Why does college cost so much?

David Leonhardt serves up a dialogue with Robert B. Archibald, and also David H. Feldman.  Archibald starts by citing the cost disease and also the heavy use of skilled labor in the sector.  I don't think they get to the heart of the matter, as there is no mention of entry barriers, whether legal, cultural, or economic.  The price of higher education is rising — rapidly — and yet a) individual universities do not have strong incentives to take in larger classes, and b) it is hard to start a new, good college or university.  The key question is how much a) and b) are remediable in the longer run and if so then there is some chance that the current structure of higher education is a bubble of sorts. 

I never see the authors utter the sentence: "There are plenty wanna-bee professors discarded on the compost heap of academic history."  Yet the best discard should not be much worse, and may even be better, than the marginally accepted professor.  Such a large pool of surplus labor would play a significant role in an economic analysis of virtually any other sector.

When it comes to solving the access problem, the word which pops up is "financial aid," not "increased competition."  Why might that be?

Matt Yglesias once had a good post on how innovation in higher education may come through the proliferation of cheaper and "inferior" alternatives; more on that here.  And here is more from Matt.


In a TGS world, the returns to certification are higher because the opportunities for labor are lowered.

Some people, ahem, believe so strongly in certification that they aren't going to like giving financial aid to an unproven college even if it increases the possibility of undermining status and increasing equality.

Too much administration (very, very expensive with little value).

Too much subsidy of private businesses. We don't educate as much as we do offsite R&D for businesses which, for a small contribution, get the use of the expensive facilities, grad students, etc., without paying the fully loaded costs. Best deal a company can make today is to align with a university, maybe pay for a wing, and direct the research.

Too little use of technology, which threatens faculty. I've offered to do a video grad program using a well known media producer (offered services for free) and the reaction was: this will take away the need to come to class and pay the big bucks. They could use the program if I weren't there.

Too little competition. Once you get tenure, sweet. And, if the French language prof doesn't have enough students, and the University should expand the Chinese department, too bad. You can't adjust the supply to meet the demand because of tenure.

Too many meetings. If you don't have a busy teaching schedule, and if you show up to work (rather than "working" at home) you can always stay busy attending meetings.

Too many non-academic amenities. Universities build swimming pools, rec centers, colisseums etc. to attact students who choose schools based on the amenities.

Too much pushing and selling debt on students. Because they want to reduce price sensitivity and sticker shock on students, schools push debt.

This all will be irrelevant in 20 years when we will be exporting students to India and other developing countries where the cost will be cheaper. Now, if we could just get the AMA to support medicare payments for foreign delivered healthcare.

Our society has it drummed through everyone's head that if you do not Go to College you're some sort of grunting primitive savage. This keeps demand high and allows colleges to boost costs as they see fit.

I think the major cause is different from all of the above: the federal student loan guarantee.

Because of the guarantee of student loans, banks can without any risk extend absurd amounts of credit to students, who generally have no idea what is in store for them on the other side of $100,000+ in non-dischargable debt. If your customers have unlimited liquidity and no sense of what kind of debt they're taking on, it's easy to jack up your prices as high as you'd like.

Does college cost "so much?" The average tuition at state four year schools is under $8,000. About a third of students graduate without debt. The average for those who do have debt is $20,000. Looking at $50,000 tuition and students who owe $100,000 and then asking why college costs so much is like pricing Ferraris and asking why cars cost so much.

Higher education and health care should be the same: government-controlled, free, and rationed.

I found this report very interesting. Its points have been touched on in a couple of the above comments, but I think the link is worth a look.

Administrative Bloat at American Universities: The Real Reason for High Costs in Higher Education

I think Peter and Brad are on to the core issue. I think Peter's point ties into the financing aspect in helping drive costs up. Brad's observation seems consistent with the idea that college is as much about image (prestige) as actual education.

There's an old article written back in the early 1930s raising the same question as Peter implies with his "drummed through everyone's head". It's by Max McConn, who was the Dean of Lehigh University writes "It was taken for granted that every boy and nearly every girl whose parents could possibly afford to send them should go on to college. But, unfortunately it has become apparent that there are difficulties with this 'new minimum standard for the great mass'". He goes on to say that college is not a good place for all young men or women.

We seem to be in the same mind-set these day, the unquestioning assumption that society, and that individual, will be better if everyone goes to college.

@peter: how do you assess quality of college? I would argue that what we call quality might rather be "brand strength". Brand strength has elements of quality of education, but also the scarcity, type of people attending, the old boys network, the selection beforehand (hey if you take on only "winners"you will produce winners, even if you don't educate them).

