"Our mission at JSTOR is supporting scholarly work and access to knowledge around the world. Faculty, teachers, and students at more than 7,000 institutions in 153 countries rely upon us for affordable and in some cases free access to content on JSTOR. Since our founding in 1995, we have digitized the complete back runs of nearly 1,400 academic journals from over 800 publishers. Our ultimate objective is to provide affordable access to scholarly content to anyone who needs it."
AHAHAHAHAHAHAHA, at 50K a year? When they won't even sell it to institutions they don't consider proper schools? Somebody needs to re-evaluate their mission statement...
50k is cheap as hell. UMass Boston had 125,000 JSTOR downloads last year. Let's assume we pay $50k (close enough), that's 40 cents an article. This is cheap as hell compared to other databases that charge more than that for a single search!
It's further harming the possibilities of being an independent scholar or even reading academic literature outside academia, though, which I think is a significant detriment to academia. I'm not as worried about the $50k a university library has to pay, but if you're an independent scholar for even one year, it becomes clear how harmful to the research community JSTOR is.
Journals that used to sell archive access to individuals for, say, $50 or $100 annually, now won't sell you a membership at all, because to save money on hosting they've moved their subscription infrastructure to JSTOR, and JSTOR refuses to sell individual subscriptions. So you end up "stealing" your access; at various times I've gotten my JSTOR access via ssh -D proxies to a friendly grad student. Now I try to return the favor by providing such access to independent scholars where needed. This sort of gaming shouldn't be necessary with a non-profit organization that is supposed to be working in the public interest, though. Hell, public domain journals from the 18th century that were scanned using public grant money are locked up behind a JSTOR paywall!
Many university libraries allow to you join as a "friend" of the library, which typically costs ~$100/year and gives you full access to all library resources.
I see your point, but you could just as easily walk into your local university, use their network as a guest, and access millions of dollars worth of content for free. Most databases allow you to download PDFs so you can download a bunch of articles and take them home, legally. Proxying through your grad student friends is breaking the law.
Even among state universities access to electronic journals for guests is becoming increasingly rare (usually prohibited in the license agreement). Also if you look to much into how scholarly communications works, the idea of charging 50k is insane since that vast majority of the cost of journals are actually paid for by universities in the first place.
edit: although I do agree with your point that in with the current state of things the cost per usage of JSTOR is significantly less than most other electronic journal collections
I would be impressed if it's an actual trend that guests were being blocked from accessing licensed content.
I work in this field and I can say that almost all databases and journals have IP authentication, generally with the entire campus IP range white listed. If the vendors don't implement that, they generally use password protected accounts. There are some alternatives such as uploading a list of barcodes, Referring URLs (I'm not kidding ProQuest for example allows this), Athens, Shibboleth, and a few others. These, however, are not commonly implemented by libraries because they are not widely adopted by vendors and require additional IT support that libraries simply don't have.
Therefore, unless libraries proxy all their users through their proxy servers (They don't for on-campus users. Usually, all links will go through the proxy, but if on campus the proxy will redirect you directly to the database to avoid the overhead), it would be near impossible to enforce the restriction of guests. I would wager that in many of these scenarios, it's all smoke and mirrors, and that access is really there if you know what you're doing (e.g. just visit the vendor site while on-campus without going through the library links).
I've been at at least one university (a smallish liberal-arts college) that used a separate foo_guest wifi network for guests, which dumped people into an IP range that wasn't in the database-access whitelist (enrolled students and faculty would connect to a different wifi network, using their username/password, that was whitelisted). I agree that that isn't usual practice, though, despite being technically required by some database agreements.
The configuration I've seen is to have not only guest accounts but also to have guest computers which are on a separate IP range, same with the guest wireless network. The important issue is actually what do the details of your license agreement say regarding guest users. Some state universities will actually request for this to specifically allowed since in many states it's considered, even if unofficially, a right of tax payers to have access if they need it
But many state schools let you join the library even if you're not a student. e.g. http://www.lib.iastate.edu/info/6197 - $20/year for a non-affiliated member.
They are probably bound by contracts with the actual content owners which limits who they can give access to. Not saying it's right but even the price is probably not all their doing.
AHAHAHAHAHAHAHA, at 50K a year? When they won't even sell it to institutions they don't consider proper schools? Somebody needs to re-evaluate their mission statement...