Hacker News Reader

Comment by wpietri | Hacker News Reader

This is kind of a weird mix of points. Ignoring his political axe-grinding, I think the value of a college degree is increasingly in question for many fields because the price has risen wildly for decades without a corresponding increase in market value. (Except perhaps for elite-institution degrees, which are more about the brand and the network than what people actually learn.)

I think we haven't seen at least a partial collapse only because most American companies are bad at hiring, bad at investing in workers, and bad at keeping them. But imagine a company where programmers are happy and tend to say for years. That company might do just as well, or perhaps better, running an in-house boot camp and apprenticeship program as by hiring new grads.

View on HN

Question: do elite institution degrees confer any addition to market value commensurate with their cost, or do they simply select for people who already have high market value?

Good question. This has been studied a lot and the answer is…unclear.

I think the kind of longitudinal studies needed combined with a small “treatment” group make this kind of investigation as hard to do as a nutritional study.

It is not the degree that is valuable, it is the network it allows you to access that is valuable. The classmates you have at certain institutions will increase the probability of you achieving certain goals.

The thing is that you frequently end making connections with alumni of those institutions in your first few corporate jobs, since they all recruit out of the same pool of elite universities. So even if you didn't go to Stanford or MIT, if you work at Google (well, in the 00s - it doesn't focus nearly so much on elite universities now) you will know people who went to those schools, and if you need to make a connection with another alumnus of them you can reach out to your friends and say "Hey, can you connect me to your classmate?"

> "Question: do elite institution degrees confer any addition to market value commensurate with their cost, or do they simply select for people who already have high market value?"

Speaking only for STEM, an elite STEM program gives the student access to a bigger variety of upperclassman courses; compare the available course list for a community college vs a large university and the difference is quite noticeable. The better programs also provide access to better equipment and advising; e.g., various University of California campuses have their own on-site chip fabrication labs. If the career trajectory you're planning benefits from those advanced courses, going to the better STEM programs is going to help tremendously. If not, well, they don't but that's hardly the institution's fault; they aren't choosing who to send your resume to, you are.

(As an aside, I use the word "access" very intentionally; educational institutions don't "confer" anything. Students get access to resources and opportunities and they choose which to take. It's entirely possible to do the minimum to fulfill requirements; for example, I chose easy courses to do the minimum fulfill my humanities requirements because it wasn't where my interests lay.)

I'm sure it depends some on the degree. A very sharp friend of mine got an aeronautical engineeering degree from a good state school. After a few years in industry he went back for a brand-name MBA. He said in 2 years of school, he learned exactly one thing that he couldn't have just figured out via general knowledge and a bit of thinking. [1] He said the real value was in the network he built up, people who would soon be placed in important positions in important companies all over.

[1] For the curious, it was the notion of comparative advantage, which was counterintuitive for me too: https://en.wikipedia.org/wiki/Comparative_advantage

And I should add that in an age of increasing inequality, a network with a lot of elite members will increase in value much faster than inflation. So the value of an elite degree might still be a bargain even at the current prices. Which probably explains why there's a lot of bribery, legal and otherwise, by the rich trying to get students into brand-name schools.

Elite institutions could do nothing and still add market value simply by colocating people with high market value who meet each other and create lasting networks and friendships.

I do think that on top of that, they add value because students can be given more challenging and stimulating coursework. The difference in Math education between a top college and a college with loose entry requirements is staggering - we're talking people doing hard proofs freshman year vs. undergrad seniors taking upper level math courses that don't even have proofs, just computation.

Private elite institutions (Yvies) offer excellent education, and networking with really rich people ;). Some people already had that, some will make economic use of it, and some not. I assume the variance within-group is very very high, so you can probably calculate stats to defend any point. And then the cost also varies, so ...

'Elite' state schools (GA Tech, Michigan State? Etc) probably have excellent ROI.

Expensive, not really elite schools are probably not worth it on average, but again, cost varies highly, so ...

It’s been studied some and there's at least some substantial evidence that it's the latter. (I posted this on another thread last week.)

> For most students, the salary boost from going to a super-selective school is “generally indistinguishable from zero” after adjusting for student characteristics, such as test scores. In other words, if Mike and Drew have the same SAT scores and apply to the same colleges, but Mike gets into Harvard and Drew doesn’t, they can still expect to earn the same income throughout their careers. Despite Harvard’s international fame and energetic alumni outreach, somebody like Mike would not experience an observable “Harvard effect.” Dale and Krueger even found that the average SAT scores of all the schools a student applies to is a more powerful predictor of success than the school that student actually attends.

https://www.theatlantic.com/ideas/archive/2018/12/does-it-ma...

That’s not “education” directly, but does suggest an extremely strong applicant pool selection effect is happening.

