The New College Degrees: The Good News and the Bad News

In Launching the Innovation Renaissance I argued that students were not graduating with the degrees that pay (see also my piece in the Chronicle of Higher Education).

In 2009 the U.S. graduated 37,994 students with bachelor’s degrees in computer and information science. This is not bad, but we graduated more students with computer science degrees 25 years ago! The story is the same in other technology fields such as chemical engineering and math and statistics.

If students aren’t studying science, technology, engineering and math, what are they studying?

In 2009 the U.S. graduated 89,140 students in the visual and performing arts, more than in computer science, math and chemical engineering combined and more than double the number of visual and performing arts graduates in 1985.

So what has happened since 2009? The good news is that enrollment in STEM fields has increased dramatically. The number of graduates with computer science degrees, for example, has increased by 34%, chemical engineering degrees are up by a whopping 49.5% and math and statistics degrees have increased by 32%.

The bad news is that we are still graduating more students in the visual and performing arts than in computer science, math and chemical engineering combined. As I said in Launching nothing wrong with the visual and performing arts but those are degrees which are unlikely to generate spillovers to society.

We are also graduating more students in communications and journalism than in computer science, math and chemical engineering combined and more students in psychology than in computer science, math and chemical engineering combined. Here’s what I said about psychology:

In 2009 we graduated 94,271 students with psychology degrees at a time when there were just 98,330 jobs in clinical, counseling and school psychology in the entire nation. The latter figure isn’t new jobs — it’s total jobs!

Despite these problems, the number of psychology degrees conferred annually has increased since 2008-2009 by an astounding 21.4%! Visual and performing arts degrees have increased by 9.7% and communication and journalism degrees are up 8.1%. Do you think that jobs in these fields have gone up by equal percentages?

Stated differently, in 2012-2013 we graduated 20,418 more students in computer science, chemical engineering and math and statistics than we did in 2008-2009 but we also graduated 20,179 more students in psychology alone! We have a long way to go.

Here is the data:



Computer science, math and chemical engineering are a bit on the tough side for most students. It would certainly be madness to launch oneself into such a degree without having some flair for the topics. One's self-esteem must be protected at all costs, mustn't it?

Another way to look at it is this. Even if industrial demand for graduates in CS/M/ChE were to decline substantially, studying those subjects probably constitutes a better education than most humanities degrees anyway.

As a graduate and degree holder (several) in STEM, who went into business, I can tell you science does not pay (that's why I got out of it and made my fortune in business as an adviser). The science incentives are just not there. And if TC's Great Stagnation theme is true, then you will find diminishing returns in the near term in STEM jobs. Hence, if you're an attractive woman who goes into the performing arts, and becomes a trophy wife for a member of the 1%, I would call that victory.

True for science, not necessarily for engineering.
Problem is that science jobs are all in academic and government labs which pay relatively little. A PhD in physics usually means a lifetime spent slaving away as a postdoc in a basement somewhere.
You go into science if you want fame, not fortune.

As an engineer with a Bachelors and a Masters degree from the #1 academic department in the US for my field, I can tell you that sometimes the job market also sucks for engineers. The software world, in particular, is full of employees who are very bad at protecting their value as skilled labour, and that is reflected in the waves of boom and bust in the software industry, each of which results in a generation of engineers cast adrift to try to find employment in other fields where their fancy technical degrees are actually a hindrance.

So Ray Lopez is completely right: a STEM degree is a bad investment, especially when you consider how much sacrifice is often involved in making it through one of those programs. I now work in law, where my colleagues and I are much, much better at protecting our labour value than coders ever will be. It's tragic and wasteful, but that's the reality for those of us who have to work for a living and want some modicum of job security as we head into middle age.

The "STEM shortage" myth is actually true from exactly one perspective: the short-term interests of employers (i.e. tech corporations), who would love nothing more than an even GREATER glut of qualified STEM graduates to be hired and fired according to the fluctuations of the market, thereby driving down labour costs even further in those fields. The number of credulous news articles being written about the STEM shortage is exactly reflective of whose perspective is being represented by corporate journalism.

My advice to young people thinking about higher education is a hard sell, but an honest one: look for skilled trades in growth industries. HVAC, medical tech, plumbing, building trades, and IT are among the most secure and remunerative careers available -- especially on a ROI basis given the costs of (overcrowded, highly competitive) 4-year vs (undersubscribed, commoditized) 2-year degrees.

I can attest to the value of the trades. I have a geophysics degree and I made a lot of money in the oil biz in the 80s and 90s. Now I have a master's degree in landscape architecture and I make a lot of money with design-build projects.

Lauren, most people would consider someone with a geophysics degree and a graduate landscape architecture degree to be in the professional/managerial/tech field, not a tradesman. You may work with your hands, but what would your pay rate be without those degrees?

Well, I was in the tech field in the oil biz, but after getting out (due to kids and Ex-hubby's transfers) there was no way to get back in. After a very few years, I was starting all over. When I went back to school I already had a landscape design business, so I got a professional degree to enhance my skills in that industry. Problem was, no landscape architecture firm was interested in a 50 year old who they considered to have no experience. A decade of actually designing and installing residential plans was beneath their radar. Without a certain number of hours' experience under a licensed landscape architect, you can't take the tests and become a licensed landscape architect yourself. Most of my 14 classmates also ended up doing what they were doing before (planning, rangeland management, real estate and finance) albeit with a larger skill set. Only 2 actually were able to get jobs that led to becoming licensed. The percentage of folks in the regular architecture program going on to professional licensure was even smaller. My point is that no, college degrees, even technical ones do not guarantee professional success, but people who get them are usually resourceful enough to monetize their skills. Trades are different. Instead of spending a great deal of time and money in school, one can spend that same time and less money learning a trade.

Correct; all the hullabaloo over STEM fields overlooks the fact that the jobs in "S" (science) are relatively scarce and do not pay well. The higher salaries are in "TEM" (technology, engineering, and math). Actually I'm not sure about the M part, math PhDs have had a tougher go ever since the Cold War ended. Don't know about bachelors degrees in math, my guess is they probably do pretty well.

One thing about engineering though: not much job security for such a highly trained and skilled profession. They get laid off in droves whenever business takes a downturn. I've known only one PhD economist who's ever been unemployed, but I know PhD engineers who've been laid off multiple times. Engineers with only bachelors degrees have it even rougher.

Low paying as compared to visual and performing arts majors? I doubt it.

@Cooper: Good point

The people who major in visual and performing arts all seem to think they are going to be movie stars, rock stars, or movie directors.

I think we tell kids too much to "follow their dreams" without any critical examination of how realistic their dreams are.
Some people would do better with advice to major in something they can make a living at and have money left off to pursue their dreams as a hobby. Very very few people can make a living as visual or performance artists. A tiny percentage make it big and the rest are basically penniless.

Cooper, ever heard of marketing? Who do you think creates TV commercials and web banner ads? STEM majors?

Apple spends more on advertising than R&D.

Math degrees are well-respected in the financial services industry.

Why are we picking out chemical engineering from all the other engineering degrees?

well, also, why is "STEM" a thing. Running a business, either I want a smart graduate, and don't care about what they studied at Uni, or I want a specialist to do a particular task, in which case I might want a Computer Scientist (or, for some fields, some sort of engineer). I don't particularly need someone who spend 3 years fiddling with test tubes, and why that is better than English Lit or History is beyond me.

Well you're not allowed to give an IQ test

The CS majors could be made easier. There are hard programming tasks and easy programmings task, there IMO are even programming where less intelligent people can do a better job by making interface that is easier to understand. Some programing task require less intelligence and more art. So perhaps there should be an easy Computer programming major. And perhaps it would make us all better off by increasing total production.

Fill disclosure I am a programmer who not so smart. When I have a difficult algorithm to write that I cannot look up I get help from a smart person.

I agree. I'm a liberal arts major in English and Information Studies (not programming), and lucked out by finding a job that trained me in administrative computing. CS majors are really needed for software engineering but for programming for basic business processes they can really screw things up, often because their communication skills aren't that great. The setup we have at my university - train liberal arts majors in computing - has worked well because they draw smart people from areas and occupations that emphasize communication and critical thinking. I'm always hearing horror stories of young CS majors who overengineered systems to the point of unmaintainability and can't be reasoned with.

You're looking for information schools, which sit in the space between engineering, business, and computer science. Some stand alone as the university's "School of Information", (Michigan, Washington, UT-Austin, UC-Berkeley, etc.) Others blend with one of the legs, like UC-Irvine's Bren School of Information and Computer Sciences.

