Who owns the copyright in works created by university faculty under US law is contested, and has recently become relevant with the advent of commercially valuable digital courses created by faculty members.
It is also highly relevant for traditional scholarly publishing, however.
Now assume, for argument's sake, that the teacher exception to the work made for hire doctrine does not apply after the coming into force of the 1976 Copyright Act, and universities do own the copyright in the work of their faculty, provided the individual employment contract does not stipulate anything to the contrary. The wording of the relevant § 101 Copyright Act is certainly broad to entertain this possibility. It appears that up to 1990s, most employment contracts with university professors did not address copyright ownership in works created by faculty, but maybe some reader has more insight. So we have a potential 20 year or so window in which the universities, not the professors, own the copyright in the scholarly writings of the professors.
When an academic publisher accepts a work for publication, the author often has to assign the copyright in the work to the publisher, or at the very least grants the publisher a non-exclusive license for publication and distribution. Problem is, "nemo plus iure transferre potest quam ipse habet" - you cannot assign a right you do not own. If the copyright in the work was owned by the university, it is being published and distributed by the academic publisher without the copyright owner's consent.
As some have pointed out - and I unfortunately do not know the original source for this theory - this could potentially expose academic publishers to massive liability. US Copyright law knows statutory damages for copyright infringement of between $ 750 and $ 30,000 per work (up to $ 150,000 in case of wilful infringement, but the publishers here hardly commit the infringement wilfully). Elsevier, the largest scholarly publisher worldwide, publishes approximately 400,000 articles annually, and its archive contains 13 million documents. If for even a fraction of those works it does not have the necessary rights for publication, the statutory damages could exceed not only Elsevier's annual revenue, but global GDP.
Since scientific publishers have been increasing subscription fees for journals to an extent university libraries publicly call "unsustainable" and privately "extortionist", universities may be motivated to sue, if only to increase their bargaining power.
One immediately apparent problem with such a suit, should it be successful, would be the recourse scientific publishers may take to the authors having assigned rights they did not own. Faculty may be in no position to actually compensate scientific publishers for statury damages awarded against the publishers, but universities should be hesitant to throw their faculty under the bus. So we may not be seeing this legal action soon, although I for one would love to see some of the profits of academic publishers being returned to universities.