Paul Graham posted a list of Yale majors yesterday sorted by how easy the grading was in each one. Economics was the toughest: only 52% of all grades were A's. History of Science clocked in as the easiest with 92% A's.
This got me curious: does easy grading attract more students? Apparently not:
There are obviously lots of things that go into choosing a major, so you wouldn't expect grades to be a huge factor. But to the extent they are, Yale students tend to be drawn to the toughest majors. Economics had 149 graduates and the toughest grading. Gender studies had only two graduates even though A's were practically guaranteed.
Things might be different elsewhere with less ambitious students. But at Yale, anyway, nobody seems to be put off by the prospect of tough grading.
POSTSCRIPT: This is solely for Yale College, which doesn't include everything. Engineering has its own school, for example, as does management.
And yet, in all majors, the median grade is an A. That's some serious grade creep.
True.
This guy was the source of the Graham table. The blather is about -- ooh! -- gender studies. But the real story is that the Ivies are an exclusive club. Hard to get into, but cushy once you're in.
Grade inflation was on the rise for years, and the pandemic pushed it into overdrive.
Tecnically not ninjad since I read this before I could post. Would I cop to being prehistoric if I said that back in the day the median grade was a C? This would be in the practically Archezoic 70's.
This isn't how it works. Students aren't drawn to the "toughest" majors. It's that every Tom, Dick, and Harry (and Henrietta) is drawn to majors they think will lead to lucrative careers. So even if they're getting D's and C's in their Econ surveys because they're not that great at it, they'll crawl over broken glass to stick with it and get that degree in the end because they think that's what will get them into an MBA and then they'll be a Titan of Wall Street™. That pushes the GPA average for those departments downwards. Students majoring in Art History or Gender Studies, or Russian Literature are probably not driven by the same motives and are drawn to those fields by some personal interest that pushes them to do well. Also, you're much more likely to get personal attention from your professor and TAs and will do well on papers and exams, leading to higher average GPAs.
Little did I know how the market for gender studies majors would become so luctrative over time ...
The cause and effect goes the other way. Departments without enough students give good grades to attract students. Departments with lots of students are more challenging so they can scare off the weak ones and focus on the strong ones.
That used to be the case in a lot of places. Remember the old professor on the first day of class? "Look to the left of you, look to the right of you. One of you will not be here after the first midterm!" It's a bit more complicated now, particularly with colleges desperate to retain students and in particular 1st gen/students of color who are less prepared than some kid who went to a fancy prep school and took 12 AP classes. Plus, if a department is clearly goosing its enrollments with grade inflation, that definitely comes back to bite them in the ass when it comes time to ask for more faculty or grad student TA-ships or whatever. You can usually spot that in student evals ("Never did any assignments and skipped half the classes -- got an A -- total waste of time.")
Eh. It's more like I am worried about my grade.
This appears to be the percentage of As given out in each subject's classes. That's not at all how I'd measure "how easy is the grading in each major" -- an obvious problem with this is that it includes _tons_ of students who aren't majoring in the subject. So plausibly a subject which attracts tons of students with only a passing interest will have a much lower fraction of As. That feels like math, psychology, and economics to me.
I think it'd be tough to measure how "tough" a major was in a way that was comparable between majors or institutions, but this seems obviously wrong? which I suppose is par for the course for Paul Graham on college education.
From the chart, it looks like it was tied with Gender Studies.
How one would calculate a purported grade average for the major is not apparent.
I assume it would be something along the lines of take all the students receiving a degree in a given "major" and look at their GPA/QPA across their set of completed courses listed as required for that "major."
There are a lot of factors going on here. A department like Art History or Gender Studies is going to have a lot of people in intro classes trying to fill breadth requirements. Those intro classes aren't going to grade very hard. Few people will actually major in those programs, and those who do are likely trying to get into graduate programs. Good grades will matter to them.
Programs like econ and computer science are popular, because people think they will lead to good-paying jobs, so mediocre students want to stick it out. Many people in the major just need the degree, and a B- counts as much as an A for a diploma. Those types of majors also often have an early course that is really tough, in order to weed out some people. That can reduce overall grades in the program.