But it's not really about requiring "less" intelligence, it's about targeting competencies that might get missed if you're focusing exclusively on engineering or business.

BTW I think one of the things that made Steve Jobs great was that he was not too smart. Wozniak could write a very difficult interface to understand and it looked fine to him, he is a genius, Jobs would look at something and say "Who can understand that? Make it simpler." User interfaces still need to made simpler and easier. Technical geniuses sometimes write user interfaces incomprehensible to the average Joe. That is not good.

Jobs' genius was his empathy for what "the rest of us" feel when confronted by technology.

Alex seems agnostic on business (and econ!) majors.

Are they fluff degrees Alex?

Ok, I'm going to sound like a snob, but the following is my honest assessment.

I'm a teaching assistant in a physics department in a well known research university. Several times I've been a teaching assistant for the introductory sequence that's aimed at future physics majors (as opposed to engineers and pre-meds). Enrollment is certainly up; last fall the first class in the sequence was well over capacity. However, I think that many of the new students aren't quite up to snuff. Calculus was invented to solve the physics covered in the course, so it's always frustrating to see students who are afraid to do integrals even though they've passed calculus classes. The weed-out rate doesn't seem to have gone up, so a lot of the weaker students advance to upper-level courses.

I've been told that there is a similar situation in engineering at our campus.

I have less knowledge about how these students do in upper-level courses. Maybe they all get up to speed. Maybe they don't; they certainly work hard and are dedicated to the major, but working hard doesn't always mean understanding the material. Granted, I think that it's good that these students graduate with physics degrees instead of psychology degrees, but my guess is that a number of these students now graduate from the department without the skills they should have.

In short, we've pushed more people into STEM fields, but simply graduating more STEM students doesn't really mean that the pool of quality STEM graduates has gone up proportionally.

Thanks for this. I enjoyed reading it. I teach History, but have an interest in science (and econ.). I think there are a lot of students who graduate without the skills they should have in many fields, mine among them. Since I am now middle-aged, and have been able to view several generations come and go, I am of the opinion that social media and gadgets play a role, too. To what extent are students poorer today because more of their time is spent on things that simply weren't around two decades ago?

For better or worse I haven't been TAing long enough to have long-term perspective.

I will say that a lot of the poor-performing students don't seem to be doing poorly because they don't spend enough time studying. Despite my negative comments above, I do have a good relation with my students, and a lot of the poorly-performing students show up religiously at my office hours; they've often been working on practice problems beyond those on the homework.

I'm not sure what to make of it. I want to believe that anyone can become a physicist if they simply work hard enough, but a lot of the evidence I've seen does not support that hypothesis.

That said, I wasn't a stellar student when I took introductory physics (although I performed acceptably). For various reasons, I didn't develop good study methods in High School, and that really hurt me in college. It wasn't until sophomore or junior year of college that I totally retooled my study methods and started performing well. Maybe something similar will happen to my students. Working smarter matters as much as working harder.

I don't think it's that some students are incapable. I think it's that they arrive unequipped to handle the introductory classes, and then fall behind, and the pace is usually too rapid to catch up once you fall behind in a STEM program. Especially if you're behind on the math.
Engineering programs would do well to add a mandatory (unless you can test out of it) semester of remedial algebra to get the freshman students up to speed before throwing them into calculus.

Hazel Meade wrote what seemed clear to me: To get through a rigorous physics program, a very good math background on entering is almost required. I guess it depends on the program but there is usually a *big* jump from the weeding out physics 101 and 102 courses to the rest of the major apart from maybe a few non-theoretical classes like electronics or astronomy.

Yes. Most of the students with strong STEM natural talent are already in STEM.

This is an extremely important point, which the STEM advocates including Alex keep overlooking. It's "correlation is not causality" all over again: simply shoving a bunch of college students into STEM majors is not going to magically increase their salaries or productivity.

That is, those higher salaries that we see the engineering majors get? Maybe it's because of the nature of the students, not of the major: students who are smarter (or alternatively, who are skilled in scarcer areas such as mathematical and quantitative ability) may be more likely to major in engineering and other STEM fields. And maybe that's why they're making higher salaries.

" Maybe it’s because of the nature of the students, not of the major: students who are smarter "

Even assuming that this is the case and that makes the education market a zero sum game, that doesn't imply there's nothing to be gained from it.

Let's assume that the IQ's for each major follow some kind of bell curve distribution. In that case, some of your visual arts students will have an IQ at the same level as the average engineering student. If you can select for those students, you can increase the amount of engineers & maintain the average IQ of those engineers. Now, the average IQ of the remaining visual arts students will drop, but that may still be a societal benefit.

Again, it goes back to the poor quality of public education in the US. The students are arriving without the math skills that the programs expect of them to jump into an introductory calculus class. Some catch up, and some squeak by and the squeakers end up lost and unequipped in upper level courses they can't handle.

Many of the students really do have poor math skills coming out of high school. Almost all of them have supposedly taken calculus in high school, but it's often hard to detect any calculus skills beyond being familiar with the terminology and being able to take some derivatives. (I've seen d/dx e^{a*x} = a*x*e^(ax) and the likes a number of times...)

Still, by blaming their math background and the quality of public education, you're just pushing the question back a level. Instead of asking if we can make more good STEM students, we're asking if we can make more good high school calculus students and so on down the line. You seem to assume that the answer to the latter question is "yes", and I certainly hope that is true, but I haven't seen evidence for it.

We know the answer is yes, because Canada and Europe do it.

They don't do any better than we do at educating white students.

An undergraduate in an analytical science or engineering does give a better education than most humanities. Not to mention, it is far easier to read the humanities later in life than it is to pick up analytical scientific thinking.

In any case, having the requisite semester hours in physics and calculus will go a long way as desired "vocational" skills for many jobs. If we add good writing skills, that student is at the top of the employability class. Of course, those writing skills are developed by writing not taking some politicized composition credits.

My book 'Your Future Job: How to Build a Career in the New Normal' has more about this.

I wonder where all these new science grads are coming from. I was checking out a college program for my eldest today (he's only 14 but still) and came across this photo of the UBC aero design team.

There was a US Presidential candidate who once warned about college, and a corrosive moral environment. I think that UBC photo is what he feared. Too many diverse smiles.

I surfed through their profiles. Only a handful are Canadian - rest are from Singapore, Taiwan, India, Sri Lanka, Hong Kong and China.

My point is that the good news about rising STEM grads might just be more Asian students. With experience in education systems in Canada and Singapore, I see huge disparities.

Why do so few rich western kids join a global engineering culture? I think there are a lot of answers. One might be misplaced optimism in lifestyle degrees. Another that "guidance" for college education has until recently been a "follow your dream" punt.

Because both medicine and finance offer more pay for easier work. Medicine also has bullet-proof job security, and will at least until the boomers pass their peak-death rate. In engineering, global trade and H1B visa programs mean that you compete against the best in the world, not the best in the country, and thanks to rapid IT advances you must constantly retrain to avoid obsolescence.

Not many people go into finance and medicine though- so depending on what he means by rich kids it doesn't explain a lot

Why would I want to take career advice from some low level public school bureaucrat who drives a Honda and has a job that everybody makes fun of?

Does any 16 year old seriously think that high school guidance counselors dreamed of being high school guidance counselors when they grew up?

Thanks for delivering a comment up to your usual standard, anon.

I don't make em up, I just remember em

At least they have some girls.
Do those girls know how good they have it surrounded by future high earning guys who out number them 3 to 1, when many campus are 55% girls!

Of course, we also know the demand for STEM jobs is almost as weak as they are in these sectors, suggesting we have a deeper problem.

Of course, we also know the demand for STEM jobs is almost as weak as they are in these sectors (arts, etc.), suggesting we have a deeper problem.

What does data analysis tell us?

That humanities majors aren't leading to employment. Not disputing that. It's the idea that STEM degrees are some sort of unrealized boom for prospective students. The post referenced computer science alot, but we already produce more compsci majors than there are new jobs in the field.

If you have an 'actual' masters degree in computer sciences you will typically start at around $130,000 -150K. Even if it is not an Ivy League college. There are very few computer science graduates who have difficulty in finding very high paying jobs.

Problem is if you are a bachelor in history who wants to join IT because you learnt the importance of money the hard way.

Your numbers are nonsensical. My wife has been debating getting a masters in CS, so I've looked at the costs/benefits of whether this is a worthwhile career move. Even top CS programs don't have their grads starting anywhere near $130k. Your 'starting salary' range is well above the median pay for experienced CS workers with masters degrees.

