Three principles for university open access policies
SPARC Open Access Newsletter, issue #120
April 2, 2008
by Peter Suber

On March 17, Harvard's Berkman Center for Internet and Society hosted a workshop on university OA policies and asked each participant to write a position statement. 

I was a participant and began writing out detailed model university policies on OA repositories, OA journals, university presses, theses and dissertations, and promotion and tenure criteria.  Although I knew what I wanted to say under each head, the draft quickly grew large and ungainly.  I changed course and decided that it would be easier to read and write a small number of principles than a large number of policy details.  In the end I was able to say all that I wanted to say as consequences of just three principles.

I limited myself to OA for peer-reviewed journal literature.  It's my top priority and already a very large topic.  With regret, therefore, I omitted OA to data, monographs, and courseware. 

The other position statements were first-rate and I hope they will be distributed soon.  Here's a slightly revised version of my own.


* Principles

1.  Universities should provide open access (OA) to their research output.

2.  Universities should not limit the freedom of faculty to submit their work to the journals of their choice.

3.  Universities now pay most of the costs of peer review, through subscription fees and faculty salaries.  They should continue to bear the costs of peer review, in order to assure its survival, while recognizing that the forms and venues of peer review are changing. 


* Annotations

1.  Universities should provide open access (OA) to their research output.

1.1.  Universities should launch OA institutional repositories (IRs) and adopt effective policies to fill them with their research output.  That is, they should actually provide OA to their research output, not just wish for it, request it, encourage it, settle for ineffective policies to provide it, or sign statements calling for it elsewhere. 

1.2.  There is good evidence that merely requesting or encouraging faculty to deposit their work in the IR has little effect.  By contrast, there is good evidence that mandates work.  However, the word "mandate" may be misleading; there is good evidence that successful policies use mandatory language but rely on expectations, education, assistance, and incentives, not coercion.  (If there were a better word than "mandate", I would use it.)  There is also good evidence that the overwhelming majority researchers would willingly comply with an OA mandate from their funder or employer.

1.3.  The OA mandate should require OA through the IR, not through OA journals.  See Principle 2.

1.4.  The OA mandate should apply, at least, to the peer-reviewed manuscripts of faculty journal articles and the theses and dissertations of grad students.  It should make exceptions for royalty-producing books and should be open to other exceptions. 

1.5.  Universities have roughly two options for creating a legal basis to distribute OA copies of peer-reviewed manuscripts by their faculty.  First, they can seek permission from publishers, and only distribute OA copies when they succeed in obtaining it.  Second, they can ask faculty to retain the right to provide OA on the university's terms (and grant the university non-exclusive permission to provide that OA), even if faculty transfer all their other rights to publishers. 

The second option can support OA for 100% of the faculty research output, while the first option would support much less.  According to SHERPA (March 10, 2008), 57% of surveyed publishers currently give blanket permission for postprint archiving.  Even if half the remaining publishers permit it on request, the second first option would still fall 20% short of the first second option.  Principle 1 requires the approach that yields the highest rate of OA.  See the Appendix for more detail.

1.6.  To help faculty who may not understand copyright law, or who do not want to negotiate with publishers, the university should adopt an author addendum which allows the author to retain the rights needed to implement the university policy.  (An "author addendum" is a lawyer-written document that authors sign and attach to a publisher's standard copyright transfer agreement.  It modifies the publisher's contract to allow authors to retain some rights which the default contract would have given to the publisher.)   The university needn't write its own addendum; there are already good ones from CIC, JISC/SURF, OhioLink, SPARC, three from Science Commons, and a handful of others from individual universities.  After adopting or recommending one, the university should allow faculty to use any author addendum which enables them to retain the same or larger set of rights. 

1.7.  When universities need to see a list of a faculty member's recent journal publications (e.g. for promotion, tenure, or post-tenure review), they should either draw up the list from the IR or request the list in digital form with live links to OA copies in the IR.  They should tell faculty that they will limit their review of journal articles to those on the list, unless the faculty member writes a special justification for the Dean.  Policies along these lines are already in effect at Katholieke Universiteit Leuven, Labortoire de Psychologie et Neurosciences Cognitives (at the University of Paris - Descartes), Charles Sturt University, and the National Research Council Canada.

1.8.  Once a university has an IR, it may want to fill it with many kinds of digital content other than its research output, e.g. courseware, conference webcasts, digital collections from the library, and administrative records.  That is desirable, but should not delay progress toward the goal of providing OA to the institution's research output.

1.9.  Universities should encourage peer institutions to provide OA to their research output.  Every university gains from every other university's OA policy.  Universities should talk to their consortial peers, their regional neighbors, and their associations.  There are already provost initiatives centered at the U of Liege, the U of Brasilia, and SPARC, to spread the message to other provosts, and there are OA initiatives from associations as varied as the Committee on Institutional Cooperation, Conference of Italian University Rectors, Council of the Rectors of Portuguese Universities, European University Association, Finnish Council of University Rectors, German University Rectors' Conference, Irish Universities Association, the New England Council of Presidents, Norwegian Association of Higher Education Institutions, the Oberlin Group, Southern African Regional Universities' Association, and Universities UK.