There are also the existence of weeder courses. In many schools the more popular majors, especially pre-professionals, STEM, and medicals, have intro courses that are explicitly meant to fail a third or more to make them drop the major. Humanities majors have never had such a course, to my knowledge. We WANT our students to learn the material. I don't know how this may or may not factor into this calculation, but it needs to be accounted for.
Also, I teach History of Science. All HoS courses are second year and above and require at least one 100 level history course. My HoS students are just better at being students than my Western Civ students so the former have way more A's than the latter. This is not because one inherently more difficult than the other.
Having been of both sides of a weeder course I've come to the conclusion that they are absolutely necessary. Unfortunately, and unfortunately kinder than any other alternative I can think of.
I love how Yale doesn't include computer science in its college of engineering, yet Iowa State does.
Makes it even more difficult.
The history of computer science departments at American universities tends to be quirky: some grew out of electrical engineering programs, some grew out of math programs, etc. The organizational structure often reflects that.
That said, at the graduate level the CS department at Yale is indeed part of the School of Engineering & Applied Science. The undergraduate program, like all of Yale's undergraduate programs including engineering, is administered by Yale College.
Unless I'm missing something, Kevin has made the rather elementary error of confusing a clearly-labelled list of subjects with majors. Confusion gets piled upon confusion when the Yale College doesn't have "subjects" at all. It offers "courses" within "majors".
There does not appear to be a course named "Economics". However, "The economics major requires courses in intermediate microeconomics, intermediate macroeconomics and econometrics. There are a variety of ways to meet these requirements, and the best option depends on the students’ background, interests, and future plans." Students are also required to take certain maths courses.
Nor, pace Graham, is there a subject named "Mathematics". Per the Yale website, "In total, mathematics majors must complete ten mathematics courses numbered 200 or higher (counting the introductory sequence and the senior seminar)." The situation is even more opaque for the Political Science major, where "Students can declare as a Political Science major at any time during their undergraduate studies. Advisers should be consulted about the choice of courses and developing a coherent program of study."
In short, I have no idea what the grade percentages cited relate to, and Graham's snarky tweet throws no light on the matter.
1) I completely agree with your first sentence -- a point well taken.
2) Your second sentence is correct, I guess, but please note that neither Kevin nor Graham refer to "subjects".
3) As for your third sentence, and most of what follows, I do think you are missing something, or else I am. Like almost all American universities, Yale has *departments*, including (to take your examples) Economics or Mathematics. Departments offer courses, with designations like "Economics 115", so it would be more accurate to say that Yale offers courses in departments rather than courses within majors. Most departments offer majors (usually with the same name as the department). A major offered by a department normally requires a bunch of courses from within the department, sometimes with a few additional courses from other departments, and almost always a bunch of options. (Yes, there are interdisciplinary courses that do not reside within a single department, and interdisciplinary majors, but this is the basic model.) That's perfectly normal across undergraduate education in the US.
The most likely way to interpret the numbers that Graham cites is that they are based on grade distributions for *all* courses offered within each department, irrespective of whether they are required for the major, optional for the major, or even aimed at non-majors. Which brings us back to your original point, which is that grades in courses in grades within the major are not the same thing.
The table Graham published is headed (my emphasis) 'Table 2. Percent A's and A-'s by subject, 2022-23'. It would have been easy to substitute "by department" if that's what he meant. You may be correct, of course, but I note that the Yale Department of Economics "offers three undergraduate majors—Economics, Economics and Mathematics, and Computer Science and Economics—with a total of 250 students graduating every year". This is impossible to reconcile with Graham's inclusion of "economics" as a "subject - enrollment greater than 500". In short, it's impossible to know what his data are supposed to mean.
My apologies -- you are right, of course, Graham does say "Subject" in the table heading, and I missed it.
I imagine -- and of course I could be wrong, as we have now seen -- that he is using "subject" as a flawed synonym for "department or interdisciplinary program". Of the "subjects" on his list, most are pretty obviously department names, but a few (e.g., Education Studies) appear to be interdisciplinary programs taught by faculty from multiple departments. To get down in the weeds for a moment: Every course will have a designation that looks like either [department name] [number] [possibly some other stuff to indicate semester or section] or [interdisciplinary program name] [number] [possibly some other stuff to indicate semester or section]. Most likely the grade data he cites is just aggregated from all courses with the same department/program designation. "Enrollment" in his data then is the result of adding up the individual class sizes of all courses with the same department/program designation, irrespective of the majors of the students taking those courses, and will count the same person multiple times if they take multiple courses in the same department/program. So 250 students majoring in Econ will easily result in Econ enrollment well over 500, since most of those students are taking multiple courses in the department. This is one very standard way of measuring departmental teaching load, etc., in academia.