You can probably start at 130K if you move to Silicon Valley. If you are competing to have the biggest salary at your 10th year reunion, this is a good strategy, even though all your increased wages are just going into rent.

If you're coming out of Stanford, maybe. If you're coming from an average CS masters program, then no, even in Silicon Valley.

I guess I was wrong to use the word 'typical' there, but all I can say is that every, every CS Masters graduate in NYC I know ( sample size 9) none of whom are from Ivy League are earning more than $120K

"If you have an ‘actual’ masters degree in computer sciences you will typically start at around $130,000 -150K."

That number looks very high. According to Payscale employees with a Masters in Computer Science and less than 1 year of experience earn $70K.

The average for a Masters and 10-19 years of experience is only $118K.,_Computer_Science_(CS)/Salary#by_Years_Experience

Maybe in San Fran?

Maybe in SF. Of course, living the SF Bay Area is around twice as expensive as the rest of the country, so that salary is more like $65k-$75k.

A good wage, but not super-lucrative either.

As DF notes, It's SF now.

San Fran? May as well put a sign around your neck, or call it Frisco.

I have a masters in computer science. I make around 150k. I only make that because of my work experience. It has nothing to do with the degree. I don't know anyone from my program that got offers starting in the ranges you listed.

As an aside I'm the guy with the bachelor's in English who wanted to join IT because I learned about money the hard way.

In other words, unlike most coders, you know how to communicate, organize ideas and express them logically to organize people into a common vision and teamwork to actually deliver large projects solving the real problems, instead of the typical effort to coerce humans to conform to the coder-machine view of good human.

How much was the avg starting salary for your batch ? I have several friends and family members in different East coast universities, are they lying to me ?

If they told you the average CS Masters graduate starts at $130K, they were terribly misinformed or they think you are pretty gullible.

Undergraduate psychology degrees are a stepping stone to work in sales.

Kids should be warned, link above, that psychology degrees have twice the unemployment rate of health sciences.

Also HR.

But does the standard psychology curriculum have any courses directly relevant to sales or HR? Or is that just where psychology B.A.'s end up? There's a lot more to being a successful sales person than what you get in a social psychology course.

Yeah, but we also graduate tons of useless Econ PhDs. I was surprised this topic was not brought up in your post.

Care to put a number instead of "tons"? It's hard to compare 50k with tons

Literal weight. Fifty econ phds probably weigh four or five tons in total.

Depending on how much gastronomy they indulge in.

About 700 PhD econ per year from US is not even a ton.

If we assume Econ PhDs weigh 80kgs on average, 13 of them constitute a metric ton, 16 a US ton.

You got that backwards. A metric ton is 1.1 US tons.

Total degrees (same source):
1984-1985 979,477
2008-2009 1,601,399
2012-2013 1,840,164
’08 – ’13 238,765
Growth ’08 - ‘13 14.9%
Growth ’84 - ’13 87.9%

Percent of total degrees:
CS CE M&S Arts Psy C&J
1984-1985 4.0%.7% 1.5% 3.9% 4.1% 4.3%
2008-2009 2.4% .3% 1.0% 5.6% 5.9% 5.2%
2012-2013 2.8% .4% 1.1% 5.3% 6.2% 4.9%

Is there really a “problem”?

@Mike W- yes, there's a problem, see the OP. Science hardly bulged in numbers while humanities did. That said, as I say above, if you're a hottie then humanities is the stepping stone to a MRS degree. Mrs. Trophy Wife.

As a percent of total degrees science increased slightly while arts and comm/jour decreased and psy increased. But the total number of degrees from '08 to '13 increased about 15%. So, the share of STEM degrees is at least the same or slightly higher in '13 as it was in '08; and, the share of psych degrees increased while the share of arts and comm/jour have declined. The conclusion would seem to be that there is an increasing share of high school graduates getting Bachelor's degrees, and while that increase is not going into the STEM fields in any greater share, those that are not going into the STEM degrees are getting more education and there is value in that ( So, I don't see a problem.

" if you’re a hottie then humanities is the stepping stone to a MRS degree. Mrs. Trophy Wife."

Yes, but that number is factored into the 1985 year stats. No, this needs to represent Mr. Trophy Husband for the up and coming female doctors. Unfortunately, while male doctors routinely marry attractive nurses, female doctors don't seem to going for male nurses. At least not to the same degree.

Who, then, are the female doctors marrying? I'm, uh, asking for a friend...

I think they assortative mate with a smart banker or lawyer, and then go off and have the smart kids, making little use of that expensive medical education.

Female doctors end up marrying Male doctors, usually their colleague or former teacher.

Most of my brother's classmates (male and female) were already with the undergrad classmate they're now married to when they started med school. The ones who were single were all male.

Based on my limited experience with medical students -- some DO take my classes in History, as an elective -- and with medical professionals, I suspect that many female doctors are quite hardworking and driven. They want to pair off with a suitable mate, then have done with it, and then get back to studying, work, the lab, etc.

"I think they assortative mate with a smart banker or lawyer, and then go off and have the smart kids, making little use of that expensive medical education. "

I've known of one who married a lawyer. But she went back to work and he raises the kids.

Ugly data. Someone needs an Excel lesson...

Maybe, but the point is, it is data. None of the comments are asking how many degrees in total and what was the increase, and, what percentage of STEM vs arts and social science and what kind of change. It appears that total degrees have increased significantly since '84-'85 and even from '08-'09 and, from '08 to '13, STEM have increased more than arts and C&J.
So, what's the "problem"?


+1 You can tell some folks didn't take a descriptive stats course.

What the kids with the soft degrees know is that non-STEM degrees are a punched-ticket to jobs in management.

Outside of STEM jobs, a college degree is just a signal that you have joined a club that considers itself to be endowed with a range of positive intellectual and social attributes that those who did not go to college lack. Spending four years drunk and landing that C-average at some junk school means you are brighter and clubbier than the guy who went to work at a factory or took a help-desk job out of high school.

I work for a huge healthcare company. All the managers are average intellects with below-average management ability, and they all have soft degrees. They are perfect supervisors for a bureaucracy that doesn't care about quality or excellence.

I tell my three kids to get a STEM degree, and if they cannot manage that, then get a degree in frisbee, but just GET A DEGREE because degrees are the secret handshake of the management world.

you want to insure that your kids can someday have loathsome, meaningless middle management jobs? I'd rather my kids skip college and become plumbers.

Really? Have you ever unplugged a toilet?

Well that didn't work, try this:

Total degrees (same source):
1984-1985 979,477
2008-2009 1,601,399
2012-2013 1,840,164
’08 – ’13 238,765
Growth ’08 – ‘13 14.9%
Growth ’84 – ’13 87.9%

Percent of total degrees:
CS 4.0%, CE.7%, M&S 1.5%, Arts 3.9%, Psy 4.1%, C&J 4.3%
CS 2.4%, CE .3%, M&S 1.0%, Arts 5.6%, Psy 5.9%, C&J 5.2%
CS 2.8%, CE .4%, M&S 1.1%, Arts 5.3%, Psy 6.2%, C&J 4.9%

@Mike W- the fly in your ointment is that population increased too. Normalize for population and I'm sure AlexT has it right, in that most kids pick humanities as a major rather than STEM.

The population increases about 1% a year...5% from '08 to '13. The total number of degrees increased 15% from '08 to '13.

Oh, sorry, I thought you were referring to 1984, not 2008. So we are talking past each other. I believe AlexT's point is about 1984, though he also makes a reference to 2008.

No, I think his point is the comparison between '08 and '13 because those figures are the change and growth rate he shows. Making comparisons between '84 and '13 asks completely different questions...e.g., the US population grew by about a third over that period while total degrees doubled.

"No, I think his point is the comparison between ’08 and ’13 because those figures are the change and growth rate he shows. Making comparisons between ’84 and ’13 asks completely different questions…e.g., the US population grew by about a third over that period while total degrees doubled." -- Mike W

So proportionally to the total population, Bachelor's degrees increase by 20%.

The window from 2008 to 2013 is drastically affected by the recession. So, I wouldn't use it as any kind of baseline.

Then why did AT use it...and what's the point of his post and this entire thread?

Isn't it an update on a past post

What leads the marginal college students, those who wouldn’t have gone to college thirty years ago, to pick their majors? When I started college thirty years ago, the first in my family to do so, I looked at my blue collar neighbors—plumbers, roofers, electricians, mechanics—and figured I would be doing a more educated version of their work, so I majored in engineering. I didn’t know any doctors or lawyers or white collar professionals. School teachers were the other model I had, and I did consider pursuing that path with a major in history or English.

