This post asks whether open access publishers that practice strict adherence to the Creative Commons-Attribution license (CC-BY) are actually shooting themselves in the foot, that is, leaving plenty of room for competitors who will be able to use their works with no requirement to reciprocate. Note that it is not the competitors who are binding CC-BY publishers - they are voluntarily doing this to themselves! CC-BY publishers can easily avoid this situation, simply by adopting a flexible approach to licensing.
How can this be?
Let's compare CC-BY with what I would consider to be an enlightened approach to licensing for an open access publisher: the decision about licensing is left to the author, with the full range of creative commons license options made available, and provision made for sub-licensing portions of a work. An author can choose to publish a work with the CC option that I would consider optimal for ensuring open access not only for now, but into the future: CC-Attribution-Noncommercial-Sharealike (CC-BY-NC-SA). The author using this license can use CC-BY material, and even material with a different restrictive CC license, such as CC-BY-NoDerivatives (CC-BY-ND), as long as the material with the different license is marked with the appropriate CC license.
This flexible approach, leaving the choice with the author, supports one of the main philosophical arguments for open access: author's rights. With respect to choice of license, author's rights means that the author gets to choose, not the publisher. A publisher that forces a CC-BY choice is no more in support of authors' rights than a publisher that requires full transfer of copyright. Engaging authors in thinking about these choices is, from my perspective, an essential step towards articulating the commons - a discussion that we as an emerging global society need to have amongst ourselves.
A journal or book publisher using the flexible approach I recommend is free to publish material that a strict CC-BY publisher would not be able to publish - including the works that are published CC-BY.
To take another example: a publisher that uses CC-BY-NC as a default can publish whatever they like from the repertoire of a CC-BY publisher, but a strict CC-BY publisher will refuse to touch any work that is licensed CC-BY-NC. I predict that the more flexible approach will give new open access publishers (or traditional publishers transitioning to open access likely to consider more limited licenses) a competitive advantage over the strict CC-BY approach.
There is nothing to stop a CC-BY publisher from adopting a more flexible approach, with CC-BY as the default license, but permitting authors to use a different license if they prefer, and allowing for differential licenses for elements included in a CC-BY piece.
As a bit of context: Some of the early commercial and open access publishers, such as BioMedCentral and Hindawi, as well as the not-for-profit Public Library of Science, seem pretty strict about use of the Creative Commons-Attribution (CC-BY) license. For some people, this is the legal expression of the most open definition of open access from the Budapest Open Access Initiative. Klaus Graf expresses this point eloquently in the Journal of Librarianship and Scholarly Communication, with Sandy Thatcher expressing the counterpoint, the good reasons for considering more limited licenses. My own perspective is that CC-BY, while superficially appearing to be the expression of the most open form of open access, actually contains loopholes which make CC-BY a weak license for strong open access. For details of this argument, see the second chapter of my draft thesis - search for open access and creative commons. This post adds to and expands this argument, by suggesting that CC-BY publishers may be putting themselves at a competitive disadvantage. If your library or institution is supporting CC-BY publishers' article processing fees - I hope that you are storing the articles in your IR for preservation purposes, just in case...