I'm not familiar with this study in particular, but there's another similar one for the elite NYC public high schools, which have historically had a similar "halo effect" attributed to them. They found that the student cohort that "barely made it" had similar long-term outcomes as students who "barely missed out".

I came away thinking that parental influence is the most important factor. If your child is well-nurtured enough to be within a crap shoot of selection for an elite institution, it probably doesn't really matter if they get in or not, they'll be successful regardless (for mysterious reasons that may be interesting to know, but don't really matter).

As such, it's better not to make your kid focus on "getting in" (which is a suspect goal for many reasons), but instead on trying their best to realize their full potential.

One dimension: In my experience one of the most valuable aspects of attending an elite institution is that many of the most prestigious, high-paying, career-accelerating jobs in the US recruit primarily or exclusively at top colleges, and this creates numerous opportunities for their students.

On one hand, the exclusive job opportunities are providing value to students which can allow students to increase their value. On the other hand, the employers are benefiting from the fact that the universities themselves are selectively gathering people with high market value. Exactly how much of whether it's the former or the latter is an open question. Whether or not a university is an expensive platform to pass a corporation's gatekeeping is questionable.

My impression is that elite institutions generally do not have a higher cost than the reasonable alternatives, so there's no need to justify if it's worth paying more for them as you're not paying more. Their credential value is strongly linked to their selection, but if they do select you, studies there can be more affordable than a "lesser" institution, especially because they are mostly funded not by your tuition fees but rather by the endowments of their alumni, so indirectly through your future income.

The highest cost seems to be for good-but-not-best institutions that have some specific niche/selling point for them but are not particularly selective and will accept almost anyone who will pay their price.

Like, if some school wants to build a reputation for having the best scientists or lawyers then they have to be somewhat selective with respect to performance and have very competitive admissions, but you also need to ensure that the particular young students with most potential can actually afford to study there; but if a school just wants to provide a good education in X and does not care who will enroll, they can just ramp up the fees until the applicants filter themselves out.

A degree doesn't cost the company. An in-house bootcamp and apprenticeship program does. Once the less-legibly-promising new hires are equally productive, they now have a resume making them legibly good to the company's competitors in the job market, so it can't make up its investment by underpaying. ("Programmers are happy and tend to stay" is something the competitors can do too.)

I'm not saying don't do it, or that the status quo is good. But I don't think your suggestion really addresses the problem that got us here, the problem that we've subsidized an expensive signaling game (cf Bryan Caplan's The case against education).

> Ignoring his political axe-grinding

This was gratuitous. I didn't even see any politics in the thread.

> A degree doesn't cost the company.

Degrees cost companies through increased wages. Student loans get paid somehow, after all.

> This was gratuitous. I didn't even see any politics in the thread.

You not seeing things is not the same as things not existing. I do see it, and wanted to focus this bit of discussion on the question of degree value, not the assorted other stuff there.

> Degrees cost companies through increased wages.

See https://en.wikipedia.org/wiki/Signalling_(economics)#Job-mar... for the model behind my comment.

You’re a hiring manager choosing between Alice and Bob. Alice has a degree, Bob doesn’t; you estimate they’ll be equally productive, and it’s common knowledge that your competitors will judge them the same. Do you offer Bob less, since he doesn’t have student loans to pay off? You can if you like, but then Bob will go work for your competitor.

> You not seeing things is not the same as things not existing. I do see it, and wanted to focus this bit of discussion on the question of degree value, not the assorted other stuff there.

You can mention having further disagreements without making it a gratuitous diss.

Quite a lot of people will indeed pay Alice more, especially if her degree is from a brand-name school. You can look up the stats yourself, but they're striking.

I personally build my hiring processes to be relatively blind to elite status signals. But interviewees tell me that's pretty unusual.

> You can mention having further disagreements without making it a gratuitous diss.

I believe it was both accurate and necessary to divert the replies away from the politics. And given that you can't even find the politics, I don't understand why you consider it necessary to appoint yourself my editor.

You misread: given common knowledge that Alice and Bob are equally promising, all info considered, do you make them different offers?

I'm not at all asking whether Alice with a degree gets better offers than Alice without a degree. We both say she does. At the time we're checking in, Alice's degree is a sunk cost. Your proposal is not: it's extra company investment in the non-degreed after hiring.

> can't even find the politics

Are you sure you want to emphasize anyone's reading ability in this conversation?

And accusing someone of political axe-grinding is a funny way to discourage replies about politics. You could've said, for instance, "Leaving aside any politics here..." That would help engender healthier discourse here on HN.

I didn't misread. I chose not to play your contrived game.

I also didn't hire you as my editor, so perhaps you can just apply your advice on comity to your own work and leave me be.

