This has not been entirely my experience. Unfortunately, I only have a half-remembered anecdote rather than data, but what I've found is that journals in some university libraries are becoming less accessible in paper form. Instead, one is directed to a JSTOR subscription or such. Of course, in order to access the electronic version, one must authenticate to the network with a university ID. This erodes the effect you describe.
Can anyone else confirm or refute this trend? Am I merely imagining it?