Defining authorized users

I'm really interested to see where this ability to tightly define users and who has authorized access to databases goes. There are certainly some pros and cons for such a system.

Tightly defining users: Pros

1. It will very likely be cheaper. If libraries work with database vendors to give only pharmacy students access to pharmacy databases, for instance, then the price of the database might go down. This might be of particular importance to an academically diverse research institution, with lots of specific departments and tens of thousands of users.
2. Most people will not miss the loss of access to other databases. As a library student, I am mostly interested in the library and information science databases and some social science databases. Pharmacy students are mostly interested in pharmacy databases. I would not lose too much sleep over my loss of access to the pharmacy school databases, and the pharmacy student would probably be okay with losing access to the library science databases.
   

Tightly defining users: Cons

1. This really goes against libraries' commitment to creating access for users. Libraries have been moving to an access rather than ownership model for decades, now. (It is debatable whether or not this is a good idea, but moving to access certainly seems to be the recent and foreseeable trend.) Strictly defining authorized users go against a core mission of the library by limiting access. (Although, see Pros #2.)
2. There might be a handful of people who would lose access to databases that are critical to their research, simply because their credentials put them in the "wrong" department. For instance, a history of science student who studies the history of pharmacies in the United States, would be put at a serious disadvantage if he did not receive access to pharmacy materials. Tightly defining users might put some students and researchers at a serious disadvantage.
3. Limiting users to certain databases is moving against a very strong trend towards interdisciplinarity in academia. More and more often, university researchers are working across disciplines to advance research on specific problems (e.g., public health, crisis management, global warming, etc.). Limiting access to databases because you are in the "wrong" department could stifle interdisciplinary research. An new example of an institute that would pose a more extreme problem is the Wisconsin Institutes for Discovery, which is both interdisciplinary and a public-private partnership. (Read: some of it could be defined as commercial activity.) How would such a varied institute fit in as an "authorized user"? Sounds problematic to me.
4. This one is the kicker: database vendors are not idiots. They have a bottom line to defend. My gut feeling is that the cost of a database would actually not be that much cheaper if you tightly defined authorized users. The databases are the same and the amount of use is still going to be more or less the same for subject-specific databases. Any cost-savings would surely be modest at best, even if a database goes from being accessible to 100% of the university population to being accessible to 1%. Any librarian that thinks that there would somehow be a 99% discount is totally delusional. It seems like libraries could be giving up broad access (again, related to libraries' missions) for what is probably going to be a meager discount. Database vendors will not likely give up nice profits to cut libraries deals based on authorized access.
   

My cons list is a bit longer than my pros. But I think that there is room to experiment with this. This might be suited to certain institutions more so than others. (For instance, it might not be worth it to implement such a system at a small liberal arts or community college.) And it might also depend on the exact implementation of such a system. Would users themselves be able to define what they might like access to? Maybe each user could have default databases assigned by status type, but also be able to choose something like 3 databases outside of their default group. For instance, that history of science student could get all of the history databases plus choose 3 relevant pharmacy databases. That seems like an interesting alternative that would add flexibility while still limiting authorized use. I'm sure the technology for such an implementation already exists or could in the next few years. The problem would come in negotiating the license to reflect such flexibility. Sigh, it never ends!