Mankiw on textbook costs

I have already posted twice about the excessive cost of textbooks, but today's post by Greg Mankiw needs a reaction.

Greg Mankiw, probably the economist that makes the most money off textbooks, and I suspect by far, argues that there is nothing wrong with the cost of textbooks. He claims that this is a market with free entry, and if the price is too high, others would enter and drive prices down with the increased supply. That would be correct if the market for textbooks would satisfy all the canons of perfect competition. But it does not.

First, students are a captive market. They have to buy the textbook the teacher assigns, a teacher who generally is unaware of the cost of the textbooks available. Also, the teacher is constantly courted by representatives of the publishers. Also, once you have prepared a class, there is little incentive to change it. This makes it remarkably difficult for a new publisher to enter the market.

Second, as the PoET website shows clearly, textbook prices show remarkably little variance, at least among the "mainstream" ones. That really looks like (open or tacit) collusion among publishers.

Third, the textbook market is twisted in a particular way: teachers tend to choose the textbooks that are easy to teach from, either because the teacher gets all sorts of teaching material, or because the material is (too) simplified. Mankiw's textbooks are a perfect example: There are lots of definitions and descriptions, but little substance in terms of intuition, and that substance is outdated (i.e., from when the teacher learned it). That is why they are so popular. The modern textbooks (the newcomers), however, struggle. That does not look like free entry to me either.