Lessons from Maryland
From the SPARC Open Access Newsletter, June Edition 2009,
The defeated policy would have encouraged green OA (deposit in the institutional repository), encouraged gold OA (submission to OA journals), and required neither.
Is the Maryland vote ominous or anomalous? Either way, supporters of OA should try to understand it. Whatever its causes, they could arise again elsewhere. At the same time, we should understand why many stronger OA policies have been accepted at other campuses.
The Maryland vote was not the first faculty vote on an OA policy, but it was the first defeat. By my count, faculty have voted on OA policies at 20 universities. At 19, all but Maryland, the votes were affirmative. An impressive majority (12 out of 19) of the votes were unanimous. An even more impressive majority (18 out of 19) of the approved policies could be considered OA mandates, significantly stronger than the policy rejected at Maryland.
Part of understanding the Maryland vote is to understand why the weakest policy put to a faculty vote was the only one to be defeated.
(I list the 19 faculty-adopted policies in the second postscript below. There are more than 19 university OA policies overall, but only 19 have been adopted by faculty, as opposed to administrators. Of these 19, 14 were adopted before the Maryland vote and the rest after. The first faculty-adopted policy was at the Harvard Faculty of Arts and Sciences in February 2008. I'm not counting one recently approved policy which the institution has not yet announced; watch OAN for the details.)
According to the University of Maryland campus newspaper, one faculty concern was that the proposed policy might limit the freedom of faculty to submit work to the journals of their choice. The policy encouraged faculty to publish in OA journals "where practical and not detrimental to their careers." Some faculty feared that the president would turn the encouragement into a de facto expectation. "Both [women's studies professor Claire Moses] and [history professor Gay Gullickson] argued the resolution's language was too strong to count as a mere suggestion and would eventually lead to university policy. 'This does not call for discussion - it urges the president to take action,' Gullickson said."
The paper also tells us that some faculty were concerned that an OA policy would kill subscription journals. "'Open access will kill the journals you need during your career,' women's studies professor and university senator Claire Moses said. 'It's as simple as that.'" (Ibid.)
Both concerns are legitimate, even if neither is "as simple as that".
As to the first, I've argued that respecting faculty freedom to submit work to the journals of their choice is one of three principles that ought to govern university OA policies.
As to the second, I've acknowledged that the rising volume of green OA might eventually trigger some cancellations. However, the evidence in the field with the highest levels and longest history of OA archiving is to the contrary, and in any case cancelling toll access (TA) journals should not be equated with undermining peer review.
Universities with strong OA mandates preserve faculty freedom primarily by offering waivers or opt-outs on request. Since Harvard pioneered the explicit waiver option in February 2008, most institutional policies have followed suit. Faculty who want to publish in journals unwilling to allow OA on the university's terms only have to request a waiver. That leaves them free to submit their work to any journal they like and to publish in any journal accepting their work.
Ironically, because the Maryland policy mandated nothing, there was no need to build in a waiver provision. Hence, no one could point to an explicit waiver option to answer fears that encouragement might harden into an expectation.
Harvard added the waiver policy precisely to answer these fears. Stuart Shieber told Robin Peek at the time, "The provision was certainly important in assuaging some faculty members' worries that they could be held hostage by the policy in cases where it wasn't serving their best interests."
That concession allowed the Harvard Faculty of Arts and Sciences to vote unanimously for a stronger policy.
A waiver provision could also have addressed the Maryland faculty's second concern about journal survival. University OA policies with waiver options can't imperil subscription journals. When a journal concludes that it can't afford to allow OA archiving, and receives a submission from a faculty member at an institution encouraging or requiring OA archiving, then it only has to ask the author to request a waiver.
No one should blame the policy proponents at Maryland for omitting a waiver provision. Before Maryland, there was no reason to think that a policy encouraging rather than requiring OA would have needed one. Now, however, I believe we have reason to consider waiver options even in "mere encouragement" policies. Either that, or we have reason to be especially clear in explaining that encouragement policies already allow opt-outs precisely by stopping short of mandates.
Because encouragement is diffuse and without sharp boundaries, because it can be enforced by opinion, and because the pressure underlying it can rise and fall without formal action, it can be more fearsome than formal legislation. Formal legislation has sharper boundaries and is alterable by, and only by, well-understood formal procedures. Whether encouragement seems more insidious than a formal rule, or more insipid, will vary from campus to campus, just as faculty trust of administrators varies from campus to campus.