Some research proves that at high school and college level, class size is not related to outcome. So college prices could go down fast by just taking more pupils. This however would kill the brand, while maintaining quality. WE pay (too much) for the brand. Why buy a Tesla when a Leaf would do for our daily commute in town?

Perhaps, just like with high schools, some measurable quality standard for undergrad studies should be created, and the value of the brand would be more aligned with the quality.

If we were really serious about breaking the signaling aspect of the college degree, we would allow prospective employers to freely test applicants on knowledge and skills. Until we do that, we aren't really serious.

A couple of reasons not mentioned above: Sports Teams, universities in Europe do not have them, sports are done at the club level. Salaries and benefits, I suspect that like government workers, they have it better much than private workers. The campuses, why do universities have these large campuses with grass field and sports facilities to the max. Some of these things started because in the beginning universities were Tory like finishing schools for the upper classes.

In VA 2011, the annual costs for public higher ed, including room and board, runs roughly between $15k and $21k. Over 4 years, that total is under $100k.

Many college kids can work during breaks and summers and earn between $3k and $10k to contribute to their costs. Over 4 years that is $12k to $40k. Taking the total cost at $100k, that means that a college student must find an additional $88k to $60k for college.

But, in VA, another option is for some students to live at home and attend a 4-year college or attend a community college for 2 years and then transfer.

Public-college tuition only in VA runs from approximately $3k annually (community colleges) to approximately $12k (UVA, W&M).

A kid needs to be housed and fed whether they live at home or at school.

Disregarding room and board, that means that the cost of college for 4 years at a VA public college runs between $30k (2 years at a community college followed by 2 years at a 4-year college) to $50k for 4 years at a public college.

And if the kid puts in $5k per year from work, that means that the after-work cost for 4 years is between $10k to $30k.

Some kids will get grants and scholarships, reducing the after-work cost even more.

And VA has a program that guarantees transfer from a community college to a 4-year college:


Last year, between freshman and sophomore years, my son worked 2 jobs during the summer (65 hours weekly) and at breaks and made over $11k, which he contributed to his college expenses. His goal is to graduate college with no debt, neither student loans nor credit cards, and then travel for a year or two, working his way around the world.

The costs of college do not include the missed wages during the 4 years, which in most cases are likely equal to the annual cost of college including room and board.

I don't believe the problem is the "consumer choosing to overpay". Fact is if you want a decent job you likely do need a 4 year degree and if you want a really good job you likely need an advanced degree from a highly ranked school.

The problem is cultural. Employers believe, perhaps quite rightly, that those who have managed to jump through those hoops of fire are more likely to make better workers. And even if the employers are wrong about that, they probably aren't wrong by much and after all: all those college grads do exist! Why not draw from that pool instead of the dropouts?

OTOH, I can picture this bubble bursting. In the 90's tech boom there were many people hired for high paying jobs who didn't have college degrees: self-taught programmers, waiters from restaurants who looked like they'd make good sales people, etc. If we saw such a boom in demand again, combined with a proliferation of autodidacts thanks to the web, the university system could be undermined. Unfortunately, if TGS is right, we aren't likely to see such a boom soon.

I like the idea of solving this maze backward. What does the exit look like? I see a world where, instead of college degrees, people are given tests to show their technical skills in whatever field. It wouldn't matter whether they were educated formally at MIT or community college or self-taught. The test score would be the relevant metric. Getting to that point will probably take a slow buildup of autodidacts, and we likely won't be there for a while. But someday. The self-destruct sequence is already built into the edifice.

"There are plenty wanna-bee professors discarded on the compost heap of academic history." Yet the best discard should not be much worse, and may even be better, than the marginally accepted professor. Such a large pool of surplus labor would play a significant role in an economic analysis of virtually any other sector.

Such a large pool of surplus labor would tend to depress costs because wannabe professors would be willing to take ever lower pay as the professorial surplus continues to worsen. Given the greater and greater use of adjuncts, and the phenomenon of "freeway fliers" who may be getting paid a full professorial salary (or even slightly more) but not receiving benefits, it appears that the cost of professors has declined.

So a sector which has ever increasing prices and costs has one of more visible sources of cost getting cheaper. Is it any wonder that economic analysis of the sector looks elsewhere?

I don't believe the problem is the "consumer choosing to overpay". Fact is if you want a decent job you likely do need a 4 year degree and if you want a really good job you likely need an advanced degree from a highly ranked school.