So what leads someone into communications or psychology or art? I can’t imagine myself having been drawn into one of those majors, but obviously there is a draw that I don’t recognize.

I double majored, and picked psychology as my second major because it interested me. I think that's enough.

Pop culture does. Television, movies, magazines, blogs, etc.

Geez, aren't there media reports of engineers getting People's Choice Awards and million dollar wages for engineering an ultra thin super tough glass that is now in the pockets of a hundred million people?

There is Elon Musk and his team of screaming, cheering, jumping up and down, engineers and technicians complete with live video shots that look like 50s sci-fi movies, but that pretty novel for 50%, 60%, 75%??, of the global population. The last time I saw that was the 70s, but those were people who were radical for not wearing jackets over their white shirts and ties, so they did do much jumping and hugging.

"I looked at my blue collar neighbors—plumbers, roofers, electricians, mechanics—and figured I would be doing a more educated version of their work, so I majored in engineering."
" I can’t imagine myself having been drawn into one of those majors, but obviously there is a draw that I don’t recognize."
Being able to look a little beyond one's neighborhood?

Communications is, like, talking and stuff...that's useful.

I think some people enroll in college as undeclared and shop around for a major that seems interesting and isn't too difficult. Fields like engineering, computer science and even economics make it known to students early on that they need a certain math ability to complete the major successfully so some students switch their majors accordingly.

"So what leads someone into communications or psychology or art?"

Judging by the work load, I'd say it's all about the ease of the degree. I was in Engineering, but took theater and psychology classes for electives. Psychology classes were ridiculously easy. I always went to class, but never studied. I rarely spend more than 30 minutes reading the material before the class. It was always an easy A.

From the social perspective, who gives a damn about engineering anyway. Many of these fields are entirely private-sector and commercial. Maybe STEM employers should pay more.

STEM employers do pay higher salaries. However, STEM jobs are not sexy big city jobs. You end up working in a chemical plant in the middle of nowhere. Mining, agriculture, infrastructure development, trade, defense and similar activities are done in small cities or towns. I was born and reared in a frontier oil town. Anyway, there were books (invented like 6 centuries ago) and 2 TV channels. Today, we have internet.

I've met people along my life whose attachment to urban life is crippling. It's not about family attachment. It's something I don't understand like being in love with the "city".

@ Millian: some people in France care about what engineers do and allow other people to do. there's people who think you need to be under surveillance at all times so you can be fined for illegal fire-sharing.

The only person fined for illegal fire-sharing was Prometheus.

Fine by some, venerated by the cold huddled masses!

The Big Dig was done in some remote rural area? Is the NYC Tunnel 3 been built over the past two decades in Utah with the final connections into the NYC distribution system over the next five years being done in New Mexico? Are the California million roofsquare being done in North Dakota? Will the fixes to all the aging water distribution systems going to involve engineering and construction in Oklahoma, not Flint, Detroit, Baltimore,Chicago, Manchester, Boston,....?

Which US oil refineries which undergo constant re-engineering and reconstruction are NOT in a large metro area?

Granted, conservatives want to do things in remote rural areas where wages are lower and there is no history to contend with, but the best choices based on those criteria is outside the USA, and Chinese workers are the educated by US university grads of the 20th century in today's Chinese universities, are the best employees.

The Intestate Highway System is close to 50K miles and it is mostly on rural areas.

It's really ironic that you mention water distribution systems. Groundwater quality below you house is not good at all. Before water is "distributed", you need to catch it in a rural area where it's quite clean and transport it to a Metro area. Did all the raw materials (cement and reinforcement steel) used in the Big Dig and the NYC Tunnel 3 were "locally produced"? It makes sense to put as closes as possible to consumers but why there are hundreds of oil rigs in the Gulf of Mexico?

STEM people needs to be on those remote places not because Conservatives can pay less, it's because the jobs needs to be done there. Grow up.

Been to a number of oil refineries. Never been to one in a large metropolitan area.

Refineries in metropolitan areas are either located next to ports or had the city grow up around them.

The refinery(ies?) on Aruba is a weird contrast with the rest of the island.

"From the social perspective"?

You mean like clean water, electricity, transportation, food preservation, non-smokey heating and air conditioning, refrigeration, steel, aluminum, plastics, medical technologies, roads, plumbing, fuels, new fibers, new agricultural methods, etc.? That social perspective?

On the other hand, the Liberal Arts produced Marxism...

"On the other hand, the Liberal Arts produced Marxism…"

Only in your mind.

I forget the chart, but it showed that the profit-per-employee for big finance was about the same as the profit-per-employee for tech firms. The difference was that finance returned the vast majority of it to employees, while in tech a big portion sticks with stockholders.

In a media-obsessed culture, is it surprising that students would be drawn to the visual and performing arts? And how many students choose computer science because they want to write computer games? It's difficult to figure out why adults do what they do, but impossible to figure out why young people do what they do. I remember my son telling me that he wanted to be a policeman. I told him that was admirable, but that he should know that policemen aren't paid that much, especially considering the danger. With a look of puzzlement on his face, he asked: "What about Magnum?"

STEM degrees don't pay either, except for graduates who are 2SD or more above the mean in IQ. The US has chosen to make a "college degree" the minimum credential for work above the dishwasher level,* but there is precious little correlation between college coursework and job skills outside of STEM. Since STEM employers only want geniuses and H1-B's, and non-STEM employers don't care what applicants might have majored in, it is economically-rational for most college students to opt for the easiest majors.

Let me state that even more baldly: for a middling-IQ middling-diligent student, neither a fool nor a flake, a STEM degree will produce no more lifetime income than a psychology degree-- the initial employment hurdle is "any degree" and in later years no one will care. In fact, since STEM courses are harder, the student who wants a good GPA (in case he unwisely wishes to apply to an economically-pointless JD or MBA program) would do better to avoid them.

The notion that psychology majors expect to become clinical psychologists is an example of poor analysis. In psychology, graduate degrees are the entry credentials for clinical work (and in general graduate degrees are only tickets to tournaments which 70-95% of entrants lose-- look at the ratio of PhD's to academic or professional jobs). Psychology is popular because people think psychological insights may help them get along in society and manipulate others, and psychology is one of the easiest "smart sounding" undergraduate majors.

*I think the government's mania for promoting college "stems'" mainly from three considerations: (1) the college lobby is very powerful-- colleges want all those warm bodies and their tuition payments; (2) politicians can't promise jobs so they promise "education" which is alleged to lead to jobs (it doesn't, but that sort of deception is the norm in politics); (3) diversion to college holds prospective new-entrants off the job market for several years. Suppose a normal "working life" would be 47 years (18-65)-- making people waste 4-5 years in college represents a roughly 9% reduction in their working life, improving the political "unemployment" figures by a similar percentage over time.

But _median_ wages for STEM degrees top the compensation lists.

Median is median, not the top 2sd

6 of the 10 richest people in America made their money through Computer Science (Gates, Zuckerberg, Bezos, Ellison, Page, Brin).

Yeah, but that doesn't really say much about going forward does it? The 90s/early 2000s aren't coming back. Those high growth salad days are over.

Pfft. And then the smart phone revolution came along. Not to mention tablets.
There is still plenty of money to be made in IT.
We're 20 years into a major technological revolution. This is like saying there was no more money to be made in manufacturing in 1900, because people had already invented mass production.

Pfft, there are so many possibilities for software that have not been developed yet. Virtual reality alone will radically transform society over the next 50 years.

...and three of the those were dropouts. One of which was a psychology major before leaving.

Wrong. They made their "money" from creating monopolies using the least labor possible.

Yahoo! is an object lesson in the fleeting durability of monopoly "wealth".

Are Apple and Google impossible to beat with the labor of a million people working for five years? Isn't the labor cost of five years paying a million workers what the rational market capitalization of those companies equivalent to?

Duplicating Amazon would be harder as long as Bezos is in charge because he sees high labor cost as translating into wealth, while Wall Street sees low labor costs as the key to creating wealth from nothing. Bezos uses his monopoly power to pay workers to expand his monopoly on productive capital assets: more computer servers, more warehouses, more control over logistics assets.

But still, would building Amazon be ten times harder than building Tesla Motors, or five times harder than building GM?

using the least labor possible.

All companies uses the least labor possible. The difference with software is the marginal cost of selling another widget is zero.

