A graduate is not a product.
Since students are not raw materials, it's silly to pretend that the university is shaping them in a controlled and predictable process to produce a well defined output. Many university managers do exactly that, though (I mean, to pretend that they can manage education as if it was the making of a product).
But the raw materials issue is nothing but one of the many perspectives under which graduates aren't products. The whole approach is definitely wrong.
A production process without any product would be completely pointless. The whole factory would be stopped with nothing else to do, except supporting itself. A university that doesn't produce any graduates at all would be (of course) rather crappy in some sense, but in spite of that it could still be fulfilling a crucial mission. People who study at the university and leave may obtain a benefit, even without any degree; actually, universities teach many courses totally unrelated to any degrees. Research could go on significantly; and even get a huge impulse, if we have to believe some so-called teachers that hate teaching. (I think it's pointless to clarify this, but I'll do it anyway just in case: I'm playing the devil's advocate. I'm not telling that teaching is not important, or that a university does not need to produce any graduates or even a few. I want lots of them; I love teaching.)
You can examine a product and clearly verify quality criteria. You can measure how good or bad your product is. You can reject flawed products, and make statistics. But you can't do that with graduates. First, the grading process itself is a verification, but it takes place before graduation. Once you've given the award, how can you judge whether your product (the student, in this case) meets quality criteria, when giving the award is a verification of how well a student meets the quality criteria? This may look like just a funny and irrelevant paradox, but it is a rather real and serious problem. This recursion renders many academic quality procedures useless, redundant and even just plain stupid. A university has already marked a student when he's given the award, and it's up to someone else to attach value to that award or not. To whom? Another interesting question; distrust yourself if you think there's an obvious answer.
Moreover, measuring the quality of a graduate is not easy. Assuming that it made sense to verify (a second time) how well a graduate performs, what parameters should we take into account? Who is the consumer for such a product? The obvious answer these neo-liberal days is "the companies", of course. I expect to write on this soon, but at this moment I'll just say that, if there are more than one possible customer, the quality indicators may be different for each one of them, and even contradictory. When you're dealing with products, you can forget about everything outside your target market. I don't think you can make this simplification with graduates.
The tradition is: for a company, producing more products, at a lower cost, with the highest benefit is the only business. Quality is defined in terms of consumer expectations; quality is good as far as it satisfies the target consumer, at least enough to sell well and give benefits. Even in this case, the definition of quality depends on the target market. A Rolex is a piece of shit for my expectations in any case, but my 10€ Casio (I'm happy with it) is less than shit for a Rolex buyer. Is the goal of a university to produce as many doctors or architects or engineers as possible, at the lower cost, with the highest benefit? Probably the answer is not so trivial. But some quality measures I've seen recently seem to consider the problem under such a trivial approach.
A graduate is not a product.