Right at the beginning of a career, maybe this is true, but how you perform that job quickly dominates the rest of your career, and I simply don't accept that you have to have the Harvard degree to get your foot in a door somewhere, or that the costs some are accepting is actually paying off in the long run. However, you can get a four year degree for a lot less than many of the college-degreed are paying.

Also, I suspect parking fines on campus lots would decrease. I still refuse to donate money to my university on the grounds that the blood they took out of me in parking fines is all they are getting.

If a high school student is involved in club sports and music lessons, the costs for those can range from $0 to over $5k annually.

Club sports during high school cost between $250 to over $3k annually, and music lessons can run from $125/month ($1500/yr) to $250/month ($3k/yr).

Those costs generally stop or drop significantly in college.

In the DC area, private (Catholic) high school tuition runs from $5k to close to $20k.

Sure, you can get an $8K degree at a land grant university. But that's not going to get you a place on Wall Street or a chance to go clubbing with a scion of Congress.

Most kids going to state schools aren't interested in working on Wall Street.

As for working on Capitol Hill, you have no clue what you are talking about. Lots of congress critters and people who work on the Hill graduated from state schools. A few members of congress don't have college degrees.

Peter, dirk and Yancey are all correct.

I think in a truly competitive marketplace reed's university would have a chance of making a go of it too, contrary to Millian's complaint, there are those people the Yancey mentions that overpay for the education, but reed's school might be able to offer a better signal and certainly would be well positioned for the consumption aspect of the college years.

What nobody has mentioned so far today is the accreditation bottleneck. This forces a homogeneity on college that is frustrating to me as a college employee and the parent of soon to be college age students.

Once the credential problem (the testing mentioned above, oh and the relevant SC case is Griggs v Duke Power) and the accreditation bottleneck are "fixed" college will get inexpensive real fast.

Home schooling anyone?

OK I read the Griggs V Duke Power briefing. Holy shit. Well, THERE'S YOUR PROBLEM. If we want education to be cheaper the first step would overturning that. Otherwise we're just pissing in the wind since we can't get the final outcome to where it should be.

Someone should sue Microsoft or Google or whoever is doing testing now and see if those guys can get the ruling overturned. I suspect they could.

As a graduate student at GWU, I paid $3,300 to take a required Econ course, along with 30 other classmates.

There's the problem. You are taking classes at GWU, one of the most expensive colleges in the US:

Did you not know when you applied, much less when you registered for courses, how much it was going to cost? If you did know, then maybe the solution is to go someplace less expensive. Like maybe a state school in whatever state you are a resident. You are probably paying at least double to go to GWU what it would cost you to go to a state school, like GMU.

Couldn't I have learned that information for essentially 0 dollars?

Sure. Why are you asking others? (I assume you are not being forced to attend GWU.)

I wonder how Austrian capital structure theory applies to this problem of human capital? I suspect that it would lead to overqualified people in jobs at the middle end, and lots of unemployment at the low end.

"OK I read the Griggs V Duke Power briefing. Holy shit. Well, THERE'S YOUR PROBLEM"

Agreed, most of education nowadays is just a very expensive substitute for a general IQ test as IQ testing is illegal.

"If testing became the norm, the tests would most likely be administered by a 3rd party. Are accounting firms prevented from judging a CPA by what he scored on the CPA test? If so, then I don't see why any other employer could get sued for looking at test scores. If not, then holy shit we do have a problem. Recent graduates often put their GPA on their resumes -- how is that any different?"

Generally white collar jobs that require licensing can discriminate based on scores but this leaves them open to lawsuits still, so they don't.

One can actually learn a great deal about why education is expensive by reading that article.

"If we could find a way of providing services of the same quality using less labor, price increases would be smaller."

It's called online classes I think. How many well known universities have made their degrees available online?

"colleges and universities are “first adopters” of new technologies because our faculty needs these tools to be productive scholars and teachers. "

How does leading edge technology help learning? It doesn't really. The same thing that made one a competent teacher in 1950 make one a good teacher today. But ego reasons dictate that having advanced - and expensive - technology is a requisite.

"In a sense, universities must meet an evolving standard of care in education that is set externally. "

Not true as whole. Most of the curricula and work doesn't change from year to year or decade to decade.

"Big state universities tend not to discount price on the basis of financial need nearly as much as the elite private schools, whose high and rapidly rising list prices garner so much of the headlines"

Big state universities also cost substantially less than elite private schools, so this actually makes sense.

"State universities are beginning to emulate...allow sufficiently high increases in the list-price tuition that wealthier families would have to pay."

