Are PhD Students Irrational?
lareviewofbooks.org
lareviewofbooks.org
"If you actually want to know why people persist in going to grad school, or trying to be screenwriters or musicians or professional athletes or actors — or, if we’re getting really real about it, journalists and writers, the kind that write for The Atlantic — it’s because our culture insists relentlessly that certain kinds of work have intellectual and aesthetic value, that they are a way to be a Somebody, and also that you never, ever give up on your dreams. There is no message that we deliver to children and adolescents more relentlessly than that they should pursue their dreams with manic focus and unflagging persistence, no matter how hard and often they fail, and that they will be assured of eventual victory. Never, never, never give up. Whether you think you can, or you think you can’t, you’re right! If you believe it, you can achieve it. Go to a middle school sometime. Look at the posters on the wall. It’s like a one-party state, propaganda of the most ubiquitous and intense variety. When I was a long-term sub at my local junior high I was amazed; it was like some sort of totalitarian reeducation camp, with the purpose being to indoctrinate in all who passed through the doors that failure stems always and only from a lack of nerve."
Amazing how few people writing about the issue see, or at least note, this basic insight.
The value of a PhD is not high because the life of a PhD is desirable. Academia provides a (flawed) society with a permanence and thoughtfulness that competes with non academic america. Leisure and camaraderie is everywhere. It provides a sense of identity which is hard to find, so supply is high and price is low.
Within the cult there are bad actors who misrepresent their value, and selfishly collect rewards. Nothings perfect.
On the other hand, a PhD is way too long. And if you stop at year 4 or so you have basically wasted 4 years. I did not appreciate how much my values would change over time and sort of got locked into trying to finish my PhD. Trying to finish a PhD when you don't have passion for the subject anymore is very stressful. This experience is disturbingly common.
Finally, the author early on implies that the job market for STEM PhDs is not good. At least in physics, this is only true in academia. I know many people who have transitioned successfully to data science, or who have gotten a post doc. Overall, the unemployment rate of physics PhDs is low.
STEM PhDs are more in demand these days, but some still don't know it and need encouragement (and sometimes a little help) transitioning to industry.
A mutual friend had referred him to me for advice on how to transition to industry. I gave him a pep-talk and explained how to build a portfolio to show skills that matter in the workplace. I think he turned things around, but for a little while he couldn't stop thinking about the "wasted" years. PhD, postdoc(s), 3-5 years of assistantship, repeated once, ... it's a tragedy for such a smart, capable person to succeed so much and yet feel like they failed.
Let us say the typical software engineer's day is filled up with the following types of tasks:
1. What patio11 describes, although in a different context: "Don't try to make a career out of optimizing the SQL queries to display a preference page on a line of business app at a company that no one has ever heard of." [1]
2. Some kind of algorithmic work (e.g. writing a compiler)
3. Big data, machine learning etc. (consuming the results of algorithms, hence different from 2)
4. Software architecting
My view is that after a while, most people want to move from group 1 to one of the other groups. This gives a good way to explain the pursuit of Ph.D. even when it is not economically rational (both in time and money cost) - it is a pursuit of something which is not mundane as long as you make the effort.
A teacher of mine said: "You are going to be spending about 40 years in your career. If you take 5 out of that to do a Ph.D. you are not going to reflect on it with regret. And at the end of it you have a Ph.D. too."
When you combine this with the possibility that you will be mostly hanging out with elastic minded students as PG once put it and you might actually have the time of your (intellectual) life, the decision doesn't look all that irrational.
[1] https://training.kalzumeus.com/newsletters/archive/do-not-en...
> for a certain group of college grads
The Adjunct scam is really, truly atrocious. Adjucts and grad students can unionize now, though, and it only takes one strike in the e. coli labs for a research university to cave...
However, PhDs do have valuable (and financially rewarding) roles in nonprofit research institutions that are not subject to the same challenges as universities (though I sure hope you like writing grants),
In "real" industry positions, a PhD can be a way to bulwark yourself against "the engineer trap", where you get promoted into a project management role and become unable to actually do anything except munge Excel documents. It is a gamble.
I have an engineering background, so I can't really speak to what it is like to, e.g., have a PhD in English. But, depending on your degree, there are also really cool gigs at places like microsoft research -- they do everything from cell biology to audio, now, as well as employ "pure theory" CS folks.
Grad students have been forming unions since the '70s, and it hasn't had any impact. There are so few tenure track positions you can't afford to even slightly irritate the people you need to get one, and everyone views their stint as a TA or an adjunct as temporary.
I think "rationality" is a stupid assumption, often wrong. Of course, people need more money, but a lot of people are also willing to make sacrifices for something else - spouses, children, parents, religion or country. Money is important, but not the overriding concern for everyone. I often feels that the correct word for a "rational" individual in the sense of economics is "sociopath".
Ralph Nader once said about his organization, that "you can bring your conscience to work" every day. This actually counts for a lot.
