The reverse is already happening although folks don’t like to admit it.
Freshman Bio classes- 500 students in a lecture hall with one professor, and review sessions taught by TA’s. Freshman “Understanding Socrates” class- 19 students, one professor, every review session taught by that same professor who also maintains open office hours.
The folks on CC who complain about large classes clearly have never looked at the enrollment in “Giotto and his contemporaries”, “The role of dissident literature in the dismantling of the Iron Curtain” or “Ugaritic and Akkadian Flood narratives”. You want personal attention, a professor who knows your name and how you like your coffee, and will help you pick a research topic, suggest extra reading, review your outline, pick up the phone to get you a job as a research assistant/editor for a colleague? Major in the humanities!
Well, I’m a data point of one, but this English major adapted her skills to a technical career but later parlayed them into upper-level management. It was always the communication, writing, and problem-solving abilities that served me best, and I always looked for those first when hiring. Among those in the C-suite, the humanities are not underrepresented.
You make (always) great arguments - and your knowledge of things like this is second to noen.
But is this happening - because there are less and less of these classes and professors, etc. best I can tell.
But yes, at any school, you can find those majors or even classes with smaller student counts - as they post the max sizes and spaces available for all to see - but some also get canceled if too small.
It amazes me people fall all over themselves to go to a UCLA, as an example and have that exact experience you just mentioned.
And students from wealthy families have less pressure to choose a college path prioritizing post graduation job prospects to pay off their student loans. Of course, labor markets and industry conditions can change significantly in four years, so choices made at college entrance may not end up with the desired results at college graduation.
This is partially mitigated by having wildly varying class sizes. At colleges like Stanford, Duke, or Harvard; it’s common for CS classes to have hundreds of students. The more popular intro classes or AI classes may have near 1000 students. In contrast, the vast majority of in-major classes in a less popular humanities major may have under 10 students.
Of course. The point is simply that the humanities aren’t hampering anyone from getting where they want to go in life and making a good living, whatever that looks like for them, including the “top” jobs.
The link earlier link was from 2022. Looking at the 2025, I see the following highest ROI LACs. Washington & Lee is at $2.9M, while other comparably selective LACs are usually on the lower end of $2.5M+. It’s a statistically significant difference, but not a dramatic one.
Highest ROI LACs: Georgetown Link
Harvey Mudd – $4.6M (2nd highest college in USA, after St. Louis Pharmacy)
– Gap –
CMC – $3.3M
Lafayette – $3.2M
Colgate – $3.0M
Bucknell – $2.9M
Washington & Lee – $2.9M
Colby – $2.9M
Holy Cross – $2.8M
Amherst – $2.8M
Swarthmore – $2.8M
Other Highly Selective LACs – Usually $2.5M+.
I believe the source is CollegeScorecard earnings. Unfortunately sample size is too small to break down earnings by major with much granularity for LACs. Among the few majors that are available, the listed earnings by major at W&L are below. Econ appears to be dramatically pulling up the average, likely through “elite” finance and consulting, while humanities appear to be pulling down the average. Among the College Scorecard sample above, the most popular W&L majors were the 2 highest earning majors – econ and business. I suspect that W&L hitting above selectivity class primarily relates to a different major distribution and corresponding different career path distribution, in CollegeScorecard sample.
I have read that computer science grads are not doing well in the job market. Science PhD’s are also facing challenges. How will schools respond to these more recent developments? Will students move away from CS and schools start hiring adjuncts for those classes as well? Will departments shrink? AI will change that field very quickly.
Fair point. On another thread yesterday someone noted that Northeastern changed its curriculum, eliminating or changing classes first years were taking. Inevitably I think it’s all about money. Have the numbers of students chasing CS majors started to fall ?
When that happens, then yes schools will react. It’s typically but not always (like the NU curriculum) reaction vs being proactive, like much of society typically is.
I touched on this in my earlier post. Stanford shows stats for number of students enrolled in CS major rather than completed degrees. As such, Stanford shows trends in majors before other colleges, with changes in the pipeline before degrees are completed. At Stanford enrollment in CS majors peaked in Aug 2023 at 865 (11.0%). In Aug 2025 the total dropped to 639 (8.8%). Aside from COVID 2020, this is the lowest CS enrollment in the past decade, less than enrollment in 2015. CS still has 3x the enrollment of any other major (2nd place is econ at 231), but this is the first time it’s been on the decline since the dot com crash period. Other colleges are also starting to show a decrease in completed CS majors for the first time since dot com crash effects, in 2025.