Waivers are not always necessary. On the very same day as the Maryland vote, across the country, the University of Washington approved a policy nearly identical to the one rejected at Maryland, including the absence of waivers. One year earlier, across the globe, Macquarie University approved an unequivocal OA mandate with no waiver provision; in fact, the vote was unanimous. But where faculty members worry about administrative encroachments on academic freedom, or green OA pressures on their preferred journals, an explicit and well-explained waiver policy should answer those worries.
I don't know all the variables in play at Maryland. (I asked a leader of the campaign for the Maryland proposal to comment on the cross-currents, but got no reply.) However, it appears that the two oddities about the Maryland policy --that it was the weakest of the policies put to a faculty vote and first to be defeated-- are connected. The first is part of the explanation of the second, even if it's not the full explanation. The gap between the Maryland policy and a mandate doesn't explain the faculty worries; other local variables explain those. But if the gap between the Maryland policy and a mandate explains the absence of a waiver option, then it explains why the faculty worries, once aroused, were so difficult to answer.
The lesson is not that stronger policies are always politically easier than weaker policies. The lesson is that weaker policies are not always politically easier than stronger policies, even apart from the question whether they are worth the trouble. When policies are strong enough to include waiver provisions, they can arouse fewer fears than weaker policies without waiver provisions. Exercising a formal waiver to side-step a requirement can be easier than bucking informal disapproval to side-step a non-requirement. If this is mysterious or paradoxical, consider how easy it is to opt out of liability insurance when renting a car (check!) compared to the difficulty of declining a dinner invitation.
Here are a few other lessons for other schools considering OA policies.
(1) Don't mess with faculty freedom to submit work to the journals of their choice.
If a given policy doesn't interfere with faculty freedom, don't assume that it will be perceived that way. Take pains to avoid interference and the appearance of interference.
(2) Green OA policies don't mess with faculty freedom.
In the past, I've distinguished two sorts of green OA mandate: loophole policies (requiring deposit in an OA repository except when a given journal doesn't allow it) and no-loophole policies (requiring deposit and requiring retention of the right to authorize OA). The first sort of policy clearly protects faculty freedom to publish in any journal. When the second sort includes a waiver provision, it clearly does so as well.
Going beyond a green OA policy to a gold OA policy can increase fears that the institution wants to steer faculty submissions to a subset of journals. Gold OA cannot be mandated without limiting faculty freedom.
As the Maryland and Washington experiences show, encouraging gold OA without mandating it sometimes arouses faculty fears and sometimes doesn't. Take the temperature of your campus and act accordingly. While waiver options can help, so can a narrower focus on green OA.
(3) Faculty leaders recommending green OA policies should make clear that green OA is compatible with publishing in TA journals. They should understand that this fact is not widely known and could even rank as the best-kept secret about OA. It will need clear and patient explanation.
Most TA journals already allow OA archiving. A rights-retention policy (for example, like those at the NIH or Harvard) can close the gap and assure that OA archiving is authorized regardless of the journal in which the author eventually publishes.
Someone might fear that a no-loophole, rights-retention mandate might lead some journals to reject work by authors subject to its terms. But that is not happening even with the NIH policy, which allows no waivers. Nevertheless, where it is a risk, a waiver option is a sufficient safeguard. Journals that don't want authors to retain the mandated rights can ask authors to request a waiver.
(4) Waiver provisions in green OA policies should only apply to OA itself, not to repository deposits. When authors obtain a waiver, then the work they deposit in the repository would remain "closed" or "dark" for a period of time matching the journal's embargo or moving wall.
Because dark deposits are not OA, they satisfy journals which don't allow OA archiving. They preserve faculty freedom to publish in any journal willing to accept their work, and still allow the institution to collect its full research output in the repository. The Harvard waiver option takes this form, as clarified in March 2009.
(5) Even weak green OA policies --encouragements rather than mandates-- could add waiver provisions to answer the fear about interfering with faculty freedom.
Some faculty will adopt even stronger policies without waiver provisions, as the Macquarie faculty proved. But once faculty worries arise, an explicit waiver option is the best way to answer them. Moreover, when a policy deliberately stops short of a mandate, then adding a waiver provision will improve its political chances without weakening its substance.