One of my pet peeves is that all these psychology majors are not allowed to do stuff that would help people like the mentally ill homeless who need often meds.

Where do you think all those psychology majors came from?

Thanks, yes this exposes another hole in Alex's thinking: the irrelevance of college majors to future careers. Most STEM bachelors degree holders do not work in STEM professions. Most economics majors do not become economists. Most philosophy majors do not become philosophers. Etc. Except for professors and engineers, few people are working at a job that is directly related to what they majored in in college (if we're talking about bachelors degrees and excluding beautician certificates, plumbers' training, etc.) Well maybe business if we define it so broadly to include "well you're working in a business so your business major prepared you for this job".

To be sure, certain majors, engineering in particular, make higher salaries than other majors. But I believe that most of that is due to the differences in students' abilities, not due to their major. That is, the smarter or more productive students can major in anything (and succeed later in life); the mediocre students can't hack it in engineering indeed they probably don't even consider majoring in it (and have lower salaries and worse employment prospects later in life). Shoving those mediocre students into STEM fields isn't going to magically raise their salaries.

And, as mentioned elsewhere, engineering is not necessarily such a great profession, with employment prospects which whipsaw up and down. I've known more unemployed engineering PhD holders than unemployed economics PhD holders.

These stats are alarming but the discussion ignores the demand for STEM majors. I recall a time when there was a glut of engineers. There is an unstated but strongly implied "build it and they will come" presumption in this article. China produces far more STEM graduates than its economy can handle. If one views this as investment in human capital, it appears the complementary resources of production are absent.

I've always assumed my daughters will go to college. I still hope they do. It's somewhat sad to admit that if I had a son, I'd suggest him to consider blue collar careers as an alternative.

Blue collar in the Washington DC area is very lucrative. All the "master plumbers" I know have big fishing boats. Here in the Philippines however, it's a different story; the plumbers, hardly masters but somewhat competent (e.g., they use pliers in lieu of an adjustable or plumbers wrench; they have a strong hand grip), work for about $5 a project. I hired one today to install a toilet float valve.

A "plumber" in Washington DC isn't the same as a "plumber" in the Philippines. In a developed country, plumbers need to know all the rules & regulations on plumbing; they need to know how to fix a dozen different brands of HVAC systems; they need quite a lot of knowledge & experience. In a developing country, plumbers just need to be able to screw two ends of a pipe together.

Same for electricians, construction workers, etc. Blue-collar jobs are nowhere near as unskilled as white-collar people think they are.

Nor are white-collar jobs as skilled as people think they are, which is why so much of it has been outsourced to India.

Yes, agreed, though arguably all the 'rules & regulations' are just make-work and barriers to entry. At some level plumbing is just screwing to ends of a pipe together, making sure it slopes gently at a few inches drop per fifty? feet of run, so the poop rolls's not rocket science.

Rules and Regulations are designed as barriers to entry so that those jobs don't become "any guy with a few screw drivers and wrenches."

Actually some plumbing, large commercial work, can be as complex as rocket science, but just like rockets, if you aren't on the cutting edge, it is more about having the technical skills to implement already proven solutions. Not a lot of new plumbing problems requiring new theoretical work out there.

If plumbing is so simple, why do so many buildings in the US have such bad plumbing. Like the townhouse I'm staying in now in suburban Baltimore built no earlier than the 70s which burbles and gulps everytime the toilet is flushed, clogs often unless lots of water is run any time even small amounts of particulate matter goes down the drain. The property is 30 feet at least above any river or stream high water in the worst of conditions and on a gentle hill with modern water and sewer, so sewer backup into this unit from the next unit is not a public sewer problem.

From This Old House, I have learned that the building codes have been strengthened this century to address less common problems of this type. These lessons come from fixing violations of what you call simple rules that aren't rocket science. Of course, my friends in model rocketry tell me that it's simply a matter of following a few simple rules.

I have perhaps the most advanced toilet there is. The man who repaired it last week was a former IT executive at a major NYC hospital who retrained after being replaced by 5 people in India. He is the only person in the tri-state area trained to repair this kind of toilet that has sensors, motors and logic boards. He is doing very well, but he can't take a vacation because the company won't hire a backup to him and many of the customers are big shots.

The fact is, I think, there are no good career paths any more. We need to get used to this idea.

Why is that sad to admit? Shouldn't the goal in one's life be satisfaction? Perhaps your son would enjoy doing something meaningful rather than the pointless bureaucratic drudgery that seems to be focus of white collar America.

Strangely enough, this blog post does not mention the largest degree category in the US: business. IIRC, about 20-25% of all degrees in the US are some sort of BSBA.

It's not strange. The article contrasts the two ends of the spectrum. STEM degree holders are likely to end up with high paying salaries. Many of the "fluffy" soft degrees are likely to end up with low paying salaries. Business degrees are in the middle. And thus aren't as relevant to the discussion.

As someone with a business degree in an area that isn't on the easy end of the business spectrum, I assure you that business degrees are easier than the "fluff" degrees. If my sister can get a finance degree, so can anyone.

Actually, I have a Finance Degree as well as an Engineering Degree. I took a bunch of electives in theater and in psychology. At least 3 each. The average person in a finance class was significantly more capable than the average person in theater or psychology. Granted, just showing up to class every class probably puts you above the average theater major.

Maybe you aren't being fair to your sister.

As a side note: The theater extracurricular parties were far better and more frequent than any that the students in Finance held. For that matter, the Engineers threw better parties.

I agree, JWatt. Each of the business courses had their own particular weed-out courses and their own particular challenges that the average liberal arts student in my school simply couldn't surmount.

Finance was more quantitative: net present value is a concept many students simply couldn't understand.
Accounting had a hassle for rules: most people just don't understand the concept of debits and credits or why pre-paid expenses are assets.
Management had a heavy demand for business writing.
Business Computing: obvious.
Marketing actually was among the tougher courses for me, because I don't care about PowerPoint presentations or dioramas.
Economics required a student to actually his/her brain, and the Econometrics class was the hardest of the optional quantitative courses for Business Majors.

These were much easier than any of the actual hard science classes, but significantly harder than any of the soft science courses (IME). The soft science students had more exposure to academic writing, though, and were encouraged to actually think. Business students were not encouraged to think, outside of Economics.

And it doesn't even really matter where you get your degree from in STEM fields, you'll still make about as much as any other grad from any other school:

Tabarrok is looking at this backward. Degrees matter less in growth fields. If you're trying to get a job in a quickly shrinking field like journalism, you need to differentiate yourself. So a degree in journalism improves your chances. Google doesn't care if you have a degree.

I think Google cares if you have a degree.

That post indicates XVO is correct. Google's job listings require degrees. They'll consider hiring smart people without the degree. However, it's pretty clear Google cares about the degree or they wouldn't be listing it as a requirement.

No, it doesn't. It says some listings require a degree, not all and even then they'll overlook a lack of degree. Here, read this again:
"More generally, while some positions list a Master's Degree as recommended, none require it. In fact, even positions that list a Bachelor's Degree as required will consider candidates with lesser qualifications."
In fact, Google's official guide to hiring does't include the word "college." And the Quora thread links to this piece that says the trend is toward more hires without degrees:

Having worked at Google, you are completely wrong that they don't care about college degrees. I never met anyone without a college degree in the three years I worked there, and never interviewed anyone without one either. Having only a BS, I felt as though I was in the minority (large minority) compared with people with Masters and PhD's.

Of course they will hire someone without a degree (geohot comes to mind:, but you basically have to be a famous prodigy.

You've got it backwards. Journalism and communications majors are a dime a dozen. The true differentiator isn't journalism majors, it's technical skills. Most employers in journalism would go for someone with a technical background and some writing chops - which can be proven via freelance work, or even a personal blog - over someone with a journalism or communications degree. Those degrees are, if anything, the opposite of a differentiator.

In general, technical proficiency + communication/writing skill is a winning combination. And seeing as it's easier to acquire the communication skills independent of studying, it makes more sense to study the technical stuff.

Also, anecdotally, it seems to me that the most "useless" degrees are in "soft" subjects that sound practical like business or marketing (though MBAs are valuable, I'm talking about undergrad here). If STEM really isn't your thing, study a social science, or even English or History.

Senator Jeff Sessions published a primer on immigration where he spoke of employers advertising a shortage of STEM workers as a means of soliciting more visas for foreign workers, even though some of those would be replacing American workers. The constant droning about lack of STEM qualifications that was setting the stage for the immigration solution at a political level, meanwhile, persuaded more Americans that STEM can be the way way forward, lured also by the idea of higher wages or a guaranteed middle class lifestyle in the advanced economy.

