Terry Martin (Law Librarian and Professor of Law, Harvard Law School) wrote in the American Scientist Open Access Forum:

Stevan, I'm sure your version is preferable to the one actually passed by FAS. Some of us urged a more forceful approach. However, those with a better political sense thought otherwise.

Note also that only the Faculty of Arts and Sciences - large as it is - has accepted this policy. It has yet to be debated at the schools of law, business, medical, education, design, divinity, or public administration.

Terry, the Harvard OA Mandate is potentially so important that I hope both political sense and pragmatism still have time to prevail, so as to remedy the few but fundamental flaws in the current draft Mandate.

The irony is that the Harvard Mandate needs a less forceful approach, not a more forceful one, in order to be much more powerful and effective.

Remedying Flaws in the NIH and Harvard OA Mandates.  An analogy with the history of the NIH OA policy is especially instructive and revealing here: The original draft of that NIH policy also had (three) fundamental flaws -- (1) that it was not a mandate (it was a request rather than a requirement), (2) that it allowed deposit itself to be delayed as long as a year, and (3) that it insisted on direct central deposit (in PubMed Central) instead of deposit in the author's own university's Institutional Repository (IR) (and then central harvesting to PubMed Central).

The NIH policy failed, completely, but it took three years to realize and remedy this failure. The remedy was (1) to upgrade the NIH policy to a mandate, (2) to require immediate -- not delayed -- deposit (allowing an embargo of up to one year, but applicable only to the date at which access to the deposit was set as OA; the deposit itself had to be immediate).

(The NIH insistence on central deposit (3), instead of institutional deposit and central harvesting, has not yet been remedied. But Harvard's institutional mandate, if its own flaws can be corrected so its policy can be adopted by all universities, US and worldwide, will also remedy this last of the three NIH mandate's flaws.)

Four years ago I went to Washington to try to explain to Norka Ruiz Bravo's group at NIH exactly how and why the three small but crucial changes (1)-(3) in the draft NIH policy needed to be made if it was to succeed. It was decided not to heed the advice, and to go ahead and adopt the draft policy as it was. As a result, three more years of NIH research access and impact were then lost, needlessly. (And during those three years, all the biomedical funders on the planet were reflexively imitating the failed NIH policy!)

Mandate Immediate Deposit.  Now NIH has it almost right. The two most important corrections have been made: (1) It is now upgraded to a requirement rather than just a request, and the (2) requirement is for immediate deposit, not delayed deposit. That's called the Immediate-Deposit/Optional-Access (ID/OA) mandate.

(The access embargo is applicable only to the date of OA-setting, not the date of deposit -- which is just fine, because of the potential power of repositories' automatized "email eprint request" Buttons to fulfill all research usage needs for Closed Access deposits during any embargo interval).

The one remaining flaw in the NIH mandate is the one that the right mandate from Harvard now would help to remedy:

Mandate Institutional Deposit.  Funder and University OA Mandates need to be complementary and convergent. NIH still insists on mandating direct central deposit in PubMed Central -- instead of reinforcing and building upon university mandates by likewise mandating deposit in each author's own University Institution Repository (IR). (From there, PubMed Central -- and any other central repository or indexer -- could harvest the deposits, with the author needing only to provide NIH with the URL!). Universities and research institutions are, after all, the providers of all research output, both funded and unfunded.

A realistic, viable mandate from Harvard now would be taken up by all universities worldwide, and it would ensure that funder mandates, too, would begin stipulating direct, convergent deposit in authors' own University IRs, instead of divergent deposit, helter-skelter, in diverse CRs. Then any central collections and indexes we desire could automatically be harvested (or exported) from the worldwide distributed network of OA IRs -- using the OAI metadata-harvesting protocol, which was created for that specific purpose, and with which all IRs are compliant.

Funder/University Collaboration to Make All Research Output OA.  Then universities and research institutions will not only be able to help funders implement and monitor compliance with the funders' self-archiving mandates (as part of their fulfillment conditions for receiving the institutional overheads from those grants): They will also be able to complement and universalize those funder OA mandates, which only cover the self-archiving of funded research output, with (Harvard-style) institutional OA mandates of their own, for self-archiving all of their research output, both funded and unfunded. That is the natural synergy that will systematically scale up to make all of the planet's research output OA -- across all disciplines, institutions, languages and nations.

Mandating Copyright-Retention is a Needless Deterrent.  But if NIH's mistake had been to make its mandate too weak (only (1) a request, with (2) an allowable year-long delay in deposit), Harvard's mistake now is needlessly making its mandate too strong, thereby necessitating the further addition of a compensatory "opt-out" clause -- which in turn makes the Harvard policy needlessly weak (indeed, no longer a mandate at all)!

The reason Harvard had to add its opt-out clause is obvious: Otherwise the policy would have faced a predictable (and justified) author revolt: (That is the sense in which "better political sense" prevailed!)