Many universities now have experience in drafting OA policies, anticipating faculty concerns, answering faculty concerns, shepherding policy proposals through a political process, winning faculty approval --more often than not by unanimous votes--, and implementing the adopted policies. At the same time, many other universities want to adopt policies and are just starting down the same road. Five years ago there wasn't much institutional experience to share, or much demand for it. But today there's a lot to share and a lot of demand.
For a sense of the rapid growth in the number of institutional OA mandates, or the rapid growth of transferable experience, see Alma Swan's revealing new graphics.
Two initiatives, both near launch, will help institutions considering OA policies avoid reinventing the wheel.
The first is a US-focused project led by SPARC. It will collect key documents from institutions which have successfully adopted strong OA policies and share them with institutions drafting their own or educating their constituents about the issues. It will also provide human help to share experiences from other campuses, answer questions, and advise on strategy and substance, as needed.
The second, more international initiative is Enabling Open Scholarship (EOS). EOS describes itself as "a membership organisation for universities and research institutions [and] a forum for raising and discussing issues around the mission of modern universities, particularly with regard to the creation, dissemination and preservation of research findings." The EOS Advisory Board will meet in Brussels in two weeks (June 15, 2009) and we can expect an official launch soon after. One of its top priorities will be to help universities adopt effective OA policies. Bernard Rentier is the EOS chairman and Alma Swan is the Convenor.
If your campus is considering an OA policy, you would do well to contact SPARC or EOS.
* Postscript 1. Since I've argued that a waiver provision will protect faculty freedom to submit to the journals of their choice, I want to point out two nuances that complicate the picture even if they don't affect the lessons I drew above.
First, the principle to respect faculty freedom to submit work to the journals of their choice applies more strongly to university policies than funder policies. As I argued in April 2008:
[This principle] 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....
As more TA journals convert to OA, and more accommodate university OA mandates, and as more universities adopt OA mandates, then universities may safely strengthen [their policies] by phasing out opt-outs or increasing the difficulty of obtaining them. If publishers accommodate university OA mandates, then opt-outs will not be necessary in order to protect faculty freedom to publish in the journals of their choice. When enough universities adopt OA mandates, we'll be there. But until then opt-outs preserve faculty freedom without reducing repository deposits....
(1) At these four institutions, the votes were unanimous in the faculty senate or equivalent body:
Macquarie University, University Senate and Council (April 27, 2008)
Boston University, University Faculty Council (February 11, 2009)
University of Calgary, Division of Library and Cultural Resources, Faculty Council (c. May 1, 2009)
University of Pretoria, University Senate (May 2009)
(2) At these eight institutions, the votes were unanimous in the whole faculty or relevant division:
Harvard University, Faculty of Arts and Sciences (February 12, 2008)
Harvard University, School of Law (May 7, 2008)
Stanford University, School of Education (June 10, 2008)
Oregon State University, Library Faculty (March 6, 2009)
Massachusetts Institute of Technology (March 18, 2009)
University of Oregon, Library Faculty (May 7, 2009)
University of Oregon, Department of Romance Languages (May 14, 2009)
Gustavus Adolphus College, Library Faculty, (May 14, 2009)
(3) At these seven institutions, policies were approved by non-unanimous faculty votes:
Stirling University, Academic Council (March 5, 2008)
University of Glasgow, University Senate (June 5, 2008)
University of St. Gallen, Faculty Senate (December 15, 2008)
University of Edinburgh, Electronic Senate (February 18, 2009)
Université catholique de Louvain, Academic Board (July 7, 2008)
Harvard University, John F. Kennedy School of Government (c. March 9, 2009) (98% approval)
University of Washington (April 23, 2009) (not a mandate) (adopted by "an overwhelming majority")
(4) These nine institutions adopted OA mandates since the first faculty-approved policy at Harvard in February 2008. I don't add them to the "unanimous" or "non-unanimous" columns either because I can't tell whether they were adopted by faculty votes or because I can't tell what the vote tally was. If anyone can help with these details, I'd be grateful.
Queen Margaret University (February 19, 2008)
Southampton University (April 4, 2008)
Napier University (now Edinburgh Napier University), Academic Board (April 25, 2008)
ETH Zürich (Eidgenössische Technische Hochschule Zürich) (September 29, 2008)
University of Liege (c. November 2008)
Vologda Scientific-Coordination Center of the Central Economics and Mathematics Institute (c. 2009)
Keldysh Institute of Applied Mathematics (c. 2009)
University of Salamanca, Consejo de Gobierno (February 27, 2009)
Ternopil State Ivan Puluj Technical University (c. April 2009)