Determinants of college majors

Students who happen to be assigned classes in one of four required subjects during the semester when they’re supposed to pick a major are twice as likely to major in the assigned subject, according to a new working paper from Pope, and Richard Patterson and Aaron Feudo of the U.S. Military Academy.

This held true regardless of how well a student did or how much they liked the course, according to the economists’ analysis of U.S. Military Academy class data from 2001 to 2015. Their database included grades, class times and students’ opinions of the course. It allowed them to control for factors such as students’ hometowns and racial backgrounds.

“Small and seemingly unimportant things can really have a large impact on people’s life decisions,” Pope said. Often students cite a specific class or teacher as justification for this life-altering choice.

In a related paper, the economists, along with Carnegie Mellon’s Kareem Haggag, showed students are about 10 percent less likely to major in a subject if they took a class at 7:30 a.m. Likewise, as students grow more fatigued during the day they grow about 10 percent less likely to major in the subject covered by each successive class…

Given how easily a first choice of major can be swayed by accidents of timing and environment, it’s perhaps not surprising that 37 percent of students eventually switch, according to a new paper from University of Memphis economists Carmen Astorne-Figari and Jamin D. Speer that will be published in the journal Economics of Education Review…

Students with lower GPAs are more likely to leave their major. But so are women of all ability levels. In contrast, men are more likely to drop out instead of sticking around and trying a different subject, according to a study published last year by Astorne-Figari and Speer.

Both men and women are most likely to abandon majors in the sciences. In addition, education and philosophy appear in the top five majors men leave most frequently, while women are more likely to leave computer science…

Students tend to switch to less competitive majors.


On net, business, social sciences and economics tend to gain the most students from major switching, while biology, computer science and medicine (medical and health services) tend to lose the most.

Here is the full WaPo piece by Andrew Van Dam.


I wonder what majors the authors of these papers changed from.

Um. the major at West Point is to become an Amy officer. And fatigue is part of the regimen.

Basically, the service academies are outliers in a number of significant and easily enough determined ways.

Like this, as a minor example -

That's the point, military academies are natural experiments.

U.S. Military Academy cadets get classes assigned by superiors, thus class selection is "random" because the cadet has no say on it. Cadets chose a major from one of the current classes they were assigned. Memories of past classes seem to be unimportant. What matters is what the cadets have in front of their eyes at decision time.

PS. read the article first ;)

'That's the point, military academies are natural experiments.'

Of how the American military trains officers.

Just read the Post article, not any of the PDFs - not sure what that has to do with pointing out that the military academies are designed to train military officers (yes, the Post article tends to be mixing a couple of things). This includes fatigue, rigid schedules involving meal times, classes, physical education etc, and a penalty system for appearance, attendance etc, which would tend to suggest a natural experiment meaning that mixing results from Pope and from Haggag would show something based on data - depending on the cleverness of the person examining the data, it might be mundane or exciting depending on presentation, using the necessary 'empirics.' (Essentially, one would likely expect a divergence between the military academies and a typical university in terms of changing majors based on class time, but who knows?)


You don't know what you're talking about.

Yes, the academies train officers, but they do so in conjunction with college education. They have actual college majors just like every other school. Their professors include civilians as well as soldiers who are working on their doctoral degrees.

Where you might be right by accident is that regardless of your choice of major in an academy, you are going to get the same pay and opportunities as any other major, at least initially. But this is part of the natural experiment. They're not choosing major for immediate financial gain.

But academy graduates can and do often return to civilian life where their choice of major matters. This could have some remote, forward looking effects.

What also differs substantially at academies is that the average HS GPA will be much higher than typical colleges, on the low side of an Ivy.

'You don't know what you're talking about.'

Then apparently the at least dozen people I knew growing up who went to West Point or Annapolis did not know what they were talking about either. (The Air Force Academy graduates I knew were all pilots anyways - the Air Force Academy appears to be a bit different, at least up until the early 1990s.)

'but they do so in conjunction with college education'

Absolutely, the officers should be well educated. Surprising that the USNA mission statement seemingly ignores that point completely --'"To develop Midshipmen morally, mentally and physically and to imbue them with the highest ideals of duty, honor and loyalty in order to graduate leaders who are dedicated to a career of naval service and have potential for future development in mind and character to assume the highest responsibilities of command, citizenship and government." Almost as if they don't know what they are talking about, right?

