It is a bit surprising that universities allow this practice — not so much because textbook costs are unbundled from tuition (things have been that way for a while) but because textbook company kickbacks may now constitute substantial fraction of an instructor's salary.
In the short term, this setup helps universities keep their teaching costs low — they can pay instructors a lower base wage and let the instructors make up the difference in textbook-company kickbacks if their conscience allows. This kind of cost-cutting might be attractive in a scenario where instructors are more "contractors" (adjunct faculty) than "employees" or "the reason the university is here and more important than administrators" (tenured faculty).
Schools should soon realize that it's a bad idea for their employees to have another revenue stream and, essentially, another supervisor influencing how they do the same job.
"this setup helps universities keep their teaching costs low — they can pay instructors a lower base wage"
It's a cost increase for students that's not as easily noticed as a "base" price increase in tuition. Which paying more to professors would require. As well as payroll taxes on the salary increase. So getting $1500 directly from publishers the same $1500 from the University costs extra in additional payroll costs (and then gets multiplied by any salary percentage increase in future years as well).
This is also somewhat analogous to what some elementary schools are doing (sorry no citation) with requiring parents to buy school supplies that were previously provided by the schools. It's a way of cost shifting that allows you to pass a tax that isn't recognized as a tax increase.
Other example may be an auto mfg. only raising the price of a new model slightly but removing and making certain features options, options that the majority of buyers will end up adding on anyway. (If the price of the car were higher many people wouldn't enter the showroom at all.)
Adjunct faculty are typically paid on a per-course-taught basis and can make as little as $400 per subject , although it's most common to see pay of about $1500–$2500 per subject. To pick one well-documented discipline in particular, psychology professors make a median of $2500 per subject taught at the graduate level .
This blog post guesses without giving evidence that the instructor may be making $5 off of each of his 250 students (who each paid $150 for a textbook), which would represent total textbook-company revenue of $1250.
To be sure, not all faculty are paid as poorly as adjuncts —  cites an estimate that adjunct faculty earn 26% less than tenured professors — but there is not a lot of money in teaching undergraduates, and for many kinds of instructors, textbook fees have the potential to become a substantial alternate revenue stream.
They sound what my uni called "bricks" - a bunch of licensed copies of reading material, all stapled together. They cost about $10-$50, as opposed to $50-$100 for a 101 text (note, this is Australia, the US has more expensive additions of textbooks for no good reason).
There are two kinds of bricks - draft textbooks, and collections of papers.
It's quite common to get a draft textbook as a brick one year, then laugh at the students the next year who pay for the textbook when it's finally published.
It seems a bit weird to publish a collection of papers as a textbook, especially at a low undergrad level, but I hear it's happening these days.
<I>It seems a bit weird to publish a collection of papers as a textbook, especially at a low undergrad level, but I hear it's happening these days.</I>
Took a macro-economics class at my local community college. One of the required texts was a collection of essays that the professor had written (much of it incoherent blabble but that's a different topic), collected into a 130-page paperback that looked like it might cost $3.99 at the local BN. Cost: $55. No copy available in the library.
Adjuncts do not generally write textbooks and get kickbacks for them. They choose one of the 'standard' textbooks, and use the online problem set for that book. They do this not because they get compensation for it, but because it's the path of least resistance.
I find it kind of sad that you would speculate like this, without any specific examples (let alone proof of a general trend!) about the motives of the adjuncts. :(
Hmm, my reading of the original blog post is perhaps a bit different than yours.
This post seems to provide evidence that people who do not traditionally write textbooks (not the top hundred faculty members in a field) are being afforded a new opportunity by textbook companies to "write a book", which either means "pick and choose some portions of the standard textbook" (discussed elsewhere in the comments) or "publish a 'course reader'-style collection of readings as a book" (discussed in this blog post).
It seems to me that this new opportunity also represents a new revenue stream for more faculty members than currently receive textbook kickbacks.