If you ask anyone what colleges and universities are for, they'll give you more or less the same answer: to educate people.  That's a good answer. It's the one I give myself. But it's only half the truth. Colleges and universities actually fulfill two separate roles. We all know about both of them. We only talk about one of them. And because of that, we misunderstand almost everything about how higher education works and how it might be improved.

Every individual college and university exists to educate: to teach people things they did not previously know. (People disagree over what the goal of education should be, but agree that education is the goal.) But colleges also confer social prestige on their graduates. Some confer a large amount, some a smaller amount, and some confer little or none. Taken together as a system, the American colleges and universities have both a teaching function and a sorting function.

I dislike the sorting function and would rather it not exist. But saying that a deeply embedded social practice ought not to exist doesn't get rid of it. Too many people are committed to it, and too many people believe in that social practice as a simple reflection of reality. American higher education labels students by the "quality" (meaning the selectivity and prestige) of the school they attended. People by and large take those labels as real indicators of students' intelligence, likely prospects, and so forth. This social labeling is part of America's class system.

We don't talk about this much, because most people don't like talking about the class system. It's difficult and embarrassing. (And before anyone finds this discussion of class upsetting or insulting, let me say two things. First, I'm not saying that some people should have more social prestige than others. I'm pointing out that in practice they do. I'm trying to describe how the system works, not endorsing it. Second, while talking to another person about their class position can be insulting and a way of putting them down, not talking about how the larger system works is a way of keeping the system in place and letting the people on top get away with things.) But even when we don't talk about class, and maybe especially when we don't talk about it, we experience it a real force in our society. We all know that Princeton alumni enjoy advantages that alumni of poor public colleges do not get. We feel it.

Our public debates about higher education are confused because we don't talk openly or think clearly about the two different functions. It's easier to pretend that colleges are doing only a single thing. But that leads us into misunderstandings because we don't acknowledge how our things really operate or, worse still, talk about one function as if it were the other.

Some people talk about the "great education" a school provides when they really mean the social cachet it provides. Some perceive socially-disadvantaged schools as genuinely providing less learning, no matter the quality of actual education there; those schools are simply "bad schools." These people conflate education and social prestige without being aware that they're doing it, and have trouble perceiving educational quality separately from educational privilege. Other people sometimes deny that anything happens at universities but social differentiation, and are prone to claiming that no actual education happens at elite schools, and so on. This is also a radical misunderstanding, and saying that nobody at Yale learns anything that people don't learn in night school is such an overstatement that it undermines the speaker's point. If two schools have different budgets and different missions, there will be real differences in educational outcomes. No one who's worked in higher education can honestly deny that.

There are schools that are good at teaching students things. There are schools that are good at giving their students social credibility. There are schools that do both well, and schools that do neither. And some can only do one. There are schools that do an excellent job of educating their students but cannot give them any social prestige. There are also, I'm afraid, schools that provide mediocre or poor educations but do better at conferring social capital.

Why does this matter? Because those two basic functions,educating and conferring prestige, have distinct and opposed economic logics.

Education is an absolute and unlimited commodity. The more you produce, the better. Teaching people things is perhaps the ultimate non-zero-sum game.

Social prestige is a relative and inherently limited commodity. The value of selectivity and exclusivity lies in the fact that most people are excluded. In a perfect world, every student could be superbly educated, but in no possible world could every student be more prestigious than every other student. "Exclusivity for all" is a nonsense slogan. Conferring prestige on students is very much a zero-sum game, where one student can only gain from what some other student, somewhere, loses.

The prestige game matters because a school's prestige is its lifeline to more tangible resources, most importantly funding and potential students. A selective institution must maintain its perceived selectivity, its reputation for being hard to get into, for fear of losing access to the most desirable students. The difficulty of getting into a school is key to the social value of getting in to it, which is why exclusive universities trumpet their obscenely low admissions rate every year. Selectivity turns out to be self-fulfilling; if you lose your reputation for selectivity, the students you want most will stop applying.

If your college or university dedicates itself strictly to education without any thought of conferring social prestige, both the school and its students will suffer for that choice. The students and the institution will be stigmatized by many people, and their actual educational achievements will often go unrecognized. Graduates will face disadvantages on the job market; the school will be starved of resources. This is the story of public higher education in America over the past thirty years.