This is funny yet incredibly sad. Apparantly the main problem is not "this costs too much" it's, "this costs too much for poor people". Unfortunately the "raise the price for the wealthy and discount it for the poor model" actually ends up being a terrible deal for the middle class, who can just barely afford it but don't recieve much or any financial aid. College is generally great if you're extremely wealthy and can afford to buy your way in to the system or if you're extremely poor and can get substantial financial aid. Otherwise, your going to struggle.

"To obtain federal financial aid one has to fill out confusing forms. "

Hardly. If you can't figure how to fill in the forms, you probably should not be in college anyway. FAFSA isn't the simplest process in the world but it doesn't represent a major barrier for a sufficiently bright person.

"a lot of learning comes from interactions with peers inside and outside the classroom."

Translation: at large state U, you'll probably never learn anything in class anyway, the main benefit is getting socialized. This is not by the way "learning" that the brochures suggest takes place all across the quad, this is "learning how to do a 30 second keg stand".

"But let me tell you the dark little secret about these grants. They do not create access to higher education for students who otherwise could not go"

I don't see how that's secret. Merit grants are about...merit not need. I know it's a shocking concept.

The article basically sums a number of convential trivialities and presents them as fait accomplia. Maybe at the end of the road colleges really don't care. Lowering costs would after all involve them getting less revenue and probably fewer benefits. Most people do not agressively advocate for that after all...

What happens when the value of a degree to employers is mainly about signaling that the employee has a certain IQ or base level of knowledge, but grade inflation and societal pressure for every child to attend college removes the value of the signal?

I would also blame credentialing pushed by certification boards and industry groups seeking to keep wages high by limiting entry of new workers. This inflates the demand for college degrees and makes demand inelastic for growing industries.

Computing is one area that has largely avoided the need for certification and licensing. Some of the most well-known programmers around do not have college degrees - Steve Jobs and Bill Gates, for example. The chronic shortage of programmers during the tech boom helped prevent the formation of licensure bodies, and even today getting a job as a programmer has less to do with your education than with your job experience and how well you do in a technical interview.

Computing is also one of the few fields where a 2-year diploma from a good tech school still carries a fair bit of weight, and where Ivy League degrees don't really impress employers. They'll hire someone from a state college who aces a technical interview and shows a personal history of real interest in computing over a graduate from Harvard who exhibits fundamental gaps of knowledge. There was a time when the for-profit DeVry Institute was seen as one of the top schools for producing programmers and engineering technologists.

I also agree that the availability of easy credit coupled with increased societal and family pressure to attend college is helping to disconnect the cost of education from the demand for it.

I have a vague feeling the rising price of college reflects rising income inequality.

The elite universities convey advantages worth far more than their cost. This may have something to do with the quality of service they provide but it has much more to do with the role they play in legitimizing uncompetitive networking advantages.

To be explicit, there are far more students who are qualified for admission to Harvard, than Harvard admits. The Harvard admission criteria do not select the most talented subset of applicants, in part because it's impossible to determine talent to sufficient precision at that stage, and in part because people actively seek to circumvent honest competition to grant themselves or their children access. (So, for instance, one pays for SAT prep for one's child, knowing this conveys an unfair advantage against poorer students.) Likewise, there are far more people who are capable of performing high-paying jobs, than there are such jobs. It would lower the pay of such jobs if not an artificial (uncompetitive) barrier were placed at entry, and the Harvard degree (for instance) serves such a role.

These effects are more significant, the more the marginal advantage of being ranked slightly higher at the top. Income inequality is the metric that signifies large marginal advantage for being ranked slightly higher at the top. So the small or non-existant advantage between a pre-Harvard and pre-Brandeis student is greatly amplified, as the difference in prestige of their degrees translate to difference in career worth lots of money.

Thus the system serves income inequality, and it benefits from it, insofar as rich people are willing to pay outrageous sums to convey to their children this advantage.

That's with the elite universities. The lower-tier schools are drawn to charge similar fees in part to make they look more like elite universities -- since most of the value of the private university degree is in the status conveyed, it's important not to give any indication of lower status -- and perhaps in part because they must, to be able to provide competitive services, given the greater charitable support (and government support) of the elite universities.

I've interviewed hundreds of software development candidates, not for Google or Microsoft but for competitive development houses in that class or the one below. I received training in the sort of questions that would determine a candidate's basic intelligence and technological aptitude without having them take any sort of objective test. I also received even more explicit training in the importance of never, ever, putting any negative opinion or objective fact about a candidate in writing. Griggs vs. U.S. Power was mentioned prominently. I have absolutely doubt that Google and Microsoft are aware that they might be sued for their recruiting practices, and have made provisions for it. It's a cost of doing business, and one that they happily pay to avoid the _vastly_ higher cost of hiring substandard software development talent (the only thing more expensive that a great developer is a bad one). The rule of thumb in the industry is that you only need to look for two things in a software developer: "Smart and Gets Things Done" (a phrase well worth googling for. The first of those is more important (although you do need to weed out the airy-fairy sorts who can talk theory all day long and never ship code).

