Why charge less when increasing the price increases demand?

That’s a question that university administrators have been asking themselves for many decades now. A couple of days ago, when asked to opine on the issue, Elena Kagan delicately suggested that, in conventional economic terms, it doesn’t make sense for Denny’s to charge almost as much for its Grand Slam Breakfast as Le Bernardin charges for its Thinly Shaved Geoduck.

[Northwestern Law School’s dean Dan] Rodriguez asked Kagan about the troubled state of legal education. Kagan demurred, stating that she was no longer in a position to advise law schools. But she did say she thinks “there ought to be more than one mode of legal education.” Since graduates of different law schools have different job opportunities when they graduate, Kagan said, not all schools should have the same teaching model—or tuition level—as “the elite law school model.”

No it certainly doesn’t seem to make sense for schools that send 5% of their graduates to high-paying legal jobs and four times as many to the unemployment line to charge 80% as much as Harvard, but such schools have been consistently charging 80% as much as Harvard since the memory of man runneth not to the contrary, aka the Eisenhower administration, if not earlier.

So why did they do so? The short answer is because they could. They could for a variety of reasons, but the one I’m going to focus on here is because the market for law degrees in this country seems to have featured an inverted demand curve — that is, one in which, all other things being equal, raising prices actually increased demand relative to competitors.

Consider a couple of natural experiments.

(1) The New York City area offers applicants about a dozen private law schools, and one public institution: CUNY. If you compare CUNY’s historical applicant pool to that of, say, New York Law School (not NYU), you’ll discover that until a couple of years ago, when after a blizzard of bad publicity NYLS’s applications started to collapse, NYLS was consistently getting about three times as many applicants as CUNY, even though it charged four times as much in tuition for New York residents relative to CUNY, and almost three times as much for non-residents. NYLS even had slightly higher entrance requirements, even though neither school carried any prestige on the hiring market (if anything, CUNY might have been the more desirable school from a prospective employment perspective, since it went and continues to go to great lengths to position itself as a law school dedicated to public interest practice).

Why, given the similar value of degrees of degrees from the schools on the employment market, would 6,000 people apply to NYLS in 2011 for the chance to pay $48,000 per year in tuition, when only 1,900 people applied to CUNY, where in-state tuition was $12,000, and non-resident tuition was $19,000? Why would three times as many people attempt to pay four times as much for what, in conventional economic terms, was pretty much the same thing?

(2) An even more stunning example of a radically inverted demand curve is provided by the tuition history of the University of Colorado Law School, where I teach.

Throughout the 1990s, CU raised its resident tuition (everybody gets resident tuition after their first year if they’re not already a resident, so for all practical purposes resident tuition represents the price of attending the school) by an average of about 8.5% per year, which incredibly enough was a markedly lower rate of increase than the typical public law school during these years. (Nation-wide, average public law school resident tuition nearly tripled between 1990 and 2002, but it only doubled at CU, from $3,130 to $6,352).

In 2003 the new dean convinced the central administration — I don’t imagine he had to break any arms or anything — to essentially quasi-privatize the school’s tuition over the next few years. This led to, even by the heady standards of American law schools, a truly mind-boggling series of price hikes, with the result that by 2011 tuition was fives times higher than it had been nine years earlier (four times higher in constant, inflation-adjusted terms).

What did this do to demand? Behold the wonders of the beneficent Market, home to sophisticated consumers, rationally maximizing their individual utility, world without end amen:

2.88% of all law school applicants applied to CU in the 2003-04 cycle.

4.13% of all law school applicants applied to CU in the 2011-12 cycle.

By increasing prices five-fold, CU increased demand by 43% relative to law schools nationally, even though over this same time average private law school tuition “only” went up by 61%, while average public resident tuition roughly doubled.

Did CU perhaps lower admissions standards, in order to make the school more attractive to a larger pool of potential applicants? Far from it stout yeoman: while the 2004 class had a median LSAT in the 84th percentile, the 2012 class’s median was in the 90th, and featured higher GPAs as well.

These experimental results are made even more vivid and robust by comparing CU’s outcomes to those at the University of Denver, just 30 miles down the road. While CU was quintupling its tuition, DU’s didn’t even double — meaning that in practical terms DU slashed its tuition relative to its neighbor’s — yet DU’s applicant pool was almost the reverse image of CU’s over these years, going from 4.18% of the national total in 2003 to 3.09% in 2011.

Note that if CU were a private corporation, its management would be violating its fiduciary obligation to the firm’s shareholders it they hadn’t raised tuition from $6,000 to $31,000, just as Honda’s management would be violating its legal duties if they hadn’t raised the price of an Accord from $20,000 to $100,000, if doing so would have led to selling 43% more cars relative to Toyota.

Now it seems, shall we say, unlikely that raising the price of an Accord to $100,000 would cause demand for the car to spike, but higher education in this country is clearly a very special sort of market.