There's been article after article on the serials crisis. I keep hearing this rhetoric that journal prices are just too high and university libraries can no longer afford them.
I'd like to see how much journals cost university libraries with my own eyes. Where should I look?
This is a complicated subject. In addition to what Brian Cody writes above, there's also Ulrichs Web, which is a serials directory with pricing data. Your academic library probably has a subcription to it. However, what makes this complicated is that librararians rarely buy journals by title. Usually they buy a bundle of journals. Often this price is negotiated and librarians often have to sign confidentiality agreements about the negotiated price. Furthermore, these bundles often come with journals the library may not want, but for a host of reasons, purchasing a bundle is cost effective. Also some vendors are allowed to license some journals and not others, but there may be overlap in what any two vendors are able to license. So that if you buy a bundle from these two vendors, the two bundles may provide content to shared journal titles, although the publication date range may be different for each vendor. Anyway, this is just the tip of the iceburg. It gets more complicated. And a lot of it has to do with copyright (which makes a journal a monopoly when copyright is transfered to them by an author).
I'd recommend starting with Journal Prices (http://www.journalprices.com), where you can search and filter journals by price, type, title, publishers, etc. The site is maintained by Ted Bergstrom and Preston McAfee, and the former's blog (http://www.econ.ucsb.edu/~tedb/Journals/jpricing.html) is full of research about journal pricing.
For summary statistics, I would check the Association of Research Librarians' site (http://www.arl.org/sc/marketplace/journals/index.shtml?) and this 2010 study by Allen Press (allenpress.com/system/files/pdfs/library/ap_journal_pricing_study_2010.pdf).