» Is the Harvard open-access policy legally sound? The Occasional Pamphlet

abernard102@gmail.com 2012-09-18

...evidenced by a written instrument... / …evidenced by a written instrument… To Sign a Contract 3” image by shho. Used by permission.

The idea behind rights-retention open-access policies is, as this year’s OA Week slogan goes, to “set the default to open access”. Traditionally, authors retained rights to their scholarly articles only if they expressly negotiated with their publishers to do so. Rights-retention OA policies—like those at Harvard and many other universities, and as exemplified by our Model Policy—change the default so that authors retain open-access rights unless they expressly opt out.

The technique the policies use is a kind of “rights loop”:

  1. The policy has the effect of granting a transferable nonexclusive license to the university as soon as copyright vests in the article. This license precedes and survives any later transfer to a publisher.
  2. The university can grant the licensed rights back to the author (as well as making use of them itself, primarily through distribution of the article from a repository).

The author retains rights by using the university as a kind of holding area for those rights. The waiver provision, under sole control of the author, means that this rights retention is a default, but defeasible.

This at least was the theory, but what are the legalities of the matter? In designing Harvard’s OA policy, we spent a lot of time trying to make sure that the reality would match the theory. Now, Eric Priest, a professor at the University of Oregon School of Law, has done a detailed analysis of the policy (forthcoming in the Northwestern Journal of Technology and Intellectual Property and available open access from SSRN) to determine if the legal premise of the policy is sound. The bottom line: It is. Those charged with writing such policies will want to read the article in detail. I’ll only give a summary of the conclusions here, and mention how at Harvard we have been optimizing our own implementation of the policy to further strengthen its legal basis.

Priest’s conclusion is well summarized in the following quote:

The principal aim of this Article has been to analyze the legal effect of “Harvard-style” open access permission mandates. This required first analyzing whether scholars are the legal authors (and therefore initial owners) of their scholarly articles under the Copyright Act’s work made for hire rules. It then required determining whether a permission mandate in fact vests, as its terms suggest, nonexclusive licenses in the university for all scholarly articles created by its faculty. Lastly, this analysis required determining whether those licenses survive after the faculty member who writes the article transfers copyright ownership to a publisher. As the foregoing analysis shows, in the Author’s opinion the answer to all three of these questions is “yes”: scholars should be deemed the authors of their works, and permission mandates create in universities effective, durable nonexclusive licenses to archive and distribute faculty scholarship and permit the university to license others to do the same.

Although Priest’s analysis agrees with our own that the policies work in and of themselves (at least those using the wording that we have promulgated in our own policies at Harvard and in our Model Policy), he notes various ways in which the arguments for the various legal aspects can be even further strengthened, revolving around Section 205(e) of the Copyright Act, which holds that “a nonexclusive license, whether recorded or not, prevails over a conflicting transfer of copyright ownership if the license is evidenced by a written instrument signed by the owner of the rights licensed”.

Priest argues at length and in detail that no individual written instrument is required for the survival of the nonexclusive license. But obtaining such an individual written instrument certainly can’t hurt. In fact, at Harvard we do obtain such a written instrument. There are two paths by which articles enter the DASH repository for distribution pursuant to an OA policy: Authors can deposit them themselves, or someone (a faculty assistant or a member of the Office for Scholarly Communication staff) can deposit them on behalf of the authors. In the first case, the author assents (via a click-through statement) to an affirmation of the nonexclusive license:

I confirm my grant to Harvard of a non-exclusive license with respect to my scholarly articles, including the Work, as set forth in the open access policy found at http://osc.hul.harvard.edu/ that was adopted by the Harvard Faculty or School of which I am a member.

In the second case, our workflow requires that authors have provided us with an “Assistance Authorization Form”, available either as a click-through web form or print version to be signed. This form gives the OSC and any named assistants the right to act on the faculty member’s behalf as depositor, and also provides assent to the statement

In addition, if I am a member of a Harvard Faculty or School that has adopted an open access policy found at http://osc.hul.harvard.edu/, this confirms my grant to Harvard of a non-exclusive license with respect to my scholarly articles as set forth in that policy.

Authors need only provide this form once; thereafter, we can act on their behalf in depositing articles.

Thus, no matter how an article enters the DASH repository, we have an express affirmation of the OA policy’s nonexclusive license, providing yet a further satisfaction of the Section 205(e) “written instrument” clause.

Priest mentions another way of strengthening the argument of survival of the nonexclusive license, namely, incorporating the license into faculty employment agreements, either directly or by reference. This provides further backup that the license is individually affirmed through the employment agreement. We take additional steps along these lines at Harvard as well.

One of the most attractive aspects of the default rights retention approach to open-access policies is that the author retains rights automatically, without having to negotiate individually with publishers and regardless of the particularities and exigencies of any later publication agreement, while maintaining complete author choice in the matter through the open license waiver option. It is good to know that a thorough independent legal review of our approach has ratified that understanding.