If your university protects its students' best interests, and its own, by working to build its own prestige (and thereby the prestige it confers on its alumni), it must enter a ceaseless zero-sum competition with its peers and rivals. This competition may strike some observers as dysfunctional, but it is an entirely rational and inevitable response to the underlying system of rewards. Since prestige is always relative, any institution must constantly be trying to keep up with its peers and stay ahead of the schools behind it. Every administration tries to move up in the pecking order, which is probably the only viable strategy for not falling further down in that order. You have to run just to stand still.

Understanding the prestige competition between universities is crucial to understanding all the other intractable, poorly-explained questions in American higher education: how admissions work, why costs keep rising, why there seems to be such emphasis on research, what the current rage of enthusiasm for MOOCs is all about. Unless you separate out the questions of education and prestige, it's hard not to misunderstand these questions.

Many people, for example, talk about the "high-quality courses" from Harvard and Stanford (which have taken the lead in MOOC production) being turned into MOOCs. But that is not an evaluation of educational quality; it is an evaluation of institutional prestige. If you don't make any distinction between teaching and creating prestige, then it seems self-evident that a class from Harvard is better than a class from Underfunded State. If you think of teaching and conferring prestige as separate things, then the MOOCs look like a pretty bad deal.

One of the MOOCs that gets the most press is a version of one of Harvard's most famous gut classes, a class universally known among undergraduates by a derogatory nickname. (I can't remember ever hearing a Harvard student call that course by its actual name. For practical purposes "Heroes for Zeroes" is its name.) I won't call it a bad course. (I've never taken it.) But there are plenty of classics professors at much less glorious places capable of teaching equally useful courses on the same material and making those courses more challenging for the students. Questions of prestige aside, taking that particular course at Harvard is not a better deal than taking an equivalent course at any number of less-glorious schools.

But the MOOC version of that notoriously-easy class is actually much easier than the Harvard class itself. At least students in the actual class write a few college papers, which get read and graded by teaching assistants who are studying for their own classics PhDs in the field (i.e. by smart people who can read classical Greek in the original). That's not possible for MOOCs, especially because of the Massive Open part. So someone taking the MOOC version just takes multiple-choice reading quizzes instead. And instead of a weekly face-to-face discussion session with one of those doctoral students (who actually knows what's going on in the material), there are lightly moderated online discussion boards. Now the educational product is very clearly inferior to taking a real class on the subject almost anywhere. Even a face-to-face version that's not quite as good as the face-to-face Harvard class is still much better than taking the Harvard MOOC. Talking about the MOOC as superior to a real course at Inglorious State is simply delusional. And replacing face-to-face classes at poor schools with MOOCs from rich, famous schools would be a rotten deal for students at the poor schools.

Does this mean that Harvard doesn't do as good a job educating its students as less prestigious schools do? No. It spends more resources on education than poorer schools can dream of spending, and that matters. But it is not going to spend the kind of money it spends on its own handpicked students on every random stranger who signs up for a MOOC. It could not and would not. Harvard will always save its high-cost, high-value educational products for its own students.

Cheerleaders for MOOCs talk about how they will make education more democratic, breaking down the exclusivity of the elite schools and making elite educations available for all. That has nothing to do with reality. The two most prominent MOOC providers, Harvard and Stanford, are currently wrestling for boasting rights over whose admissions rate is lower. (Stanford's rate finally fell below Harvard's by a tenth of one percent. They want everyone to know. Harvard wants everyone to forget.) Both universities are intent on turning away more applicants every year, and publicly boast about how many excellent students they have turned away. These are not schools committed to breaking down exclusivity. These are schools committed to being the most exclusive. Exclusivity is their business.

If you take a MOOC produced at Harvard or Stanford, you don't get the full educational value that the real Harvard or Stanford version of the class provides. But you get absolutely none of the prestige that Harvard and Stanford gives its students. Part of the cachet of going to those schools is getting into a school that turns down more than 94% of its applicants. MOOCs take everyone who logs on. Harvard and Stanford have enormous social value because they are clubs almost no one can get into. MOOCs are clubs that will take anyone as a member.

In fact, the point of a Harvard or Stanford MOOC is to remind you that you are NOT at Harvard and Stanford, that you are NOT one of the chosen few who gets to take the real class. They get to go to the selective school. You get to wish that you were one of them, with your nose pressed up against the monitor glass. The point is not for Harvard and Stanford to reduce the educational difference between the haves and the have-nots. The point is to increase the prestige difference between the haves and the have-nots. It isn't democratic. It isn't even very nice.