Here is the key metric: the discounted price of Swarthmore tuition, as a percentage of per capita nominal GDP (PCNGDP), has not changed much since at least 1954, when my father was at Swarthmore.

My grandfather paid $2000 in tuition, room, and board in 1954. PCNGDP was $2333 (according to eh.net). Swarthmore comprehensive fees were $26,565 and PCNGDP was $39,959 in 2004.

O.K., fees are not exactly constant: it has gone from 86% down to 66% of PCNGDP. 🙂

To be fair, the 1954 fees are not “discounted”. I probably should knock only about 10% off for financial aid, since “scholarship” cases were far less common then. For example, my grandfather was a sociology professor at a state university, yet he was expected to pay full tuition (not the case today).

Considering the undiscounted price, what people making >3X the median family income pay, the price has risen only a little: from 86% to 100%.

This rise in undiscounted tuition represents an attempt by the college to counter-act the other major change from the 1950’s to now: the incredible rise in the income of the top 20% of households, at the expense of the median and below.

For example, while real per capita GDP has doubled since 1970, the real cost per hour for unskilled labor has only risen less than 5%. This country is incredibly more productive than it was in 1970, but the lower half of the income curve is not seeing any additional income.

By the way, if you don’t believe my analysis of income disparity growth, see the CIA Fact Book’s description of the US economy. That communist organization makes the same point. 🙂

While it will hurt me in the wallet when I send my child to Swarthmore in a couple of years, it will hurt less than it did my grandfather 50 years ago, particularly since our household is firmly in the top 20%.

For the top 20% of income earners, Swarthmore is about as cheap as it ever was, by the only metric that matters: fraction of household income. For the rest, the practice of needs blind admissions, and huge tuition discounts based on need, means Swarthmore is cheaper now that it was in the 1950s.

We are living in the golden age of cheap, excellent, private education.

By the way, healthcare for students is a red herring. There is no demographic to which providing health services is cheaper than college age, non-reproducing, students. With the exception of a few ACL surgeries, this group may actually cost less now in real terms (because of increased productivity in the health care sector) than it did in the 1950s.

The solution you suggest (using a test and work history) is the logical one; it is, unfortunately, not entirely legal in the US. An employer can use a college degree as a criterion for hiring so long; however, any test must either be directly related to the job to be done or have no disparate impact by race.

In theory, this might also be extended to all kinds of public non-military service, in police work, social work, firefighting, and so on. Considering how strapped some localities and states are for good people within budgetary confines, it isn’t impossible that states and local governments might bypass college grads directly and go for high school grads. Once having demonstrated a track record of working for an organization, these non-college graduates then become excellent hiring potential, the more so if they do any IT-related work while putting in their (low-paid or volunteer) time, which is silver.

Developing a standardized test and getting it widely accepted would be no small thing, but I don’t see any reason to think it’s impossible As college education gets more expensive, the incentive for such a test would grow. Also, I keep hearing complaints about incompetent college graduates. I don’t know how sound they are, but if the complaints are reasonable, that’s another reason for wanting a test.

The work history is an alternate way of checking on whether a job candidate is willing to keep showing up and (one hopes) working.

Yes, two students to a room should lower the cost of housing. But, on the other hand, start pricing what is essentiall a gated residential community on the open market that offers the beauty of Swarthmore’s campus plus 24/7 walk-in health clinic, counseling services, big screen TVs, “free” high-speed internet in every unit and a full service cafeteria for three meals a day. I don’t think you’d find that for what Swarthmore charges.

However, this misses a key point. College pricing has little or nothing to do with the cost of services rendered. If it did, the prices would be double the current price at most higher-end schools (and all state universities).

Pricing is purely market driven. If the seats are full on the airplane, then the price is right. If the plane is flying with empty seats (or if the seats are filled with lower SAT scores than the school needs for its desired level of prestige), then the discount price will be lowered to attract more customers. Those discounts may be, and often are, specifically targeted at wealthy, high SAT scoring customers because $40,000 of discounts buys four or five of those highly desireable customers rather than one low income (probably low SAT) customer. The average SATs (i.e. prestige) of the school goes up, thus increasing the appeal to full-price customers down the road. The increasing use of merit aid discounting as a pricing strategy hasn’t happened by accident.

The spending levels are largely determined by other factors: per student endowment, per student state subsidy, etc. Technically, most schools lose money on every student. That’s why they can’t “make it up on volume” and why growing the enrollment to “increase revenues” is often a fool’s errand in higher education — unless you increase enrollment without adding dorm rooms or hiring a new professor to teach them (i.e. lower your quality). Swarthmore is a financial heavyweight (one of the top per student endowments in the country) in large part because it has not significantly grown its enrollment. It is the enormous per student endowment (roughly $750,000 per student) that allows it to spend $68,000 a year and only charge $26,500. If Swarthmore had doubled its enrollment, the per student endowment would be sliced in half, thus eroding the ability to spend so much above the market-driven price.

In fact, Williams encountered just this problem when they rapidly increased the enrollment in the ’70’s from 1400 to 2000 to accommodate the admission of the “new-fangled” female students. As a result, they fell significantly behind Amherst and Swarthmore financially because their endowment was diluted. Only by holding their enrollment to zero growth over the last ten years has Williams’ per student endowment caught back up to Amherst, just slightly behind Swarthmore.

Anyone in college housing must buy into the meal plan, purportedly because the dorms don’t have enough kitchens for everyone to cook. But that can’t be a real reason. So maybe college housing loses money, and we’re subsidizing that? I don’t know. Is this information available (and I just haven’t found it)?