It is one thing to demand that the article be deposited in Harvard's IR. That just costs a few keystrokes, and brings palpable benefits to the author and institution. But it's quite another thing to demand that the author accept the risk of failure to successfully negotiate copyright retention with his journal of choice, and thereby being forced to publish in a lesser journal.

An Opt-Out Mandate Is Not a Mandate.  (Whether or not this risk is real, it is definitely a reasonable, perceived risk for publish-or-perish authors, even at Harvard, and hence a risk that rightly obliged those with "better political sense" at Harvard to add the opt-out clause. Without the opt-out clause it is unlikely that the policy motion could have passed at all.)

But with an opt-out clause, a mandate is no longer a mandate: it's just a request! And the reason the NIH request had failed was that it had been just a request rather than a mandate!

Allow Opt-Out Only From Copyright Retention.  The right remedy is hence to modularize the mandate, so as separately (a) to require immediate deposit, with no opt-out, and also (b) to request/require copyright-retention -- but to allow an opt-out from the latter.

Copyright Retention Is Desirable But Unnecessary for OA.  In reality, for at least 62% of refereed postprints and a further 29% of pre-refereeing preprints, there is no need for copyright retention at all in order to provide OA because this 91% of journals have already officially endorsed immediate OA self-archiving in one form or the other.

(Online access, free for all, by the way, moots all of the other uses for the sake of which authors -- and lawyers and librarians -- imagine that copyright would need to be retained and licensed: Virtually all the other uses already come with the territory, once a paper is made free online; and whatever doesn't, soon will, once all papers are made free online.)

To fulfill all immediate research usage needs for any of the articles from those journals that have not yet endorsed immediate OA self-archiving (38% for postprints, 9% for preprints), the IRs will all have the semi-automatic almost-immediate, almost-OA Button.

(This will not only provide for all immediate usage needs during any embargoes, but it will also soon bring the rest of the journal policies into line with those that already endorse immediate OA self-archiving, under mounting pressure from the worldwide research community's growing experience with -- and increasing reliance and eventual insistence upon -- the palpable benefits of OA, thanks to the growing number of mandates.)

Weaken the Harvard OA Mandate To Strengthen It.  So nothing is lost by weakening the Harvard policy, as recommended, so as to make immediate deposit mandatory, with no opt-out, allowing opt-out only on the (unnecessary) requirement to retain copyright.

Eliminate Need For Opt-Out Loophole.  In the current draft of its mandate, by bundling the two together, and allowing opt-out on the entire package, Harvard instead gets the worst of both: the author deterrent effect of insisting on copyright retention, plus the resultant loophole from Harvard authors simply choosing to opt out of depositing altogether.

Harvard's rationale was that its authors, if they have to opt out paper by paper, will find it too tiresome to keep opting out, and will prefer to try to renegotiate copyright with their journals instead. Much more likely, authors will draft standard form letters for the Provost's office saying "The right journal for this work is journal X, and journal X does not allow copyright retention, so I regret but I must opt out for this work.").

An opt-out letter is ar easier (and more likely of success) than the (real and apparent) risks of trying to renegotiate copyright. And the Provost's Office is certainly in no position to argue with Harvard authors on the appropriate outlet for their work.

Authors Will Not Deposit Unless Mandated.  Journals would no doubt be quite pleased if Harvard decided to over-reach and needlessly mandate copyright retention, with opt-out, instead of just mandating immediate deposit, without opt-out:

That way the journals that have endorsed immediate OA self-archiving could appear progressive on OA, confident that absent a deposit mandate very few authors bother to self-archive spontaneously (as the first NIH policy and many other non-mandatory policies have repeatedly shown: indeed, that was exactly what gave rise to the deposit mandate movement).

So because the current wording of the Harvard policy does not mandate the deposit itself, but only the copyright retention, along with the option of opting out, the journals can again count on most authors taking the path of least resistance: opting out. (Publishers are already singing the praises of this opt-out option as "author choice.")

Don't Wait Three Years to Correct the Flaws, as NIH Did.  As with the failed NIH policy, if Harvard does not upgrade its policy now, it will take three years to confirm empirically that the draft policy has failed.

Yet Harvard's mandate is so easy to fix pre-emptively now, with no loss at all in any of its intended effects, only the gain of a far greater likelihood of compliance and more OA. The ground has already been tested by other universities who have as already adopted a deposit mandate.

There Are Already 37 Successful Deposit Mandates.  It is not even clear why Harvard feels it needs to independently re-invent the OA-mandate wheel, when there are 15 successful university mandates and 22 successful funder mandates adopted already, not one of them needing to mandate copyright retention or to allow opt-out. All that is needed is for Harvard to adopt ID/OA, and all the other universities of the world would follow suit!

The notion that copyright retention is the solution to the research accessibility problem is far from new: It was prominently mooted a decade ago in Science magazine by 12 co-authors and has gotten precisely nowhere since then, despite being repeatedly revived, notionally, in university after university, by either the library or legal staff, year after year.