You are also right that none of that maps directly onto majors. Graham's shot at "math majors" is just snark, and clearly was intended to be taken as such. Kevin is the one who has compared that grade data to major counts, and again I completely agree with your opening main point on how that that's a significant error.
"that he is using "subject" as a flawed synonym for "department or interdisciplinary program". "
That should be enough toskip the rest of what he writes. If you are actually interested in the issue, there are more reliable sources, and if not, why take in information from somebody is that lousy?
It would be helpful, of course, if Graham had linked to the source of the mysterious "Table 2", but he merely links to another Twitter account, which mentions "the report" but doesn't provide a citation. https://twitter.com/sfmcguire79/status/1730568207700771279/photo/1
"another Twitter account"
It changed its name: "another Xitter account".
Correction to the postscript: Yale College is the undergraduate school of Yale University; all undergraduate degrees at Yale are granted by Yale College, including those with majors in engineering, art, and music. (There is no undergraduate major in management.) I do not know why Graham did not include courses in those subjects in his analysis, but it was not because they are distinct from Yale College. The other schools Kevin mentions are all at the graduate level, so not relevant to this topic.
The full list is here: https://catalog.yale.edu/ycps/majors-in-yale-college/
Do grades mean the same in all fields, for example to employers or graduate schools? Evidently they don't mean much in the fields of gender studies or history of science. Maybe all schools are expected to grade less leniently in economics. But how much do grades mean in any field anymore when most people get A's?
I believe that a better analysis would be outcomes (% of students with a job offer at graduation, average starting salaries, % accepted to grad school etc) rather than grades. I am on an advisory board at another selective university (not quite as selective as Yale but a 10% admissions rate...) and career outcomes and demand for the major have a high correlation.
Thus, I bet at Yale the career outcomes for Economics, Computer Science are far better than Gender Studies or the History of Science.
Interesting data. Not sure if it's meaningful without knowing which students are preparing, with an English or History of Science degree, to continue into grad school. I am very curious which majors all the unmarked red dots represent.
Would it be possible to move cat blogging up a few days this week?
I would love to see similar data for a non-selective public college. Nobody gets into Yale without being an academic star in high school, so Yale students are probably less likely to be scared off by tough grading than the much weaker students you'll find at unprestigious institutions. With that said, I'm inclined to agree with previous commenters who argue for reverse causation, with professors in popular majors feeling like they can afford to be tough while the departments that have a hard time attracting students give easy A's because they can't afford to lose anyone.
Now do male vs female in the hard vs easy degrees...
This is not just easy snark.
The pattern seems to be that plenty of boys (substantially fewer females) relish the very idea of a challenge, or are simply so interested in what they're interested in that they don't care about minor details like grades.
Whereas the bulk of female students seem to care about the penumbra of social concepts around a degree ("respect", "get a job", "get parents' approval") in a way that is a lot more grounded in grades and succeeding in eventually graduating.
So of course one way to view degree inflation is catering to the influx into the academy of a new population who care much less about the original goals of the academy ("pursuit of truth" no matter how difficult it might) and who care a lot more about the social side effects of the academy, of being able to boast of being an academic (or at least having a degree).
Do you have any evidence this is true or is it just your misogyny speaking?
Do I have evidence that the sky is blue? No I don't.
Which is the part you find so hard to believe? That men (as a class) take more risks than women? Or that women (as a class) obsess over social status more than men?
Do you SERIOUSLY think that I'm wrong? You know, when you're not virtue signaling to your friends? Look at those courses and tell me that you don't expect 90% of Physics, Math and Econ to be male. And Gender Studies and English to be Female.
All this data proves is that the quality of instruction in math and sciences also sucks at Yale, and hence the low grades, just like in less reputable institutions ... and that meaningful learning happens more in the humanities and social sciences.
If you take out Economics, I don't think the trendline even has a visible slope. Economics is an outlier and there is no relationship between grading and enrolment.