There's a thoughtful discussion over at Dr. Crazy's about full-time faculty workload. (The post was a response to Tenured Radical's own discussion here.) Within a recognition of the importance of context, Dr. C notes that what looks on paper like a static workload has actually been increasing in insidious ways over the years. (It's all the extra, off-the-books stuff, like advising and assessment, that consumes the extra hours.) In her estimation, some faculty have made themselves martyrs, and others have "half-assed it" in teaching, since there's really no institutional reason not to. Recognizing the limits of the strategy, she has allied herself with some colleagues to look for ways to streamline the extra tasks to allow for a more sustainable workload that doesn't shortchange students.
She notes in passing the different understandings of 'productivity' that underlie the work speedup. More students per class increases the tuition generated per professor, at least in the short run; that's one version of 'productivity.' More students per class decreases the amount of individual attention the professor can really give, which leads to a decline in the quality of feedback; that's another version of 'productivity.' Both versions are internally valid, but they don't necessarily mesh with each other. And that's where the real problem is.
From the standpoint of an individual instructor, the controllable variable (at least to some degree) is the quality of instruction. That's also what you care the most about, what you pride yourself on, and at a really basic level, why you're there.
From the standpoint of trying to make payroll, though, the opposite is true. A thrilled student doesn't pay any more than does a barely-contented student. (There's presumably a minimal level at which attrition becomes an issue, but I'm assuming at least basic competence.) Students pay by the credit, the course, or the year; they don't pay by the breakthrough. The 'extras' that a great class can generate don't show up in the budget. Worse, some students actually prefer classes that don't ask very much of them. (If you doubt the truth of this, spend a day at in-person registration, just listening.) The mutual non-aggression pact between an instructor who doesn't ask very much and students who'd rather not be bothered is one of the open secrets of American higher ed, and it fits short-term institutional needs disturbingly well. There's a reason that Rocks for Jocks and Physics for Poets still exist.
(I'll add here that I agree with Dr. C that in some classes, there's really a minimum size beneath which quality can actually drop. A public speaking class with two students isn't really a public speaking class. I once had a section of six very quiet personalities, and teaching that was painful. A couple of sparkplugs would have enhanced the class tremendously. But this is really a side issue.)
The endemic conflict is that beyond a minimal level, and outside of the elites, there's no economic incentive for the institution to do better than okay in the classroom. Once you understand that, the rest follows. (There's a moral imperative, but that's a different issue.) For a college that's struggling to stay afloat financially, the short-term cost of stuffing a few extra seats into each class is dwarfed by the tangible and immediate tuition gain (or labor saving). You can blame pinheaded administrators for that if you want to, but you'd be missing the point; the math is what the math is. When the college is flush, it's possible to make a choice to have your cake and eat it too; when the college is strapped, though, the contradiction is unavoidable.
And that's the core issue. If you want to be paid for quality instead of quantity, you have to charge by quality rather than quantity. You have to align the incentives.
The elite SLAC's do a version of this by selling exclusivity. If you charge 50k a year for a small residential college, small classes are part of what you sell. There's a market for that, and you'd defeat your own niche if you watered that down too far. But most colleges and universities don't do that.
And it's not entirely clear how to do that. In olden times, I'm told, lecturers went out on public circuits, and the audiences paid them according to how impressed they were. It was lecturing for tips. But I don't see that (or anything terribly close to it) making sense now, if it ever did.
One could argue that philanthropy is a very delayed response to 'extra' quality --- quality above and beyond what the student paid for --- but the length and uncertainty of the delay makes it a difficult sell. I agree that community colleges could and should do a better job courting philanthropic resources, but I'm not convinced that this will tip the balance in most cases, particularly in the short term. One could also argue that 'prestige' is a proxy for quality, but anyone who has t.a.'ed intro classes at prestigious places (or who has taken those intro classes with t.a.'s) can tell you that the connection between prestige and quality is problematic at best. A good adjunct will often do a better job than a full-time professor who's "half-assing" it, and will do it for much less. As long as price isn't connected to quality, these perverse incentives will arise.
Since I haven't cracked this particular nut, I'll crowdsource it. Wise and worldly readers, is there a sustainable way for colleges to charge by quality rather than quantity?