Copyright Retention is a Needless, Disabling Deterrent.  The reason requiring copyright retention is a nonstarter is that it contains a gratuitous, disabling deterrent: needlessly putting at risk the actual and perceived probability of being accepted by the author's journal of choice. Hence the need to allow opt-out, which in turn effectively reduces any mandate to a mere request.

An ID/OA mandate does not have that deterrent. Deposit mandates have already been demonstrated, repeatedly, to successfully approach 100% compliance within two years of adoption. ID/OA does not require copyright retention, nor an opt-out clause.

Deposit Mandate Will Eventually Lead to Copyright Retention Too; But OA First!  And, most ironically of all, ID/OA will almost certainly lead, eventually, to copyright retention too! But to do that, it must first reach 100% OA. And universal Deposit Mandates will ensure that.

Needlessly attempting instead to impose the stronger mandate first will not.

I hope Harvard will make the small parametric adjustments needed to maximize the likelihood that its historic mandate will succeed, and will be emulated by all other universities worldwide.

I close with a re-posting of the specific small but crucial changes in the wording of the mandate that are needed to prevent the copyright-retention requirement from compromising the deposit requirement.

Current Draft of Harvard OA Mandate.  First, here is the draft Harvard OA mandate as it now stands. [Passages that are flagged for modification are in brackets]:

Text of Motion on behalf of the Provost’s Committee on Scholarly Publishing:

The Faculty of Arts and Sciences of Harvard University is committed to disseminating the fruits of its research and scholarship as widely as possible. In keeping with that commitment, the Faculty adopts the following policy:

[COPYRIGHT RETENTION POLICY] Each Faculty member [grants] to the President and Fellows of Harvard College permission to make available his or her scholarly articles and to exercise the copyright in those articles. In legal terms, the permission granted by each Faculty member is a nonexclusive, irrevocable, paid-up, worldwide license to exercise any and all rights under copyright relating to each of his or her scholarly articles, in any medium, and to authorize others to do the same, provided that the articles are not sold for a profit.

[OPT-OUT CLAUSE] The [policy] will apply to all scholarly articles written while the person is a member of the Faculty except for any articles completed before the adoption of this policy and any articles for which the Faculty member entered into an incompatible licensing or assignment agreement before the adoption of this policy. The Dean or the Dean’s designate will waive application of the policy for a particular article upon written request by a Faculty member explaining the need.

[DEPOSIT MANDATE] To assist the University in [distributing the articles], each Faculty member [will provide] an electronic copy of the [final version] of the article at no charge to the appropriate representative of the Provost’s Office in an appropriate format (such as PDF) specified by the Provost’s Office. [The Provost’s Office may make the article available to the public in an open-access repository.]

The Office of the Dean will be responsible for interpreting this policy, resolving disputes concerning its interpretation and application, and recommending changes to the Faculty from time to time. The policy will be reviewed after three years and a report presented to the Faculty.

Proposed Revised Draft.   Now here are the small but crucial changes that will immunize the deposit requirement against any opt-outs from the copyright-retention requirement. Note the re-ordering of the clauses, and the addition of the underscored passages. (Other universities may also omit the two indented clauses preceded by asterisk ** if they wish):

Proposed revision:

The Faculty of Arts and Sciences of Harvard University is committed to disseminating the fruits of its research and scholarship as widely as possible. In keeping with that commitment, the Faculty adopts the following policy:

[DEPOSIT MANDATE] To assist the University in providing Open Access to all scholarly articles published by its Faculty members, each Faculty member is required to provide, immediately upon acceptance for publication, an electronic copy of the final peer-reviewed draft of each article at no charge to the appropriate representative of the Provost’s Office in an appropriate format (such as PDF) specified by the Provost’s Office. This can be done either by depositing it directly in Harvard's Institutional Repository or by emailing it to the Provost’s Office to be deposited on the author's behalf.

**[COPYRIGHT RETENTION POLICY] Each Faculty member is also encouraged to grant to the President and Fellows of Harvard College permission to make available his or her scholarly articles and to exercise the copyright in those articles. In legal terms, the permission granted by each Faculty member is a nonexclusive, irrevocable, paid-up, worldwide license to exercise any and all rights under copyright relating to each of his or her scholarly articles, in any medium, and to authorize others to do the same, provided that the articles are not sold for a profit.

**[COPYRIGHT-RETENTION POLICY OPT-OUT CLAUSE] The copyright retention and licence-granting policy will apply to all scholarly articles written while the person is a member of the Faculty except for any articles completed before the adoption of this policy and any articles for which the Faculty member entered into an incompatible licensing or assignment agreement before the adoption of this policy. The Dean or the Dean’s designate will waive application of the policy for a particular article upon written request by a Faculty member explaining the need.

The Office of the Dean will be responsible for interpreting this policy, resolving disputes concerning its interpretation and application, and recommending changes to the Faculty from time to time. The policy will be reviewed after three years and a report presented to the Faculty.