We have several mass digitization initiatives underway. And there seems to be an expectation that these will continue, that more of our current collective book and journal collection will be digitized. There are a variety of drivers for this, both for libraries and for the other organizations which have stepped up to serve and to resource such initiatives.
We are used to thinking that the library 'owns' its print collections, that subject to certain restrictions, it can do what it will with them. Among these restrictions are copyright ones.
Our Google 5 analyis suggested that more than 80% of books in the Google 5 library collections were published post 1923. This means that about 20% of books in those collections are out of copyright.
We do not currently have very easy ways of knowing which post-1923 works have gone through the copyright renewal process.
And one of the interesting service requirements to emerge around the mass digitization initiatives is rights tracking and notification. Libraries want to know the copyright status of materials at various points in potential workflows, including at the point of selection for digitization.
What this means is that a large part of library collections is still in copyright. The library 'owns' the cost of storing it, shelving it, keeping it at the right temperature, and so on. It can be shared and borrowed in its current form. However, the library does not 'own' it to the extent that they can freely re-format it and allow it to be used by many parties.
In this sense, the gap between the materials that libraries 'own' and the materials that libraries license is smaller than we are used to thinking about.