However the basic point of the article, that the Universities are conning students is perfectly valid (source: ex-PhD student, can relate to the frustration). This point needs to be better made, avoiding slinging mud on the intellect of the hapless students. This article is an example of how you can write a dubious article, even in the presence of good data.
If you can't, the it's bad for obvious reasons.
As you point out, there are plenty of non-monetary benefits to academia: passion, interest, fulfillment, pride, etc. These things have different value to different people depending on their own interests and goals. If the total benefit (tangible and intangible) is worth more to someone than the opportunity cost, then, according to economics, that person is still acting rationally in such pursuits.
I'm not an economist, but this is seriously undergraduate-level Introductory Economics material.
Payscale has an interesting ranking of graduate degrees by program, which is more useful than raking all holders of a particular graduate degree together.
http://www.payscale.com/college-salary-report/grad?page=49
Unfortunately, it doesn't break out MS and PhD holders by subject studied. So while you do get to see a specific ranking for an MBA or JD holder from UCLA or MIT, you only get overall salary info for PhDs from MIT, not PhD in Computer Science from MIT vs Electrical Engineering from Berkeley. That would be far more useful.
As it stands, a PhD doesn't show up until spot 29 on this list, but then again, a PhD in CS might, so hard to say.
To me, the attrition rate from PhD programs is an under appreciated aspect of this discussion. A lot of people from elite Law, MBA, and MD programs are floored when they hear the attrition rate from PhD programs. Seriously, the attrition rate form an elite law or med school tends to be well below one half of one percent. Attrition rates form elite engineering and science PhD programs range from 35%-50%.
In any case, I'm always glad to see this discussed. The only people who seem to think there is a "shortage" or STEM graduate students are people who have a financial interest in hiring STEM graduate students. Almost every other analysis concludes that people with the freedom to choose their career in the US (free of visa restrictions that limit their career choices) are largely acting rationally by pursuing other graduate degree programs (or no graduate degree program).
Many PhD programs offer full tuition, health care, and a yearly stipend with teaching duties. OTOH, there may be less incentive to give someone the boot if they're bringing in large sums of money by paying their way through, which is how most master's and professional degrees work (as I understand).
It's a tough comparison of course, in that the degree itself is a very different beast. Half of a med or law degree doesn't help you, a strong student with a BA in History can't skip law school and just go be a lawyer, whereas anyone who can get into MIT's CS PhD program is already almost certainly capable of getting a good paying job. So attrition does mean different things.
That said, we are a nation that officially (by that, I mean our top elected officials, most of whom are lawyers) fret loudly and constantly, often with a tech CEO at their side, about the shortage of US citizens getting graduate degrees, especially PhDs, in STEM fields. I have never heard one of them ask why a 50% attrition rate in some programs isn't part of the problem. I think this is because, as lawyers, they think that elite PhD programs in STEM are kind of sort of like the elite law schools they attended.
There is legislation, for instance, to "staple a green card" to all PhD degrees in STEM fields. I can't argue that these talented people should be excluded from the US, nor would I want to. But if we're so short on PhD holders in STEM, to the point where we need to empower US universities to bestow green cards upon STEM graduate degree recipients (but not other fields, it has to be STEM)... shouldn't we at least ask them why they're throwing out half the class? Test scores and grades for the students who drop out or fail out are sky high, elite law and med schools don't seem to have a problem getting everyone through the program. This is a lot of power with very little accountability. US STEM PhD programs need to change if they are going to attract highly talented people with the freedom to choose their careers, unencumbered by visa restrictions, and I'm not really in favor of giving them special powers to avoid that reckoning.
Do we have data showing that entry-level jobs (the first three to five years) in STEM industries have 35-50% attrition rates?
A PhD is a great way to continue learning.. And good for the overall economy because companies rely on math, science and engineering to produce products.
I would be interested in comparing the chances of a graduate student becoming a tenured professor versus a fresh grad starting a start up that becomes a unicorn.
Far better chance of becoming a tenured professor (I'd say roughly 1 in 10-20 in my field, geoscience). Plus, like has been said elsewhere in this thread, after 4-7+ years of being underpaid and overworked, you've got a PhD.
However, a freshly-tenured professor at a major research school might make 100k, and might make 150k by full professor. Very few outside of business/medicine/maybe some engineering make more than 200k by full professor.
If you found a unicorn, you probably do a bit better than that, financially.
You can also hit 'unicorns' as a scientist, for example a major discovery (nobel prize) or innovation (millennium technology prize), which should be taken into account somehow if we want to compare the two.
In start ups there are lots of discussions and analysis of long tails and capturing unicorns (implying lots of failures along the way) but analysis of careers in science never gets the same treatment, although I find many connections between the two (many small failures but big wins). It is always statistics of what an average scientist would achieve rather than what a top scientist would achieve. It would be nice to apply some of the same analyses we do on careers in start ups to careers on science.
