Paper, printing, and binding (PPB) are typically a small fraction of the retail price of the book. It depends on a lot of factors, but as a ballpark figure, if you take the total cost of PPB for the nth edition of a big-selling 3-color college freshman textbook, and divide by the number of copies sold, it's probably ~$10, or ~ 1/10 of the ~$100 retail price.
Not only that, but the incremental price is even smaller. In traditional printing (not print on demand), printing costs are almost all setup costs. Once you have the print run going, the cost of producing one more book is nearly zero. Again, it depends on a lot of factors, but a ballpark figure would be about $1, or ~1% of the retail price.
This explains why publishers are seemingly so wasteful about sending out these unsolicited copies so indiscriminately to professors. It's not wasteful at all, because the incremental cost of printing a book is so low.
I always just pass the books on to students for free. This is good for the student who gets the book, and the money spent by the publisher to produce the book is so small that essentially no cost gets passed on to students. If I give it to a student, and the student then turns around and resells it, I think that's fine, too. It has the effect of undermining the $&%^#& evil publishers' exploitative pricing by helping to maintain a healthy market in used copies, while the nth edition is still in use. Part of their reprehensible scheme for maintaining exploitative prices is to kill off the used marked for edition n by rapidly bringing out edition n+1. The evaluation copy they send me might get out on the used market early enough to get some use before then.