"Recent data from the Census Bureau confirmed that a stunning 3 in 4 Americans witha STEM degree do not hold a job in a STEM field—that’s a pool of more than 11 million Americans with STEM qualifications who lack STEM employment.This is a constantly growing number: Rutgers Professor Hal Salzman, a top national expert on STEM labor markets, estimates that “U.S. colleges produce twice the number of STEM graduates annually as find jobs in those fields.”22 Many of the students, no doubt choosing to pursue STEM degrees in part due to bogus claims of STEM labor shortages, now find themselves with massive amounts of debt and no prospects of a good-paying job. Salzman goes on to report this shocking fact: “guest workers currently make up two-thirds of all new IT hires”—so even as half of Americans with STEM degrees can’t find STEM work, 2 in 3 new jobs in the information technology field are going to labor imported from abroad."

If I would pick one theme for this page it would be:

People with poor numerical skills do not appreciate STEM degrees.

(My high school friend and I each got physical science degrees. He went to work in a stock brokerage and I went to work in biomedical engineering. Neither was specifically in our fields. We both ended up with higher than average returns. This was not a loss for anyone, as far as I can tell. Our state subsidized educations generated returns. We did well. It would be utter folly to say that there was a loss here.)

If you want to show STEM is bad, and I know you can't, show your math. (But if you want to go beyond that, and do even better, dive down to which STEMs are better and worse, which non-STEMs are positive outliers.)

@anon- you're just bright. Even with no degree you would have done better than average. It's not STEM but self-selection. I know a smart guy who set up a certain business in the DC area and got rich, made more money than I, who got a STEM degree, and he only has a high-school diploma. Statistically, the above 110 IQ crowd does better than the sub 90 IQ crowd, regardless of degree.

IQ helps, but I think there is a broader mismatch between what we want to do, and what really requires a 4 year residential degree. We push people toward college, but it doesn't really help society if a guy with IQ 140 spends four years on fine art. It might not help that guy much more than an arts apprenticeship either.

Interesting, here is someone's collection of SAT scores and estimate of average IQ for a given major:

A bit to digest, but it seems that quantitative scores are rewarded.

Why are you replying to me? I don't think STEM is bad or overrated and expressed no such thought. I remembered the Jeff Sessions paper and thought it provided a partial explanation for the jump in STEM graduates that Mr. Tabarrok highlighted. And I agree with Mr Lopez that bright people will generally do well anywhere. Most degrees are just dearly bought signalling mechanisms, when an IQ test or some other means of preselection would serve most employers just as well, without the prospective employee missing out on years of experience or earning wages for an education of dubious worth.

Not to mention having to borrow and pay back 2-6 years of their income.

STEM is good. But so are lots of other majors, if we simply look at the median salary 10 years out. There is philosophy ($85,000), history ($74,200), and, of course, economics ($98,500).

I think the best thing a kid could do would be to start at the top of such lists, and work their way down to something they can see themselves doing. STEM is just shorthand. Or, if you really want to be a "starving artist" be aware that there is some deep history to the phrase.

STEM shortage is the greatest BS ever. H-1B Visa holders could do the job for cheaper than any American Citizen.

I think if we really believe that education is more important than signaling, the STEM subjects should be made easier and more enjoyable. Teach more people to think like stem people. Teach more people as much valuable information as they can learn. Maybe we need easy STEM and regular STEM so if you cannot keep up in regular STEM classes you drop to easy STEM were you learn at a slower pace. Maybe a general STEM major where you take classes like the Physics class my college had that was for RN's.

The state schools should cut back on the number of students who can take Sociology, psychology, history, communications etc. That would raise the bar making it very hard for students to get into those colleges.

The athletes would end up taking this easy STEM/Science major.

Sometimes science does not take so much genius as lots of people trying different things.

Such a major might reduce the number of anti-vaccers.

"I think if we really believe that education is more important than signaling" - but we don't, do we? If that was true, you could become a chess grandmaster with 10000 hours of practice. I know that to be false, as I've put almost that much in 20 years of chess practice, and have only gone from Class D to Class A (though I'm on the borderline of Expert at the moment; taking lessons from Filipino masters helped a lot).

Hi Ray,

I agree with everything you said but in science unlike chess being OK at it is useful. Actually principles of science are not that hard it is the math that describes quantifies the principles that is difficult.

I spent a tremendous amount of time in my youth trying to become and NBA player but was not even good enough to make a division 1 college team. I have spent many hours programming but I am not great compared to some people around me but I am not useless. There are many things that could be automated in business with not so difficult programing.

The signal is important so perhaps the best idea is to have hard and easy STEM degrees.

I don't know if this is possible. Introductory courses or electives (e.g. the infamous "rocks for jocks"-type courses) can cover the basics of a subject without lots of complex math, especially if the professor is someone who is especially talented at presenting and simplifying ideas. Past a certain point, though, the point of a major is that it is supposed to offer "deep dives" into the topics touched in the first one or two intro courses. There is only so much science you can understand without using calculus.

Even with the calculus (and the correct), there's quite a bit of physics I couldn't _understand_.

Even with the calculus (and the correct answer), there's quite a bit of physics I still couldn't _understand_.

Putting a cap to the "Social Natural Philosophy" would not improve the quantity of STEM majors. Instead, institutions end up making monetary lost. The "Social Natural Philosophy" students are subsidizing the STEM majors!

1) STEM degree programs, particularly EM ones, are typically more difficult than humanities and social sciences programs. Consequently, they're going to both attract fewer students and weed out more than less demanding majors.

2) Not everyone can learn everything. People and their talents are not interchangeable. You're assuming that all those kids going into psychology and visual/performing arts are just as capable of succeeding in a STEM program. I very much doubt that. Arts majors tend to be very right-brained and thoroughly mediocre, if not completely shit, at STEM subjects. I'll also submit that above average to high verbal intelligence may be substantially more common than above average to high mathematical/logical intelligence, so STEM subjects may have a lower "natural" ceiling of enrollment than humanities and social sciences. (Anyone know of data to test that hypothesis? Aggregate ACT/SAT scores, perhaps?)

3) Given American popular culture, visual/performing arts degrees and the like are sexier and more fun than STEM programs. Keep in mind, too, that the vast majority of college students lack real world life experience, have unrealistic expectations of both the job market and how far they can go, and have been told all their lives by their parents, friends, teachers, relatives, etc. that they can be whatever they want to be if they just try hard enough. I guarantee you that far more of them dream of being a famous actor, pro athlete, artist, or fashion designer than do of being an upper-middle class engineer or corporate IT drone. Most young Americans are deeply naive and have had their expectations of life set way too high.

4) Contrary to popular belief, the real tsunami of college degrees over the past few decades has been more in business and vocational majors than in the humanities.

I submit that STEM programs aren't losing potential students to the humanities and social sciences so much as they are to Business Administration, Management, Marketing, etc. How many students consider a STEM major, get a taste of it in their lower-division gen. ed. courses, decide they can't hack it, and become business majors instead?

It is commonly believed that true competence in quantitative reasoning is more rare and more difficult than equivalent skills in language.

But that is because everyone uses language and judges himself competent. And is blind to their insufficiencies.

High competence in language is extraordinarily rare, and extraordinarily difficult.

As Thomas Mann once said: Writing is more difficult for professionals than it is for other people.

What about economists? Does America have as many economists as she needs to prosper?

Probably not. The undergrad economics degree has become CS-lite with a little extra statistics, and those people are very useful and employable. Be very wary of econ departments that lack a serious programming track and exposure to a lot of languages in project-style classes.

"and those people are very useful and employable."
In other words, they are not what we used to call a economist.

Is it that hard to stick these numbers into word or excel and generate a pie chart? A picture says a thousand words, a good graph even more.

Does anyone know where to find this kind of statistics about major, separated by universities? Or at least by type of universities?

Many engineering and computer science undergrad programs, particularly at top school, 'weed out' students---directing them to other disciplines at their school. They do this by either requiring separate applications to the programs during the admission process, or by thinning the ranks in the first year.

Directing more students from 'soft degree' to hard engineering at the college stage may not solve the problem as the additional engineering students won't be able to compete in their field.

This does not mean that it may be possible to increase the number of students able to compete by greater efforts in K-12.

Yes this is exactly right. Most STEM classes have a strict curve. "Soft degree" courses typically do not. Many freshmen and sophomores who are figuring out what to major in still aren't sure whether they want to go to grad school or not, where undergrad grades will be a huge factor. Get straight As in psych and have a blast in college, or struggle and work your ass off to get Bs in a STEM major, potentially also knocking you down a tier in grad school?