HWC’s comment is extraordinarily useful. For one, it brings to the fore the crucial point which is frequently not at all well-understood, that many small liberal-arts colleges at the upper or luxury end of the market are spending far more per student, *even* the students who are not charged at a discounted rate, than they take in through tuition.

However, coming off HWC’s observations, one of the things that I do think consumers could reasonably ask of colleges and universities, parallel to some of the demands now being made upon hospitals, is more transparency in breaking down the composition of the cost on the overall bill. When hospitals have been required to do this, it’s been revealed that internal pricing metrics at any given hospital can be wildly different from a very similar hospital nearby: one can charge 11 cents for an aspirin, another $60.00 for the same. That suggests that internal pricing is an arbitrary practice designed to tally up a relatively consistent total bill which covers costs adequately. The problem with that is two-fold: it may conceal from consumers and the general public what the true costs of medical care in a hospital actually are (and thus prevent us having a useful national discussion about those costs) and it may obscure even from the hospital itself where real savings might be possible.

A college bill which broke down precisely into the particular payments for every major service might not only help consumers understand what exactly it is that they’re paying for, but might help any given college to understand where its costs don’t compare favorably to local market pricing of a particular service. My students observed yesterday, with some fairness, that while you couldn’t find rental housing in our local market at the price we charge in sufficient numbers for our undergraduate student body, some of the housing that we actually provide for that price is, well, pretty crappy: single rooms with two or even three students in them. You can’t buy that kind of housing in our local residential market because there’s nothing quite that squalid available within a mile of the campus. This is not to say that there aren’t arguments in favor of the residential facilities we make available to students, in terms of the positive forms of community building that come from being roomed with someone (or at least the practical skills involved in negotiating privacy and autonomy under those conditions), as well as practical and aesthetic considerations (building a dorm big enough to house students in their own apartments or rooms would require a building much, much taller than anything on campus now, or building out of the current tight clustering of the college’s footprint, plus we don’t have the money to build a huge new facility right now). Still, we owe it to our customers to price housing out in such a way that it allows them to reasonably observe the relation between our prices and comparable local housing markets.

More transparency would also expose discounting practices to clearer view. This might make for some uncomfortable moments, when it becomes clear that virtually all private institutions follow a pricing strategy that might fairly be called “soak the rich”, but I think it’s again only fair to make that fully visible to the paying customers.

Here are several issues to consider:

a) The real price is not the “sticker price” published by the college, but the net tuition, room, and board charged after discounts (finanical aid). For example, Swarthmore’s sticker price is $40,000+. However, in Fiscal 2004, the average Swarthmore student actually paid $26,565 after discounts.

b) College pricing follows the airline model with each customer paying a different price. Some pay full sticker price. Some pay nothing. Like the airlines, colleges adjust the proportion of expensive to discounted fares to insure that all the seats are sold.

c) The nature of the tuition discounting reflects institutional priorities and, in many cases, sophisticated enrollment management techniques. Some colleges use their available discounts to attract need-based aid students. Others limit those discounts in favor of offering merit-aid discounts to attract wealthy, high-stat customers. Merit aid discounting is a pure form of price competition.

d) When the amount of discounting required to fill the seats gets too burdensome, the college is faced with a choice: reduce the discounting (and accept the corresponding decline in average SATs) or reduce expenditures. Smith has recently reached the point where its discounting has pushed the budget out of equilibrium. They have responded by reducing the size of the faculty by 8% along with other cost-cutting measures, including increasing the size of the loan component in their need-based aid.

e) The real “value” is not determined solely by the net price charged, but must also include looking at the expenditures per student. For example, Swarthmore charged $26,585 per student in 2004, but spent $68,304 per student. That’s a heck of a bargain. Is it any wonder that customers are lined up around the block? Is it a surprise that the handful of schools offering this kind bargain, for a sustained period of time, are the most pretigious, high-demand schools in the country, clustered at the top of the USNEWS charts? To put this value into perspective: If Swarthmore reduced its spending to Oberlin’s level ($40,135 per student), Swarthmore could eliminate all student charges entirely, i.e. make it free for every student. In fact, there are several schools (Berea, Olin) that have adopted exactly this pricing strategy.

f) The real price-pressure fun will begin in about three years when the echo-boom demographic glut has passed through the system and demand starts to decline. I expect that some of the excess capacity in the system will fall by the wayside.

Once you start wrestling with these aspects of college pricing, you can start to identify a wide range of consumer options. For example, Grinnell can’t generate the same demand as Swarthmore because it’s in Iowa. So, they charge less (a mix of need-based and merit discounting) and compete successfully in the quality end of the market based on price. They spend less and charge less, thus offering a terrific value for a price-conscious consumer. Conversely, Swarthmore and Williams compete at the luxury end of the market, spending more and charging more (in higher education terms, “charging more” really means having enough prestige to attract more full-fare customers).

Here are the 2004 numbers (per student spending and net per student price after discounts) for several well-respected liberal arts colleges:

Swarthmore: spending $68,305 – price $26,585
Williams: spending $66,936 – price $25,734
Haverford: spending $54,223 – price $28,896
Grinnell: spending $46,615 – price $20,399
Davidson: spending $45,499 – price $25,727
Oberlin: spending $40,135 – price $22,239

I think this is a statement you have to defend. People tend to think of most of an institutions budget as fixed, but in reality the entire budget is reflective of a series of decisions that each institution chooses to make over and over when it prepares its annual budget.

For instance, Swarthmore could kill the Honors program, significantly change its leave policy, close any number of administrative offices, or alter its faculty:student ratio. We don’t, because we assume that we believe these things are roughly good and/or not worth changing. Faced with some sort of radical change in consumer behavior, we could.