Comment by mjn | Hacker News Reader

Comment by mjn | Hacker News Reader

Parent Full thread

Depends on whether they actually do research in the private sector, though. Historically most fundamental science advances have come either from academia, or from big academia-like research organizations in industry (Bell Labs, Xerox PARC, Microsoft Research), or from government research labs (Sandia, Livermore, CERN, etc.). Most industry R&D tends towards different kinds of problems, applied research with 1-5 years commercialization horizons, sometimes pushing 10. That's a valuable segment of research, but not all of it, and it depends on a pipeline of less-applied research to keep it fed.

But due to the difficulty of monetizing indirect future advances that basic science enables, it's difficult to do it profitably in industry. A scientific discovery that at some point in the future turns out to be useful in developing technology is valuable for humanity, but you cannot easily capture that value, as it simply diffuses to the general benefit. Knowledge advances, but anyone can use it, and you don't get to own it; you can get some kind of satisfaction in old age in looking at how this physics discovery you came up with in your 30s is being used to improve chip fabrication processes 40 years later, but you don't get royalties on the chips. Patents try to change that within a 20-year timespan, but they are not supposed to apply to scientific or mathematical discoveries, 20 years is too short for some advances, and they have a whole host of problems anyway.

I guess I'd divide the research world like this:

1) Pie in the sky research that is beyond the 20 year time horizon of long-view companies and 10 year time horizon of VCs.

2) Research that is amenable to teaching people how to research.

3) Research to advance one's career.

4) Research that is relatively close to Development.

There are overlaps, and it can be hard to differentiate between the types until after the fact, but let's go with this model.

#1 has to be funded federally (i.e. NSF), via philanthropy (i.e. Gates Foundation) or as a monopoly concession (i.e. Bell Labs). This can belong in the University or elsewhere, and may or may not be tied to teaching.

#2 belongs in Universities, as it helps generate the next generation of scholars.

#3 should not be financed by taxpayers. It's also hard to justify this in aggregate coming from tuition. Do we really need more papers on Ancient History, or should we invest the same money in better teaching? I understand you need to be an expert to teach, but it seems like a lot of publishing is purely to generate citations.

#4 If these people can get real world jobs, why not send them to the real world where rather than stay in a profession with very scarce jobs?

> Do we really need more papers on Ancient History, or should we invest the same money in better teaching?

This seems to me like a pretty dim view of the humanities in general. Research in the humanities is not in the same category as research in Computer Science, for example; there is often little to no free-market value. People who enter the humanities tend to do so (at least in my experience) because knowledge for the sake of knowledge is something worth pursuing. Your next company might not make a fortune knowing more about agricultural practices of the ancient Macedonians, or the concept of fatherhood in the music of Robert Schumann, but people doing that research aren't doing it simply "to advance their career," and not necessarily just to teach new people how to research either.

It has traditionally been the function of universities to foster this sort of research, as a general goal of "making the world a better, more knowledgeable place." It's a societal goal that goes way back: people say that humanity was set back something like 1000 years when the library at Alexandria burned down. This type of research doesn't fit neatly into any of your categories, and it is (I think) a more trenchant problem. Disillusioned humanities PhD's often have problems finding work in the "real world," since companies aren't keen on hiring someone who knows everything there is to know about Macedonian agriculture.

-- a fairly disillusioned Ph.D. student in Music Theory

I have a dim view of research in the humanities. I have a strong view of teaching in the humanities.

In my mind, much of engineering, computer science, math, and non-lab science can be efficiently taught online. Probably a lot of business school coursework too. For at least the intro and intermediate courses, it's about having a body of material that can be mastered. Discussion isn't as important if you can learn the material on your own.

Comparative literature, history, philosophy and similar subjects are very important, and are more likely to require live in-person teachers. Discussions are important. Subjective grading and coaching of individual papers is important. Multiple choice exams are less likely to work.

I'm also a strong believer that humanities degrees complement professional degrees. An accounting major can do my taxes. An accounting and English major can break arcane financial reports into plain English for investors. A math major is ince. A math major who studies French literature may also be able to read French proofs in their original language.

What I question is the relative value of the teacher's research. What's the right break-up of teaching to research time for humanities? Is it the same as the sciences? Is the current ratio correct?

I believe that the current ratio is not correct. One needs only to look at how expensive small liberal arts colleges are. If their teachers spent more time in the classroom and less doing arcane research, the costs would come down. Perhaps if you're a top 10 school it's ok, but not as a lower tier school.

I have a right to complain as a parent, as well as a taxpayer whose government is subsidizing student loans and paying for schools.

Now disillusioned Phds is another problem... Systemically everyone is incented to produce more Phds than required. It's definitely a situation of letting the buyer beware early on.

> One needs only to look at how expensive small liberal arts colleges are. If their teachers spent more time in the classroom and less doing arcane research, the costs would come down.