While grad schools do value a STEM degree much more than a "soft degree", I doubt their value is enough to make up for the grade difference. In other words, maybe B+ STEM would be taken over A- English, but I doubt B STEM would be taken over A English.

I confess I don't get it. Where are those extra STEM graduates coming from? MIT, University of Phoenix, State U., Eureka College? How can America be educating 50% more chemical engineering students than it used to educate a few years ago? Does it mean there was enormous idle resources at universities' STEM departments ? How flexible is the American educational system to answer to changing students' preferences? Isn't the admission to STEM courses more competitive than to other lower-paying divisions?

At my midwestern engineering alma mater, admission was extremely UN competitive. Acceptance rate was something like 85%. (Admittedly some self-selection happened for application.) But the first year chemistry and calculus classes were used as weeders. Can't understand covalent bonds in a 300-person auditorium lecture with high stakes testing? Sorry, you're out. First year attrition was approaching 50%.

Engineering degrees as a proxy for intelligence is well known. I recall struggling through two years of calculus plus differential equations, and the third-year principles classes, only to have the capstone courses ignore all of that to focus on how to use tools and best practices. Ask any professional engineer the last time she or he calculated dynamics on paper, instead of putting it into a $5,000/seat software program and letting the computer do the work.

OK, I don't know how typical your alma mater is of American universities, but I guess this philosophy is plausible. However, note that Mr.
Tabarrok is talking about students who were graduated, so it is not just a "let Newton sort them out" thing and drop-outs freeing desperately nedded resources for other students. Apparently America can graduate up to 50% more STEM professionals than a feears ago without noticeable fall of quality and no extra financial efforts if the whims of students go this way. If it is true, it is miraculous.

Does it mean there was enormous idle resources at universities’ STEM departments ?

Yes. Enrollment in STEM has been under capacity for a LONG time. And academia is not an environment that sheds excess capacity quickly.... or ever. There's this thing called tenure.

What kind of idle resources? Auditoriums, teachers' working hours, money, computers, desks? I
just don't understand how there can be idle resources if students must compete for positions at universities. Doesn't it mean that universities accept as much students as they can teach (evidently, some are more sought after)?

"What kind of idle resources? Auditoriums, teachers’ working hours, money, computers, desks?"

All of the above. Freshman engineering classes might be full, but they are taught in large auditoriums, so just adding one instructor might double capacity. And beyond the Freshman year, most classes weren't full. Many had less than half the seats occupied. Weed out rates are high and pretty standard for engineering.

"Doesn’t it mean that universities accept as much students as they can teach (evidently, some are more sought after)?"

No, American engineering schools are highly selective. Just the entrance requirements are stiff. Then they intentionally set certain Freshman classes as hurdles. Freshman calculus is certainly a weeder class.

They don't try to maximize their graduation rates.

And there are plenty of professors who either aren't teaching any classes or who could easily pick up an extra class if needed.

But I'd disagree on acceptance and graduation rates. Just about anyone who is interested in an Engineering degree can get into a program if they want to. maybe not in a top tier school, but certainly in a state school. And they don't use weeder classes like they used to. You can take the same class over and over and get endless tutoring until you graduate. Add in grade inflation and eventually the school will pressure TAs and Professors to just pass the student so they can move on.

"Freshman engineering classes might be full, but they are taught in large auditoriums, so just adding one instructor might double capacity."
OK, this I can understand, although I guess the practical classes may demand a little more investment than just an extra teacher.
"American engineering schools are highly selective. Just the entrance requirements are stiff."
It means it is not easy to get in, right? In Brazil, minimum requirements aside (some universities impose also their own, more stringent requirements), universities accept as many students as they can teach (say, 100 freshmen for every course) and let competitive exams weed out the extra candidates (so it is more difficult to get inside the most prestigious courses/colleges, some of them sporting a 100:1 candidates: spots ratio-- beating the other candidates IS the requirement), i.e. all the vacancies are taken care of unless:
a) there is a shortage of students coveting the spots (specially in low prestige or too expensive colleges).
b) there is a shortage of quaified candidates.
c) dropouts free some spots mid-course (but usually they are filled by transfered students).
Barring these exceptions, every possible spot is already filled. So the only ways here of admiting more students is spending mone to create more spots at state colleges or spending money lending money to students or outright "buying" spots in private colleges for students that would not be able or willing to pay. Aside that, we can only redistribute openings (affirmative action is much more common than ever).
I guess it is great if America can graduate much more STEM professionals than she used to without hurting the standards. I never heard of anything like that..

"And they don’t use weeder classes like they used to. You can take the same class over and over and get endless tutoring until you graduate. Add in grade inflation and eventually the school will pressure TAs and Professors to just pass the student so they can move on."
So the limiting factor in not-so-top programs may be candidates.

"Just about anyone who is interested in an Engineering degree can get into a program if they want to. maybe not in a top tier school, but certainly in a state school."

Hazel, that's just not true. All the engineering schools I'm familiar with require a minimum testing score of some kind.

For example, University of Tennessee Knoxville:
"A Success Prediction Indicator (SPI) number of 60 and a math ACT of 25 or a math SAT of 570 are minimum standards used for admission to the College of Engineering. The admitted class may also be limited by space available in the College. The SPI is calculated by adding an individual's ACT mathematics score to 10 times their core high school GPA (based on a 4.0 scale). "

"SPI EXAMPLE: A student with a high school core GPA of 3.5 and an ACT mathematics score of 28 would have an SPI of 63 using the formula (3.5 X 10) + 28 = 63."

For reference the average ACT Math score in the US is 21.

Fracking liquids, analysis of sediments, and separating the games and fluids from a well are best done by a literature major who is good at composition, analysis, and editing out the waste.


University classrooms are not specific to certain disciplines and you can squeeze in more students.

They may be empty a lot of time, too.

Adding teachers is easy.

Fire code! Firefighters and the Fire code regulation requires a cap per square footage of buildings. There is a cap to how much you can squeeze in to a classroom.

In my experience University buildings during teaching hours tend to be far below capacity. Often they're nearly empty after 4 pm most weekdays and on week ends.

Squeezing people in is not the same as educating them. Even if it were, it would mean the limiting factor was people willing to study STEM subjects, which would mean the previous situation was uncompetititve (at least, outside the top schools), which doesn't make any sense. Maybe it is retention that got better. I can believe that teachers and students are getting their act together ("more together" than a few years ago, anyway). It would reflect well on America.

Yeah, experience doing chemistry, physics, biology can all be done without equipment by computer simukation, while sports is impossible to simulate on a computer and requires hundreds of millions in capital assets just to run a program for a couple hundred student athletes.

Why just look at all the computer games simulating building rockets, boring tunnels under cities and rivers, building refineries and polymer plants and plastics molding equipment and products, plus the games playing surgeon operating on brains and hearts and especially joints like knees through small incisions. No need for actual capital investments in the sciences because software is a better substitute!


"In 2009 we graduated 94,271 students with psychology degrees at a time when there were just 98,330 jobs in clinical, counseling and school psychology in the entire nation. The latter figure isn’t new jobs — it’s total jobs!"

AT needs to get off campus more....why base the conclusion on only the jobs available in "clinical, counseling and school psychology"? What about marketing and sales...a psych degree is likely better preparation for those careers than a Marketing major in the Business School.

Did we forget about signaling? Having a STEM degree is a proxy for "smart and hardworking" - there are a lot of folks out there with a BS in mechanical engineering who haven't professionally touched a CAD diagram.

STEM degrees do have a high starting salary and a decent median salary, but they definitely do top out in the low $100s. Then you have to move into management, in the traditional corporate path, or start/join a young startup and ride the elevator up.

"STEM degrees do have a high starting salary and a decent median salary, but they definitely do top out in the low $100s. Then you have to move into management, in the traditional corporate path, or start/join a young startup and ride the elevator up."

Sure, but so what. You are talking about a small fraction of the population at that point. 96% of US salaries are less than $100K.

Seriously. 100k is an awesome salary, particularly without the onerous student loan debt of other advanced degrees.

Another path is that a STEM degree can be parlayed into a career in patent law, which is very lucrative.

"Here is the data:"
Uh, no.
Here ARE the data. or Here are these data.