Sure, the academic job market is very competitive (though CS arguably has more opportunities than other fields), and landing a solid industry research position is no cakewalk either. But the difficulties in finding a research job don't preclude doing software engineering work to pay the bills.
It's not uncommon to decide a few years into a graduate program that you'd rather just be making more money and drop out for a more lucrative industry/industry research position. For CS PhD students, the job market only seems terrible if you have a very narrow definition of a job (which seems irrational).
> We’ve presupposed a scenario in which there really is a massive oversupply of PhDs, and thus PhD students must be irrational for treading into an oversupplied labor market. But that’s simply not true. PhD “oversupply” is just a euphemistic way of talking about the fact that colleges and universities haven’t met student-generated demand with a commensurate supply of full-time, tenure-track faculty.
PhDs aren't irrational for wasting 5 years on something that won't net them a job... universities are shirking their responsibility to provide a job to those who demand them.
This reverses the normal definitions of 'supply' and 'demand' to make a weak point. Universities are the buyers, i.e. they have demand for faculty. The PhD graduates are the suppliers (of their labour), i.e. they can compete to fulfill that demand.
So to say universities have a responsibility to hire people, just because those people want/need jobs, goes against how we normally think about the way the world works.
No one owes me a job.
EDIT: zasz is correct. I read that quotation out of context, and as a result my comment doesn't address the point. I still disagree with the assertion, but for a different reason. Undergraduate students don't demand 'full-time' or 'tenure-track' faculty. Those aspects don't directly affect how well they are taught, or the value of the credentials they gain.
> Instead, we’ve rendered the majority of faculty contingent, increased administrators and administrative staff by 85 and 240 percent, respectively, over the past 40 years, and created a massive holding pen of temporary postdoctoral positions in STEM. If we look outside of academia for good measure, we see similar evidence of increased dependency on contingent labor, decades of stagnant wages, and no increase in leisure time to accompany increases in economic productivity. In this light it becomes harder to claim that PhD students are especially irrational or shortsighted, since so much of the broader US workforce is facing similar problems.
However, if I see it through the lense of someone who desperately wants to move the world forward -- and at any personal cost -- then I start to entertain the thought that, maybe, it's unfortunate that that some can't find jobs, and they're deserving of my sympathy.
Simillarly, I work on a bunch of OSS. A lot of it is based in academia and unlikely to become mainstream. I'm also unlikely to benefit directly from it, and I don't really enjoy the process of creation within the field of software. However, I feel that the stuff I'm working on -- even with the slim odds of mass adoption -- are worth dumping my time into because it'll move the industry forward if it's successful. You can bet that I'd love to get paid to work on this stuff, but I also understand that's unlikely to ever happen. And sure, I could let other people work on these projects so I can instead go off and have fun, but I believe that my background fills a particular niche that would likely be left open if I didn't sacrifice my time and energy in these pursuits.
Maybe it's the same with these PhD peeps. Each person believes that they're essential to progress, and their principles prohibit them from abandoning their self sacrifice in the pursuit thereof. I think I can understand the frustration; perhaps someone really ought to give these selfless people some money so, you know, they can make all of our lives better.
(In other words, I think "passionate (if not masochistic) altruism" might be mistaken as "entitlement" here.)
Maybe it would work if PHDs benefited from some kind of basic income, as long as they dedicate their life to research or teaching. Tenure should not be the only way to achieve the basic necessities for researchers. In the long run society benefits from all the fundamental research being done, which would be difficult to sponsor in the private domain, where immediate profits are most valued.
It would be more accurate to say that universities are exploitative than that the existence of a self-reproducing class of professional scientists is irrational.
>But perhaps the best way to get answers is to talk to the parents of college students, the ones footing the bill. Ask whether they’d rather pay that ballooning college tuition for lazy rivers on campus, for fruit baskets and housekeepers for executive staff, and for other rising non-classroom costs; or for more top PhD talent, smaller classrooms, more one-on-one attention from faculty, more independent studies and research opportunities with faculty, and a stable faculty who won’t have to leave halfway through their student’s college career. Then you’ll get a sense of the real, unmet demand for PhDs, and the irrationality of a market that isn’t delivering.
My willpower is wearing down, considering the effort it takes to achieve a full Coq development of something serious like that.
I also have a friend who's trying to Solve Programming Languages by building a dependently-typed systems programming language with modules and type-classes and such, so I can pass him the code and proofs if I really so burningly desire to see my work "ship" somewhere, somehow.
Smart. Might work out.
" building a dependently-typed systems programming language"
I've seen a few of those: ATS, Idris, Bedrock in a way. Has your friend published any work showing off his? Or is it still in pre-alpha development?
I know at least 3 people that knowingly got PHDs in majors that will not amount to any sort of high-paying job. They chose to do this and it's not really the job of the university to determine the market value of their degrees.