1.10.  OA here means at least removing price barriers (making content free of charge).  But when possible, the policy should remove permission barriers as well (making content free of unnecessary copyright and licensing restrictions).  For example, the U of Auckland not only provides free online access to student theses and dissertations, it also releases them under Creative Commons licenses. 

2.  Universities should not limit the freedom of faculty to submit their work to the journals of their choice.

2.1.  If it weren't for Principle 2, universities could require faculty to submit their articles to OA journals rather than deposit them in an OA repository (a gold OA mandate rather than a green OA mandate).  But there aren't yet enough OA journals; there aren't yet first-rate OA journals in every research niche; and even one day when there are, a university policy to rule out submission to a journal based solely on its business model would needlessly limit faculty freedom.  Not even the urgent need for OA justifies that kind of restriction, as long as we can achieve OA through OA repositories.  That's why all university and funder OA mandates focus on green OA (through OA repositories) rather than gold OA (through OA journals). 

But of course OA journals still deserve support.  See Principle 3. 

2.2.  If annotation 2.1 doesn't stand on its own, it may be because it presupposes another premise.  As I put it elsewhere:  "The purpose of the campaign for OA is the constructive one of providing OA to a larger and larger body of literature, not the destructive one of putting non-OA journals or publishers out of business. The consequences may or may not overlap (this is contingent), but the purposes do not overlap." 

2.3.  If it weren't for Principle 2, universities could require faculty to deposit some version of their peer-reviewed journal articles in the IR, for OA, with or without an embargo, and faculty would have to avoid journals that did not allow OA archiving on those terms.  But that would needlessly limit faculty freedom to submit to the journals of their choice.  To respect faculty freedom, universities must allow exemptions (waivers, opt-outs) for faculty submitting to journals that do not allow OA archiving on the university's terms.  However, when enough universities adopt OA mandates, then all journals would have to accommodate them, and therefore the first type of policy (no opt-outs) would no longer limit faculty freedom or violate Principle 2.  But until we approach that point, Principle 2 requires the second type of policy (with opt-outs).  Moreover, allowing an opt-out on OA is compatible with not allowing an opt-out on IR deposits themselves.  See the Appendix for more detail.

2.4.  The strategy to require OA archiving, and to require researchers to avoid publishers that will not allow it, was pioneered by the Wellcome Trust.  The WT's example has been followed by some other funding agencies, most notably the UK Medical Research Council and the US National Institutes of Health.  Because I support these policies, as well as annotation 2.3, I should therefore point out that Principle 2 is designed for universities, not funding agencies.  Funding agencies are essentially charities, spending money on research because it is in the public interest.  They have an interest in making that research as useful and widely available as possible, and virtually no competing interests.  Universities have the same charitable purpose but many competing interests, such as nurturing researchers more than research projects, nurturing them over their entire careers, and erecting bulwarks of policy and custom to protect academic freedom.

2.5.  If we hasten the day when all or most journals allow postprint archiving, then we hasten the day when universities could adopt no-opt-out OA policies (as opposed to both no-opt-out deposit policies and opt-out OA policies) without violating Principle 2.  One way to do that is for universities to demand the right for postprint archiving when negotiating licensing terms for subscription or renewal.  OhioLink publicly committed itself to this strategy in 2006, the only library consortium I know to do so.  (OhioLink is a consortium of 86 academic libraries in Ohio representing more than 600,000 faculty, students, and staff.)  Several major universities are also trying this strategy, but so far without a public announcement.  Public or private, I recommend that all universities do what they can to negotiate better terms for their authors, not just better terms for their readers.

3.  Universities now pay most of the costs of peer review, through subscription fees and faculty salaries.  They should continue to bear the costs of peer review, in order to assure its survival, while recognizing that the forms and venues of peer review are changing. 

3.1.  Whenever a university library saves money in its serials budget from the cancellation, conversion, or demise of toll-access (TA) journals, it should spend the savings first on peer-reviewed OA journals.  This will insure that the same money which formerly paid the costs of facilitating peer review at TA journals will continue to pay the costs of facilitating peer review at OA journals. 

(BTW, my second priority for the savings is to replenish the book budgets suppressed by the high costs of TA journals.)

Many publishers argue that the growth of OA will undermine peer review.  It's important for universities to recognize that the key factor in the survival of peer review is not the growth of OA but the willingness of universities to support peer-reviewed OA journals with the same funds they formerly used to support peer-reviewed TA journals.

3.2.  Universities should create a fund to pay reasonable publication fees at fee-based OA journals (or subsidies to no-fee OA journals).  Such funds are already in operation at the U of Amsterdam, U of California at Berkeley, U of North Carolina at Chapel Hill, U of Nottingham, and the U of Wisconsin at Madison.