I think that horse left the barn ;-)

I would guess a small minority of people in STEM in TECH have engineering degrees. Many have related..
Bootcamps, Business Info Systems, Community college classes..other

Some decent comments in this thread. My $.02

Average is over. amirite? If that is the case, why swing for a double with a STEM degree. Why not swing for the fence with a "people person" degree (which includes business and econ degrees)? Not only do you get to party more while in school, you'll probably make more connections with those folks who end up in leadership positions.

btw, what are the trends with the advanced trade degrees like MBA, JD, and MD?

Full disclosure, I hire developers. Why some of them waste 4 years getting a somewhat worthless CS degree is beyond me. Oh wait, I guess it gets back to that party time thing. OK, carry on...

I'm a computer programmer who gets pinged by a number of headhunters each week fishing around to see if I want to change jobs. These days it's usually through LinkedIn.

I'm curious to see these headhunters' backgrounds. It is almost always somebody with a communications, psychology, political science, English, marketing, visual/performing arts, or journalism degree. I'm increasingly seeing people with general business degrees.

I always have to wonder about those headhunters, "When you were a starry eyed 18 year old going off to college, is cold calling and emailing computer geeks what you imagined you'd be doing for a living?"

At the same time, actively promoting STEM majors is not a good idea. Market signals will do a better job of matching the skills needed with a young person's skills/desires/aptitude than will a $4 billion push by Obama to make every kid a coder.

While there are a lot of STEM fields out there that do pay well if you have the qualifications, they are not always plentiful and are often scarce even in medium to large cities.

It can't be that bad. I know a couple headhunters. They all make 6 figures. I know one who just received a $250k job offer.

Using 1985 as the baseline is a bit misleading:

Source is the work of Ed Lazowska of the University of Washington in Seattle.

Thanks, that's good data. 1985 was a local maxima and also 2008 was a local minima. So, the data from the original post is highly misleading regarding CS degrees.

I really think folks in the comments are creating too complex a model of the 18 year old college freshman picking a major. They're not looking at employment statistics, they don't have insight into how employers will view their degrees.

They pick a major that seems interesting to them, and major in it.

"They pick a major that seems interesting to them, and major in it. "

I agree, but perhaps they ought to be exposed to a few power point slides before they sign for the huge student loan. At the least, they should be shown the average wages of last years graduating class for their major and college. (Well probably graduates from two years ago, would paint a better picture.)

Perhaps they should be exposed to realistic interest rates and loan terms for the major they chose.

If student loans were private, then terms and interest rates would be much less favorable for psychology majors.

This is a bit like the special mortgages available to doctors.

The interest on my wife's Department of Education loan is higher than the interest on her private market loan. It'd be cheaper for my in-laws to mortgage their house and me to pay them back than to continue forking money over to Useless Sam.

She majored in a STEM field and her honorific is "Doctor." Not a psychologist.

Young adults do not understand interest. I certainly didn't, nor did my Sister, nor did my Brother: we just were children of an Accountant who had an attitude of "NO DEBT NO DEBT NO DEBT" and so we all graduated without debt.

Yes, indeed, the economy would be just fine if people just behaved the way they were supposed to.

Funny, before the last century or so, being a math genius was not an especially useful or lucrative skill. You'd be better off being a skilled orator or craftsman. Three-quarters of us were farmers. Now, if you're not a math-whiz by age thirteen, you are just a burden on society and would be better off dying to decrease the surplus population. I doubt even the Social Darwinists in the Human Biodiversity Movement would argue that evolution works that quickly, but I could be wrong.

And the people who make it to the "top" of the income ladder get there not by means of superior intelligence, but via an extroverted personality, an ability to completely focus on their chosen field to the exclusion of everything else, and feral social cunning.

All of it obscuring the fundamental fact that Neoliberalism can't produce decent living standards for the majority of us. Sad.

Ahh, the quaint term "neoliberal". Reminds me of the 90s.


You're soaking in it.

You need to be a math wizard to harvest 3000 heads of lettuce per day at 3 cents each, trimmed and bagged for the store display?

There are two key flaws in your argument that undermine the entire thing.

First, the people at the "top" of the income ladder, as a group, are of superior intelligence. Perhaps they aren't more intelligent than "physicists", as a group, but they are more intelligent than people at other positions on the income ladder, as a group. Obviously, these assertions depend on how one measures intelligence. So you can let us know what metrics and data you are using.

Are the living standards of people in developed countries not higher than those of the farmers of a century ago? The living standards of common folk in developed countries are quite high by any benchmark. There is a certain renewed backlash against capitalism (neoliberalism) that is fashionable at the moment. It's unfortunate that the labor share has decreased and that median real wages have been stagnant, but welfare programs that ship benefits almost entirely to the bottom 10-20% will actually make the situation much worse for the median.

I get the impression that the reason the number of computer science degrees is stagnant is that it's possible to learn computer science on your own without spending tons of money on college, and computer jobs are sufficiently meritocratic that you can get one if you know what you're doing even if you don't have the credentials.

Really it's a microcosm of what makes the whole "we need more skilled workers" argument for increasing education so ridiculous. Actual workplace skills are easier to acquire than ever. College isn't a very good way to get those anyway. The increasing need for a degree in the modern economy is almost entirely about credential inflation, not any productivity increases from having that education.

Once again, people are looking at the bombers that came back and drawing the wrong conclusion about what parts to reinforce. The STEM degree is not the valuable item. It's what is required to get a STEM degree (intelligence and conscientiousness).

A bright, hard-working person with a liberal arts degree is likely to have the same life outcomes as a bright, hard-working person with a STEM degree.

An unintelligent, lazy liberal arts degree holder is likely to fail in life.

An unintelligent, lazy STEM-intender won't survive the program and won't get the degree.

College is consumption, not investment. V&PA is a great degree if it's consumption. Ain't nobody got time to consume a Chem E degree.

Most psychology majors don't look for jobs as psychologists. Likewise, most computer science majors don't look for jobs as computer scientists.

First-time commenter, have mercy....

I graduated in May 2015 with a chemical engineering degree. Slightly less than half of my class had job offers at graduation, and most of those, myself included, were not in a traditional chemical engineering field. 8 years before, the hire rate was effectively 100%. What's changed? Among other obvious things, class size. My graduating class was over 50. The class 10 years before was under 25 (I think it was 23). This is not a result of dilution of quality. The academic standards at my (public) school, to the extent that they are measurable, have risen markedly as higher caliber students are pushed out of private education by increasing costs. In that same 8 year time period, the starting salaries of those students that could find jobs have declined in not just real, but nominal terms as well.

If there ever was a shortage of engineers, there certainly is not one now.

Engineers kill jobs by costing society too much. That's why chemical engineers were not hired a couple years ago to cut Flint's water system costs.

Look at all the jobs created handing out bottled water and ineffective water filters by cutting costs by not hiring chemical engineers who would have forced outrageously high costly wasteful investment in capital assets.

This is academic malpractice. The question is, if you "arranged" degrees to match the job market, what would you get? Are there anything like enough "professional" jobs to absorb everybody? Maybe all those in oversubscribed fields should just take business classes...

I wonder why "choosing a major" is the single thing that people cannot be trusted to do on their own, since the main message of this blog is "market know better, and people are mostly rational".

Keen observation. I think back to that period of time and recognize how clueless I was, and similarly recognize how clueless I probably still am ;-) My entire philosophy of the time was similar to Red Forman's outlook: work hard, presumably at something difficult, be trustworthy, and it'll all work out for the best.

My first choice was to be a rock star, but even I was clued in enough to know that wouldn't happen. But at that age I was mainly following the lead that going to college was best for me, and if I studied something difficult and worked hard it, things would work out.

Obviously the demand for those degrees reflects a range of reasons other than simply post graduation employment prospects. A degree is still useful in signalling ability, and formative education confers benefits elsewhere. Or in other words, money isn't everything.

Alex, knowing all this, if you could go back in time and do it over would you have done a different major?

At least at George Mason University in the '90s (when I went there), very few students were officially Undecided...while many were officially Psychology majors. The impression I got was that students felt they had to declare something, and Psychology was considered the standby.

OECD soft degress by countries

I see no good reasoning just *why* the country focuses on a STEM shortage.

STEM is a field where having a big brain is so disproportionate to success that having every single person below the 50th percentile major in stem will just result in a *lot* of crappy code and engineering. In fact, it would be a national health disaster, with so many moronic civil and environmental engineers.

Whats the data on people with many options majoring in STEM? My guess is its still going up, as CS is a major fad right now...deservedly so for the very smart.

Are the very smart their own worst enemies though for job security, with the trends in machine learning?

Comments for this post are closed