I got an idea: If you want to reduce the cost of college, eliminate all student loans and watch enrollment drop. With significantly reduced enrollment, schools will have to lower costs in order to attract students.

But.....the reduced costs of production would be at the expense of......(a) school, (b) the faculty, (c) the administration, or (d) all of the above.

Sorry, that won't fly.

The worst quarterback in the NFL gets paid a nice but astronomical salary for a year or three, doesn't see much game time, and then doesn't make the roster. He then makes a living doing what the best quarterback who didn't make the NFL has been doing the last few years. There is no sudden discontinuity in their lifetime earnings, and the guy who didn't make the league is healthier is his old age on account of a few less years of regularly having the living daylight beaten out of him.

The worst person in a highly compensated role doesn't stay there long.

@Cyrus -- good point. But what if you take the worst person who is just good enough to make a career of it vs. those not quite good enough to make a career of it? At some point there is a dividing line between those who succeed and those who fail, yet those just on both sides of that dividing line probably don't differ much in ability. Take any field and there are, almost by definition but also due to bell-curvature, many more mediocre performers in that field than superstars, and those mediocre performers aren't to a material degree better than many others who could do just as mediocre a job yet don't have the job.

What about, "Because they can."?

Of course, you are comparing apples to oranges in the context of this thread, and completely ignoring how difficult is to even take the exams you mentioned without getting the degree first, right? How many states let you take and pass the bar exams, and actually practice law without completing law school? Is it seven or eight now? How many of those don't have nearly equally time consuming requirements of law office study? One or two?

Most companies, and not guilds, forgo such exams because the burden is on the prospective employer to show that the exams are job performance related, and that is why they use degrees as a proxy for screening candidates and actual hires. Believe me, any one who has taken part in the hiring process in corporate America is trained about what is and is not acceptable in testing an applicant's knowledge and intelligence (and how to do it). I and some of my former colleagues were trained by two lawyers whose practice involved suing employers who violated Title VII in just those ways. An urban myth, it is not.

As one of the unwashed heathens who has no formal 'higher' education, I'm proud to say I've done just fine so far. I taught myself what I needed to know when I needed to know it. And without a mass of debt, I've never stressed about how much I was earning.

True, I don't have that awesome network of well-connected pals I might have collected had I attended college, the 'net is making most of that irrelevant, anyway.

For a weeding out phase, why don't Google and Microsoft do what grad schools figured out long ago? Hire people, don't pay them very much, and don't let them touch anything important, only let them teach.

In engineering they could lop off a year (and maybe even two) from the curriculum. At my alma mater, they actually did that in effect as the time to degree was creeping up towards 5 years they reduced the hours required. However, you could easily test out of many Freshman classes and Senior classes are either highly specialized or resemble graduate prep courses. Both of these are almost never used in industry. Much of the senior year could be optional.

Frankly, the ever-increasing financial aid budget and the Duke Power case are both major causes of tuition inflation, but I don't realistically see either of these being addressed. Neither is politically tenable.

At the relevant margin, demand is highly inelastic. This is, by the way, very similar to healthcare. Some people are priced out of the market entirely. For those that are still buyers, they will pay come hell or high water. They may bitch and moan about the cost, but when push comes to shove, they will pay whatever it takes for {treatment}{Ivy League}.

I'm recognizing those alleged Microsoft pre-employment questions as Car Talk puzzlers.

You may think A is copying B, but it may be B copying A, or both of them copying someone else.

I really would like to know what the latest is on someone suing Google or Microsoft. I've also worked at software companies that tested extensively, and we really didn't worry at all about being sued because we didn't hire someone who didn't know how to use gcc.

I remember one comment from the CTO after getting a specific applicant's answers: "forget this bozo." He cc'd the entire tech team to make sure we all knew the applicant sucked.

Wow, all these comments and just two mentions of health care! Virtually every university administrator you talk to will tell you that exploding health care costs for employees are behind much of the rapid increase in college costs. Colleges and universities employ a lot of people, you know. Now, for why universities appear to be able to raise their rates as a result of rising health care costs while most private businesses cannot (at least to the same extent), that's a subject for debate...

Comments for this post are closed