Teaching vs Research
I am inclined to agree with Ron that universities should give more weight to teaching relative to research. Yet I am not fully convinced about my own judgments here because I do not understand what market failure leads universities to overweight research.
Approximately only the top 20% of published journal articles really make any difference at all. Most of the rest are read by scarcely anyone but the editors and referees that decide the papers' fate, and do little more than earn their authors tenure and merit pay.
On the other hand, writing one of these papers in the lower 80% can be a useful exercise. It keeps you connected to the literature, the questions top researchers are asking, the methods they're using, the results they're finding. It helps keep you as part of the discussion.
Still, I feel that the social benefit of published journal articles outside the top 20% is way smaller than the private benefit, and too small to justify their overwhelming importance in promotion and tenure decisions. I have colleagues (not necessarily in my department or college, but maybe) who publish a high quantity of articles in "C" journals, get $2500 merit pay raises every year, and are living large. But if you read one of their papers, you have to wonder what the contribution is, why anyone would bother reading it.
The ability to do research in this lower 80% serves as a barrier to entry and reduces competition among teachers and schools. I wonder if this might be why schools stack the incentives toward mediocre, unimportant research and against more socially meaningful things like exceptional teaching that truly makes a difference. If someone knows of a different reason, please say so. I would much rather hold a less cynical view about this.
Note that Harvard may now be moving in the direction of elevating the role of teaching. Only time will tell whether this change is real or just "cheap talk" aimed at satisfying alums and other donors.