But small liberal arts colleges are the places that already do primarily prioritize teaching, and expect professors to do relatively little research. Professors typically teach more courses per semester than at research universities, and are hired and promoted primarily based on teaching ability rather than research portfolio. This was the case at the Claremont Colleges, at least, where I went for undergrad; most professors just did a little research on the side, maybe one paper a year, and spent nearly full-time teaching. They're expensive mainly because they have very small class sizes (big emphasis on small discussion seminars), and typically much smaller endowments and fewer donations than the famous research universities do. Plus, since none of the salaries or lab facilities are paid for by research-grant money, those have to come out of tuition, too. And, since there are no grad students, professors typically do most of the teaching work themselves, rather than offloading a significant proportion to a TA, as is done at research universities: sections, tutorials, grading, office hours, etc. are all the professor's responsibility.

Big research universities, like MIT or Stanford or University of Texas or UCLA, are the ones that primarily emphasize research track record.

This has been my experience as well. I went to a small liberal arts college for undergrad. My teachers there did very little research, but all taught their small classes by themselves with no TA.

My particular field (music) tends to be much more expensive per student since private lessons must be taught one-on-one. One physics professor could teach 15 students for three hours a week, but a piano professor needs essentially needs 15 hours to teach each of those students for one hour a week. Some of this work can be done in small groups/classes, but higher music education is much more individual.

Some of my colleagues who have graduated now teach at liberal arts schools where their tenure requirement is essentially one presentation at a regional conference or one publication, with a strong record of teaching. Compare this to a large institution, where the expectation has traditionally been that you need a book deal in order to obtain tenure, whether or not you're much good as a teacher of undergraduates.

Point well taken. I appreciate your insight.

So here's a secondary question... Could the overall cost of the education be reduced if the intro classes were taught online? Or the non-humanities intro classes taught online?

As the other poster has mentioned, teaching is part of what makes these colleges expensive.

Teaching heavy departments (the humanities) have one source of income: student fees

Research heavy departments (science, engineering, medicine) have both student fees and grants - not only to pay salary, but also to cover things like keeping the lights on, computing power, administrative support, etc. This is often a substantial amount of the total cost of a grant.

Going to pure teaching means everything has to be carried by either student fees or tax payers (for public institutions).

I appreciate your insight.

Let's do a very back of the envelope oversimplified calculation...

Let's say assume the following: - Semester system - Each teacher teaches 5 classes per year (2 one semester, 3 the other) - Each student takes 10 classes per year - Each class is 20 students (assume very low - liberal arts college) - Each teacher earns 100K in total comp - Each teacher requires an assistant making 40K in total comp - Each teacher and assistant pair carries an admin making 100K in total comp - Each course costs $5K

The math I get says: - 50K total tuition per student - 20 students = $1mm revenue - 20 students = 200 course-students = (20 teachers/class * 5 classes * 2 teachers) = 2 teachers + 2 assistants + 2 admins - Staffing cost = 200K + 80K + 200K = 480K

In this oversimplified model of small classes, tuition covers the teachers, their graders and admin overhead 4 times over. Are buildings really that expensive in rural America?

Even if all the classes are 10 people each the #s don't add up. Are people really getting 50% discounts? Are football stadiums for these colleges so expensive?

The astounding ratio I follow is tuition divided by fully loaded faculty salary. If it's 1/2, that tells me something is wrong in how the faculty are spending their time.

Working on mathematical models for a living, if the number at the end comes out as "That can't be right..." odds are its a problem with the model, not the system it's modeling. You've made a lot of simplifications, both on the income and expenses end of your envelope math that I very much doubt are correct.

Universities, even small ones, are fairly large, complex organizations that are extremely capital intensive. This is complicated by the fact that many of the institutions we're talking about are private, so we don't have many numbers.

But to the major point of moving toward teaching, and away from research, we can look at the balance sheet of a large state university system:

For 2012, the UNC system took in $320,535,000 in student fees and $725,846,000 in grants and contracts. While UNC is a fairly good deal for it's students (if you're in-state) its hardly free, and grant income outpaces student fees by a healthy margin.

Do the grants include general state support for in-state students? Since it's a public school the math I laid out would change. Certainly much larger average class size.

To prior points, it is hard to get to the bottom of financials. If like to know where all the money goes if bit to teaching.

What do you see as the value-add to taxpayers of humanities degrees?

Would a patronage system work better for these types of degrees?

The value-add to taxpayers is much less concrete in the humanities. My argument was that research in the humanities is essentially done for the sake of doing research: to learn more about the world and the people in it. It is hard to see definite, monetary benefits, which is one of the reason arts/music/creative writing programs are being slashed cut in favor of match/science in primary education in the US.

The benefit of studying the humanities is one of personal edification. Students who study art or music in school are likely to be more disciplined and score better on standardized tests than their non-musician/artist colleagues. Not to mention, young students who have a strong background in the humanities are often happier with their quality of life (I know that music was the highlight of my day/week when I was in school).

A patronage program would be great for this type of degree...if any VCs out there want a personal music theorist I'd be happy to apply! My argument is that it is traditionally universities which have been the patrons of these degrees. This goes back at least to the earliest universities in Europe, where students learned the seven liberal arts: grammar, logic, rhetoric, arithmetic, geometry, music, and astronomy. The idea of the university was not to churn out good economic workers (as it tends to be today) but to produce well-rounded human beings. We in the humanities still think this is a worthy goal (and consequently, tend to make much less than our STEM counterparts).

You left out DARPA, which is the source of all things internet.

Reply on news.ycombinator.com