In which I make what, coming from me, will sound like a rather odd objection
In the CoHE, Boyd Zenner (sorry, behind the paywall), discussing the sorry state of contemporary academic publishing, places part of the blame at faculty feet: "It's pretty hard to understand why so many card-carrying members of the knowledge industry can't, or don't, or won't purchase the instruments necessary for the plying of their trade." Zenner concedes that "[p]eople complain that scholarly books are expensive, and of course they are, relatively," but that "of course" covers up some rather astonishing sins of omission--like, for example, any mention of how much scholarly books actually cost. An "inexpensive" monograph from OUP will be about $65 (and can easily run into the $100s and even $200s); from Cambridge, in the $80s (ditto). Cornell makes it into the $50s. Virginia, a "relatively" inexpensive press, retails literary criticism in the $30s and $40s. Books published by the commercial academic presses--Palgrave Macmillan, Ashgate, Routledge--frequently sell for anything from the mid-$70s to the low $100s. (My own Book One, published by Ashgate, currently goes for $110.) These books are not merely expensive in relation to other books; they're expensive in relation to groceries and the gas bill. One does not acquire a $200 book from Cambridge without much soul-searching (not to mention belt-tightening). That being said, Princeton, Harvard, Yale, Ohio State, and the University of Chicago manage to regularly stock books in the $30s and $20s--because, not coincidentally, they tend to do relatively rapid paperback releases. If I'm going to purchase a retail copy of an academic book, it's most likely going to be from one of the last five. After all, I would rather like to have some money left over after the transaction, which may be considered an odd predilection in some quarters.
For most academics, buying scholarly monographs is a strictly utilitarian activity--and for those of us in literature, one that competes with buying literature. (I work on the history of historical fiction; what else are contemporary historical novels but the "instruments" of my "trade"?) As a graduate student, the biggest mistake I made was buying monographs that looked interesting, as opposed to the ones I was actually going to use more than once. Interesting stuff got read once, sat morosely on the shelf, and then got donated to other graduate students. If something looks like an interesting read, I check it out from the library or ILL it; if, after perusal, it turns out to be useful for more than one project, then I buy my own copy. Given that many publishers key their book prices to institutional buying power, it cannot come as a great shock that most of us prefer to leave the monograph-buying to the institutions. (Pace Zenner, most of us don't expect our books to sell particularly well, unless we're the second incarnation of Jacques Derrida or something. We expect them to sell to...wait for it...libraries.) Most article-length projects call for many loans and few purchases--which has nothing to do with the value of the books, and everything to do with the nature of research. I'm always happy to make an impulse purchase when I see an intriguing title available cheaply (sort of) at a secondhand store, but that still leaves me with a very small subset of monographs that I need to own.