'Where you might be right by accident is that regardless of your choice of major in an academy, you are going to get the same pay and opportunities as any other major'

Nope - but that is based on actually knowing graduates from service academies, and am aware of the importance of specialties related to areas that favor advancement. Admittedly, the connection between major and future career track are not as intricately entwined. McCain was a (very) poor student (and a bit of jerk according to one person I know who went to Annapolis the same time McCain did), but being a carrier pilot certainly brought him career advantages (along with the opportunity to demonstrate under torture what Annaoplis teaches, which is not about a major).

'But academy graduates can and do often return to civilian life where their choice of major matters.'

Not in my experience (admittedly, no longer up to date, and involving career officers) - the 20 0r 30 year career means the civilian job they get with a company is almost always related one to one with the program they had been involved with - someone working on a cruise missile program works on that cruise missile program in civilian life (it certainly helps to already have the necessary security clearances already).

'This could have some remote, forward looking effects.'

Not if they have even the tiniest bit of experience how the military works - and expect to be a career officer. Certainly not everyone going to a service academy expects that to be the case.

'What also differs substantially at academies is' is the mandatory letter of recommendation from a member of Congress (in the case of Annapolis, each member is limited to 5 constituents attending at the same time).

"What also differs substantially at academies is the mandatory letter of recommendation from a member of Congress": how much does one have to pay to get such a letter? Or is it purely a matter of privilege rather than money?

Yup, any statistic about students changing majors needs to account for the nature of the major (bio, education, econ, etc.); the nature of the student (strong vs weak, male vs female, interested in science vs doing pre-med due to lack of other ideas); and the nature of the institution (West Point vs Evergreen State College).

I'm inclined to think that any overall averages as quoted in these articles are pretty much useless if they don't account for those basic attributes of the major, student, and college. Or if not useless, then limited to their domain of research e.g. West Point.

Education dead last in grading standards, 27th out of 29 in SAT scores.
If that is true in other schools that would explain a lot.

Also highest number of PhDs by far, indicating the standards are low.

The Ed major is the joke of higher education in the US.

On net, students switch to easier subjects. Your tax dollars at work!

Such is life in Trump's America.

And was in Obama's America.

A good indication that the students know it is the signal the magic parchment sends that is the most important outcome of college.
Learning something difficult, that requires disciplining of the mind and regulation of the heart is a very distant ninth or tenth place.

I picked my second major because of a class I was taking when we were supposed to declare. It was probably a better method than the basis for my first major - my favorite class in high school.

One obvious reason why students are more likely to major after taking a required course is that it counts toward degree completion. That is, if you are forced to take Econ 101 and 102, you are two course closer to meeting the requirements for an Econ major and have satisfied the prerequisites for upper division work.

It isn't at all surprising that students will leave hard majors for easier ones, even if their grades aren't that bad.

Education is among the least rigorous of all college majors, sucking up the dregs, hand wringers, and idealists. The best thing we could ever do for education in America is to eliminate the Education degree.

Eliminate it or harden it.

One problem, without teachers the students would not even make it to university to join the other majors. Even economists had to learn to add and subtract somewhere. Given that education has low status and low pay, of course, only the idealists would dare choose it as major and buck a culture where money and status reign supreme. I am not defending schools of education, they are largely complacent and not preparing teachers adequately, but only those who have never taught a room full of 12-year olds would say that we should eliminate the preparation of teachers.

What does an education degree have to do with preparation of teachers? As you point out, that is the problem.

In high school(this would have been around 2005) they told us if we wanted to be teachers, we shouldn't major in education, we should get a major in math or history or whatever subject we wanted to teach. Only half of high school teachers are ed majors, and they get the undesirable jobs.

Agreed, high school teachers in the major discipline areas (e.g., math, history) are often those who majored in those areas and then added on a few of education courses. This is the transition (as I see it): K-8 educators largely studied how kids learn; 9-12 teachers have focused more in a content area and learn a little about how to teach; and higher ed professors generally know their content area but little (or nothing more typically) about how to help others learn it. None of these are optimal and they are all failing us... with the first and third coming up the most short in my view. The pressures are building though and the cracks are showing, so we have to figure out a reasonable alternative soon. Technology is helpful, but it is not the whole solution. Someday in the future, even if cognitive implants can get the data/information into our heads, we still have to learn what to do with it. Teaching will not go away, it will just have to look quite different and we will have to prepare people for it (rather than let them stumble into it without adequate preparation).