3.3.  Why is it important to recognize that the forms and venues of peer review are changing?  So that universities can support effective forms of quality control wherever they occur, not just the traditional forms which may over time represent a shrinking percentage of peer-reviewed literature.  Universities should use promotion and tenure criteria to reward all forms of excellent scholarship, not just the subset delivered in certain conventional packages.  When new peer-reviewed journals and new forms of peer review are worthy, universities should not discriminate against them or create disincentives for faculty to submit their work to them.  Universities should continue to try to weed out second-rate scholarship, but they should not do it with such crude criteria that they rule out much that is first-rate at the same time.  They should not disregard or devalue new forms of high-quality review or (by disregarding and devaluing them) put a brake on their continuing evolution.


* Appendix:  Quick summary of seven policy options

A.  Require submission to OA journals. 
(Unattested, though once considered by the Australian Productivity Commission.  Included on this list only for completeness.)

B.  Request and encourage OA through the IR, except when publishers do not allow it.
(Examples:  Athabasca U, Swedish U of Agricultural Sciences.)

C.  Require OA through the IR, except when publishers do not allow it. 
(Examples:  UK Economic & Social Research Council, Canadian Institutes of Health Research.)

D.  Require OA through the IR, and require faculty to avoid publishers that do not allow it. 
(Examples:  Wellcome Trust, UK Medical Research Council, and US National Institutes of Health.)

E.  Require OA through the IR, and permission from faculty, and allow opt-outs when faculty want to publish with publishers that do not allow it. 
(Examples:  Harvard Faculty of Arts and Sciences, draft policy of U of California.)

F.  Require deposit in the IR, no exceptions, and make the deposits OA whenever the author or institution can obtain permission. 
(Examples:  Recommendation of European Research Advisory Board, draft policy of Irish Research Council for Science, Engineering & Technology.)

G.  A blend of E and F:  Require faculty to grant permission for OA through the IR (with an opt-out), but also require deposit in the IR (without an opt-out). 
(Examples:  Recommendation of Stevan Harnad.)

None of these is perfect.  Policy A violates Principles 1 and 2.  Policies B, C, and E permit exceptions, falling short of Principle 1.  Policy D violates Principle 2.  Policies F and G permit OA exceptions, even if not deposit exceptions, falling short of Principle 1. 

Some shortfall from 100% OA is probably inevitable, like the friction in a machine, and a small shortfall is harmless.  I even believe that some deliberate exceptions (as opposed to unintended failures) could be desirable, for example, to muster support to pass the policy and to accommodate unexpected circumstances.  We don't have to pretend to anticipate unanticipated cases; it's enough to make OA the default and ask those seeking an exception to bear the burden of proof.

My preference is to think about which policy will bring an institution closest to 100% OA (Principle 1) without violating academic freedom (Principle 2).  I rule out A because there aren't nearly enough OA journals today; the shortfall and violation of academic freedom would both be large.  I rule out B because mere requests and encouragement generate low compliance rates; the shortfall would be large.  I rule out C because about one-third of journals do not allow OA archiving; the shortfall would be large. 

I rule out D because it violates Principle 2. But I only rule it out for universities, not funders, and only for the present.  As I argued in annotation 2.4, Principle 2 binds universities more than funders, and as I argued in 2.3, over time policy D will violate Principle 2 less and less.  These qualifications are important because only D-type policies eliminate the shortfall caused by author or publisher opt-outs.  When the violation of Principle 2 becomes negligible, D-type policies will be the clear winners.

I like E, F, and G, even today.  In each case, it appears that the shortfall would be fairly small.  If true, then in choosing among them the stakes are low and I would applaud any university for adopting any of them.  However, we should still look closely at which would have the smallest shortfall and be ready to refine our policies in light of what we find.

As I argued in my March 2008 newsletter, over time policy E will fall short of Principle 1 less and less.  For similar reasons, over time, policies F and G will fall short of Principle 1 less and less.  Whatever their initial deficiencies, I'm persuaded that all three of these policies will steadily, even if asymptotically, close their loopholes and improve their conformity to Principle 1.


Read this issue online

SOAN is published and sponsored by the Scholarly Publishing and Academic Resources Coalition (SPARC).

Additional support is provided by Data Conversion Laboratory (DCL), experts in converting research documents to XML.


This is the SPARC Open Access Newsletter (ISSN 1546-7821), written by Peter Suber and published by SPARC.  The views I express in this newsletter are my own and do not necessarily reflect those of SPARC or other sponsors.

To unsubscribe, send any message (from the subscribed address) to <>.

Please feel free to forward any issue of the newsletter to interested colleagues.  If you are reading a forwarded copy, see the instructions for subscribing at either of the next two sites below.

SPARC home page for the Open Access Newsletter and Open Access Forum

Peter Suber's page of related information, including the newsletter editorial position

Newsletter, archived back issues

Forum, archived postings

Conferences Related to the Open Access Movement

Timeline of the Open Access Movement

Open Access Overview

Open Access News blog

Peter Suber

SOAN is licensed under a Creative Commons Attribution 3.0 United States License.

Return to the Newsletter archive