In think the rumors of CS majors not doing well in the job market are greatly exaggerated. Post grad outcome reports at highly selective colleges that are regularly discussed here show relatively little change from previous years, still generally showing the overwhelming majority of CS majors employed (or in grad school) and median earnings above all other majors. However, there are more comments about CS majors applying to a large number of positions before getting their job or specific companies that are higher fewer CS majors than in the past. I expect these types of rumors and fears of AI effects contribute to the decreased enrollment.
As an example, I see than CMU just added 2025 to their post grad outcomes dashboard. A summary is below. Other less selective colleges I checked show worse stats than CMU. However, like CMU, they did not show a notable decline from 2023/2024.
CMU CS + AI Majors
2023 – 4% Seeking, Average Salary = $145k
2024 – 2% Seeking, Average Salary = $150k
2025 – 2% Seeking, Average Salary = $155k
Welcome to “Recruiting Land” where companies have cut their targets for summer interns, new grads (both Bachelor’s and Masters) AND what is often referred to as “EP’s”– Early Professionals, meaning folks who are not coming directly out of campus, but who have worked one or possibly two full time jobs already. Targets cut. Not trimmed, cut.
My educated guess is that you will NOT see this showing up in the numbers (yet) at the top tier U’s. You have already seen it at one notch below, and are seeing it aggressively impacting numbers at the “very nice but not especially competitive” CS programs which proliferated in the last 20 years.
I think I was at a meeting in 2012 or so and heard a presentation on the difficulty colleges were having hiring CS faculty. There was a list of about 20 colleges that were trying- desperately- to fill their open slots. But the salaries in industry were so high (base, bonus, options) that even PhD’s that wanted an academic career were opting for industry– “how can I turn down that much money?”
My guess is that colleges will NOT have trouble filling faculty positions in CS going forward, no matter how esoteric the specialty. My guess.
CMU is not a benchmark program anyway. They’ve got deep interdisciplinary connectivity across robotics and all the engineering disciplines, so their BS grads have the luxury of pivoting professionally. Newer programs with less depth are watching their grads get jobs in IT- plain vanilla IT- help desk, integration, etc. and breathing a sigh of relief that these kids are employed. Getting a job at an insurance company trouble-shooting when a customer is having trouble with the online claims system is called- employed. It ain’t CS. But that kid doesn’t show up in the dreaded “still looking for a job” category.
Are there any “one notch below” or “very nice but not especially competitive” CS programs that are showing noteworthy declines in CS employment stats from previous years? If so, I haven’t seen them. As I touched on above, Class of 2025 stats are still limited to only a small handful of colleges, but that small handful doesn’t seem abnormal to me. An example from UMass Manning College of CS & Information is below. The higher % seeking that other colleges partially relates to capturing stats on a rolling basis rather than at end of x months, but my point is 2025 doesn’t stand out notably as worse employment outcomes than usual.
2023 – 17% Seeking, No stats on salary, Most common employers = Amazon, Dell, Liberty Mutual
2024 – 24% Seeking, $82k to $110k “typical”, Most common employers = Amazon, Microsoft, Liberty Mutual
2025 – 16% Seeking, $93k to $105k “typical”, Most common employers = Amazon, Fidelity, Travelers
Founder majored in English literature, says technical background needed may fall sharply and
“The things that make us human will become much more important instead of much less important,” she told ABC News. “And what I mean by that is when we look to hire people at Anthropic today, we look for people who are great communicators, who have excellent EQ and people skills, who are kind and compassionate and curious and want to help other people.”
Interesting. Time will tell. Of course some humanities departments are already gone!
Creative problem solving, managing people, and coordinating projects should continue to be highly valued skills. However, all of these skills will require a significant comfort level with technology. Maybe the key is to integrate soft skills instruction into STEM courses & majors. Or, as I suggested above in an earlier post, to strengthen distribution requirements to make sure that all essential areas are covered.
The need to adapt quickly and to implement & manage projects efficiently are likely to remain important functions for both humans & technology.
UF 2024 57% employed, 32% seeking, 11% will look after graduation. Mean = $94.2K, Median = $87.5K. Shows 87% in programming/coding - the things we are AI will take away.
UF 2025 54% secured, 35% looking, 11% will look after graduation. $99.7k mean, Median $90.3k. Shows 87% in programming/coding
Don’t know when the 25 snapshots are vs. 24s so maybe that explains the slight drop in 25s - but the money has not dropped off.