From Alex Tabarrok, a few years ago:

“The 80,000 hours research charity, co-founded by William MacAskill, can be a bit preachy but they have assembled and reviewed a large amount of research on careers–not just on what makes a career useful but also what makes it enjoyable. Young people spend surprisingly little time thinking about a career. There’s a lot more advice about choosing and getting into a college than there is serious advice about choosing a major let alone figuring out a practical plan towards a career.”

Something like 25% of students regret their choice of major at graduation. IME, besides those who picked an alibi (i.e., whatever their parents told them to study), most of those got sucked in by the “we teach you how to think, not what to think” spiel, as if studying sociology or General Studies, or whatever was some sort of CrossFit for the mind.

Also: [Indignant snort at the continued absence of Linguistics in any list of majors]

Seemingly trivial things can determine the choice of a major, but how important is a college major for later job choices? I would be willing to bet that for most people, the answer is "not a lot". Simply because courses of academic study often aren't closely related to non-academic employment.

It's a classic correlation vs causality confusion, which too many policymakers unfortunately fall prey to. It's easy to get statistics on the average salaries of say petroleum engineering majors compared to education majors. And to then assign all sorts of causal importance and life-changing outcomes to the student's choice of major.

I believe that you are correct that students' choices of majors are relatively unimportant, and to the extent that they have predictive value for life outcomes it due to being an effect rather than a cause. It's a classic endogeneity problem.

"but how important is a college major for later job choices?"

You are highly unlikely to be hired for software development or electrical engineering if you did not major in those areas (yes, some gifted programmers can avoid college, but they are rare).

If you don't major in Physics (or perhaps math), you don't have much chance to go to Physics grad school and get on the Physics research track.

Good start. Now please account for the other approximately 99.99% of college majors...

washington post sandwich narrative
"They may be retreating from an overtly or implicitly sexist or racist environment."
got it so a so now a rigorous field is implicitly racist/sexist
really? whats the evidence for this ?

isn't it more likely that students in stem fields are avoiding
the overt/implicit racist/sexist environment of the sociologists&the washingtonpost!

It's an insane system! The experience in a university classroom bears almost no relationship to that of the work that a major prepares you for (unless that career is higher-education instruction -- and even then it's not all that representative). And high-school graduates get almost no help in understanding what different kinds of careers are actually like on a daily basis.

Nudge iffy people into accountancy

'Students tend to switch to less competitive majors.'

In college biochemistry was known as 'pre-nutrition.'

Another reason why freshmen should not be allowed to fulfill their general education requirements with classes in sociology or gender studies or communications.

It would be really nice if all colleges would simply require a solid, difficult to pass math course plus another one requiring technical competence in essay writing for all students in all majors. I know, I know, I'm such an idealist.

As long as composition is consigned to the English department competence in essay writing will falter. Writing is a vocational skill for the student and should be taught by teachers who are no experienced teachers, not frustrated grad students or those hoping for "real" teaching position in literary theory.

That these comp courses are now used to promote sociology ideological positions just makes it worse. Students know the class is BS and apply the appropriate effort.

At my college, majors varied dramatically in the extent to which they "protected" their upper level classes. Some majors had an intro sequence (one or more 101-level courses) that were a pre-requisite for any upper level courses in the major. You simply could not take an upper level course, as an elective or otherwise, without completing the intro sequence. Other majors let just anyone sign up for any class. And these were all humanities fields, not places where you needed certain math skills to understand a higher level course or anything like that.

I ended up in a major that let anyone take upper level classes, after taking a few upper level classes as a freshman and liking them.

I'd like to see a controlled trail where students are fully informed about expected salaries based on major (as well as student loan default rates by major, see:

"Arts majors have the highest overall default rates, while STEM majors default at the lowest rates. Both Business and Vocational majors default at higher rates than STEM majors, but at rates closer in magnitude to STEM majors than to Arts majors."

I think it has more to do with one's math abilities. Sciences and engineering demand deep dives into math that most people do not do well in.

Comments for this post are closed