Is open access biting?

News that Elsevier is changing its policy on allowing authors to archive papers suggests that the publisher may be feeling the pinch of open access, as users turn to archives for copies rather than to the journal.  I don't know how many individual copies of papers Elsevier sells, in addition to its subscription list, but a copy obtained from an archive is a copy unsold.
The new policy requires institutions to have specific agreements with Elsevier:

...deposit in, or posting to, subject-oriented or centralised repositories (such as PubMed Central), or institutional repositories with systematic posting mandates is permitted only under specific agreements between Elsevier and the repository, agency or institution, and only consistent with the Publisher’s policies concerning such repositories.

 This policy applies to the manuscript version of an accepted paper.  For actual published papers, the policy is even more restrictive:

Elsevier does not permit the posting of PJAs (Elsevier-provided PDF or HTML files) on any open web sites.  This is to ensure that the final published version of an article, which has been edited and peer-reviewed according to the publishing standards of an Elsevier journal, is always recognized as such only via the journal itself, whether in print or electronic format.

Publishers behave this way because they believe that they are the owners of the research output and the answer to their stranglehold on the distribution of research findings lies in the hands of the institutions and the researchers.  It is only by refusing to transfer copyright and by insisting that the publisher only has the right to publish a specific version of a paper, that authors will be able to do what they wish with their research outputs.
The response of the Swedish open access programme to this change in Elsevier's terms is as follows:

•    We strongly object to Elsevier’s new policy which requires separate agreements for author’s rights and we urge Elsevier to withdraw the new clause.
•    We recommend Swedish universities with open access mandates to not conclude separate agreements with Elsevier. Instead, this issue should be transferred to the negotiations of the national license agreements with Elsevier.

It seems likely that institutions will be in conflict with the major publishers to an increasing extent as more and more adopt policies on open access to their research outputs. The SciComInfo newsletter reports, for example, that the Karolinska Institute, one of the world's leading medical research institutions, has adopted an open access policy, and many institutions are now using the deposit of papers in their institutional repository as one way of assessing research performance. Limiting open archiving in any way will run counter to these trends.