So contriving a simplest concrete example, after the abstract explanation apparently failed to get through, is a "contrived game".

Comity is not the same as healthy discourse. Objecting to a cheap shot is not hiring myself as your editor. I do agree on ending the thread here.

> Once the less-legibly-promising new hires are equally productive, they now have a resume making them legibly good to the company's competitors in the job market, so it can't make up its investment by underpaying.

It's not like other industries haven't already solved this problem. Law firms will finance your law school and make you pay them back if you leave the firm.

Maybe ... I think IBM, for example did, and others might.

OTOH, it may be easier and cheaper to make an agreement with their local community college.

I think what makes the discussion hard is that people say 'college' and mean wildly different things. The author seemed to mean elite colleges in some comments, overall statistics in others, and for-profits in others. I think there are very different problems in different sectors.

> Except perhaps for elite-institution degrees, which are more about the brand and the network than what people actually learn.

Even for a place like MIT/Caltech?

Yes.

Even the way you ask the question shows that you consider them somehow different. Probably better. Which presumably means that "better" people go there or teach there. So that's better networking.

And the fact that you know them by name as better places shows better brand.

I'm increasingly loathe to hire a new college grad for enterprise software positions due to four years of terrible instruction that we have to deplorable. What are they even learning in modern day institutions?

Meanwhile, I have zero qualms about hiring high school grads with a github who can talk about the base language intelligently, let alone frameworks like spring.

Price and value are decoupled for college degrees. STEM students are subsidized by liberal arts students, and many liberal arts degrees are not, by themselves, economically valuable (I said this as someone with just such a degree).

> STEM students are subsidized by liberal arts students

Can you provide more context for this? How is the cost of teaching a STEM student higher than the cost of a liberal arts student? The classrooms are roughly the same. STEM student labs might be more expensive to manage, but that equipment is/(can be?) funded by research grants.

STEM professors have to be paid a lot more than liberal arts professors (due to competition for STEM talent). Equipment for teaching isn’t covered under research grants, but might come from overhead on those grants, but more likely from donations (from big corps), grants, or tuition.

The national center for labor statistics says labor costs are about 30% of the spend [0]. STEM programs also include Liberal Arts courses, so only a fraction of the 30% goes to STEM professors pockets. AFAIK, STEM professors justify their value by generating money for the university via patents and grants.

While there are a massive number of private liberal arts schools, there are relatively few private STEM schools. Most STEM degrees come from public university that are significantly cheaper than private universities.

[0] - https://nces.ed.gov/fastfacts/display.asp?id=75

STEM professors bring in billions of dollars in research grants, of which the university takes over 50% for "administration". If you aren't aware labs have to pay a big cut of any grant income they get to the school. It used to be that in exchange the school would maintain buildings, fund build outs of equipment, etc. but nowadays it's wasted on DEIABCDXYZ vice chancellors of provosts.

> of which the university takes over 50% for "administration".

Hence why I said "overhead from research grants."

Business professors are paid even more.

Research grants pay for research equipment. They do not cover student labs in any way.

Yes, they do.

(1) Many grants have a mentorship / training / giving-back component. This is one of the merit-based criteria NSF reviews on.

(2) Research grants have overhead which feeds back into general budgets. At elite schools, this is about 2/3 of the money. A typical split might be 1/3 to the department, 1/3 to the school, and 1/3 to the project. It's kind of a financial scam. Nominally, these cover buildings and admin time. Practically, these feed into general budgets which do include labs and teaching. Corruptly, a lot of the money gets funnelled in creative ways to improve professor's lives through fancy faculty clubs, get-aways, and in some cases, creative (but legal) embezzlement with money ending in people's pockets.

It is largely anecdotal based on some inquiries made at my school. Different schools will vary- is it a research school? What types of STEM degrees are offered? Etc. Others likely have more concrete insight than I do.

One conversation that stuck was the me:

Back when I was in school, one of my mandarin teachers wanted to offer a "business mandarin" course outside of the general language program geared towards business students who might need basic fluency in a corporate setting. The school denied him because there wasn't a budget for it, which struck me as asinine at the time as it wasn't like there would be students who weren't paying tuition for the course anyway. The school had rooms available, and students have to buy all the materials the class would need anyway.

Would offering that course have spread the same n students across m + 1 courses? (Hence an increase in instructional costs, but no increase in tuition revenue.)

This works fine for research universities, but in teaching universities (couldn't find the statistics, but there are a lot of them out there) someone has to subsidize the new STEM programs until the alumni start giving grants. In my experience, a lot more STEM alumni give large donations than humanities and social sciences. But that does mean a bigger initial outlay by someone.

Reply on news.ycombinator.com