100 academic mathematicians making > 200K. This, in a field that's famous for attracting brilliant, hard-working people.

Where do you see a Breaking Bad picture on this page? The comparison they make to drug gangs is fair and the gangs themselves are _not_ the root of the article. It's an article with statistics and cited sources, it seems reasonable for them to make an interesting introduction.

Now, in business you can get close to "the top" by riding on investors' money, but someone still has to make profit on every step of the way, only that "every step of the way" may be more like "the aggregated profit of all companies fueled by one investment fund in a period of time" or something like that.

To be sure, a tenured professor's salary is comfortable (unless you are in NYC, SFO, Tokyo, London, Paris, etc), but it is also has a definitive ceiling. It's not like you can work double the hours and get double the salary. Even if you generate a great discovery, the salary increment, if any, is comparably modest --- orders of magnitude smaller than for example the bonuses one gets on Wall Street for a successful quarter of sales of shady investments to unsuspecting investors.

As an illustration: A colleague of mine who's at the top of his field just negotiated a raise, (using a competing offer from an Ivy League school), and do you know what his raise was? ... He got an additional $15,000 per year to his salary of around $140k. After taxes he'll get around $600 or $700 per month extra. Now tell me again about the riches of being a tenured prof? If he was working in the private sector (Wall Street, or consulting, or medicine, which he could easily do given his skill set) he would be making integer multiples of this salary.

Also on the whole we should take into account that professors didn't have to pay for grad school unlike medicine / business and instead got paid for it.

The Robert Taylor Homes were poor, isolated, and cut off from the rest of Chicago by a highway. The young men who joined the gang didn't have many other options. Most of the men who lived there worked ad hoc jobs. Sudhir Venkatesh coached young drug dealers for interviews (ex: to become a janitor).

I suspect that low-level drug dealers start dealing because it's easy for them to get into and they don't have better options. The gangs probably hold more appeal as a social organization than they do as a way to make money.

Doing a PhD is fun and interesting, and that's the main reason to go with it.

The pay during the PhD is indeed not good compared to what we could make by working in the industry, but it's not that bad: we certainly make more than we would by working at MacDonald's (I don't think it is a relevant comparison but that's the example at the beginning of the article).

We do a PhD because we enjoy our life with it right now, and because we want to do research (and possibly teaching at uni level) as our job later, without thinking about the salary (some PhD students don't even really know what are the salaries in academia, only that it's not as much as in the industry).

The position of a PhD student have nothing that is comparable to the drug dealing field as described in the article. It's sometimes hard, but it's not risky/dangerous and you're not exploited by your hierarchy, or have to fear it in any way.

[EDIT to clarify my point] There's is no trade-off such as "I'll take risks and live poorly for now in the hope of becoming rich and famous later" when you choose to do a PhD. If you get to do a PhD, you already have what you want, which is freedom to work on the subject that interests you, very flexible work-hours, the possibility of teaching if you want to (and have some more money for that, which is a nice plus but I guess most PhD students who teach would do it anyway). [/EDIT]

I recently wrote a blogpost which did not get any traction here [1], in which I present a quick overview of my PhD work until now in the hope of showing how much I'm enjoying myself as a PhD student :-), and this in order to try to balance a bit the tendency that HN has to present PhD as a waste of time or generally with negative connotations.

[1] https://news.ycombinator.com/item?id=6755727

Although I enjoyed many parts of my PhD experience, and the subsequent Postdoc, I rarely harboured (after my first year of grad school when I saw how competitive faculty positions are) the belief that my PhD (from a competitive school in a hot area) would allow me to work on anything I wanted for a reasonable salary.

I will say this: PhDs go a long way to maximizing freedom, but they are not a panacea. Positions for research are typically quite limited, and it's more likely, you'll end up a terminal "Research Associate" (little pay or autonomy) or working in an Industrial Lab (which has its own plusses and minuses).

Real freedom comes from being the top 1%.

Concerning the pay, I question what you call "little". Salaries in public research institutes in France may be little compared to what you could make at a good position in the industry and given your diplomas, but it's certainly not a bad salary in the sense that you can live off of it with reasonable comfort.

The research of Paula Stephan covers the economic decisions made by US science research institutions[1]. Her hypothesis is that academia maximizes PhD student numbers while the number of tenure-track faculty positions decline. This presents an untenable situation in the long term, where a flood of researchers with PhDs will be unable to find permanent employment, and yet the demand for cheap labor filled by PhD students intensifies.

As you can imagine, this type of trend can be terrifying to some PhD students, despite the "freedom" that their research allows. A lack of job security makes one question their salary as an extension of their worth.

[1] Her latest book is an exhaustive look at the subject: http://www.amazon.com/Economics-Shapes-Science-Paula-Stephan...

What "corresponding age cohort?

E.g., in, say, the OECD countries, the Ph.D. degree holders who got their Ph.D. in year 2000 make up about 1% of the population, and the Ph.D. degree holders who got their Ph.D. in year 2011 make up about 2% of the population?

It's a bit tough to read the vertical axis labels on the right. If save the graph as a file to disk, unfortunately a JPG, good for landscapes, instead of a lossless file type such as GIF or PNG, and magnify the image, then can read the labels on the right axis. Alas, find that the labels are 0, 1, 1, 2, 2, 3, 3, 4, 4. I have team of researchers working on what those axis labels mean and will post here again when they submit their solution!

But yes, that figure (all the figures in the article, really) could have used a bit of editorial guidance.

[0] vague!

[1] other people "of an age to be getting PhDs"? also vague!

i.e. see every third world country.

And while in some fields there's the need to "import people" (IT basically), there's a chronic lack of jobs in some other fields. Source: spent 2 years working in Europe.

Academia has those positions, maybe not enough to go around, but they're there. I'm not sure that drug dealing does.

An increasing number of PhD graduates,sure... But not an increasing number of brilliant PhD graduates.

There are market forces (e.g. Government incentives) for universities to admit more students to PhD programs and my experience is not that the additional students are brilliant. They tend to be people who complete the PhD (or just a terminal masters) and exit academia quite definitively

The most important factors for your future academic career are intelligence, mentoring, good project choice, good self-management and always some luck with outcome. Fail in two of these and you don't have much of a chance to get your academic carrer back on track.

IMO failing at least two of these factors are usually the reasons for people to leave academia that are subsumed under "lack of success". There are also many students that are successfull and leave for the money or because they are female and want to have childern before 35.

Then again, I've never been at a "top tier" university, so maybe things work different in Stanford.

http://www.thisamericanlife.org/radio-archives/episode/487/h...

The truth is that the world's full of suckers who can be paid in promises rather than actual rewards. So many people will believe anything they are told.

Academics and startup engineers have less of an excuse than drug pawns, though. The soldiers have felony convictions (often) and generational poverty issues against them, putting regular employment out of the question. Academics and startup engineers are privileged people who refuse to fight for themselves; most of them exhibit the very process by which wealth declines-- a refusal to look out for one's own interests.