Thursday, July 26, 2012

OA advocate Stevan Harnad withdraws support for RCUK policy

When on July 16th Research Councils UK (RCUK) published its updated Policy on Access to Research Outputs the Open Access (OA) movement greeted the news with enthusiasm. This was hardly surprising: unlike the recommendations in the controversial Finch Report (published a month earlier), RCUK stressed that it continues to view both gold OA publishing and green OA self-archiving as equal partners in any OA policy.
Stevan Harnad

Gold and green are the two strategies outlined eight years ago when the OA movement was born, and are viewed as being essential components of any successful transition to OA.

By contrast, Finch concluded that the main vehicle should now be gold OA, either via pure open access journals or via hybrid journals, and that this should be funded by article processing charges (APCs).

At the same time, Finch argued, it was time to downgrade green OA, and reduce the role of institutional repositories to merely, "providing access to research data and to grey literature" and assisting in digital preservation. 

Set alongside the Finch proposals, OA advocates quickly concluded that RCUK’s policy was a godsend.

One of the first to applaud the new policy was long-standing OA advocate, and self-styled archivangelist, Stevan Harnad. The minute the report was published a relieved Harnad began flooding mailing lists with messages congratulating RCUK on coming up with a policy that not only defied Finch, but was stronger than its current OA policy. 

But as Harnad set about talking up the policy, and seeking to win over sceptics and doubters, he himself began to have doubts. And eventually he was driven to the conclusion that he had no option but to withdraw his support for the RCUK policy — which he now characterises as “autistic”, and a “foolish, wasteful and counterproductive step backwards”.

How has what at first sight seemed so desirable rapidly become something terrible? Curious to find out, I contacted Harnad. Below I publish the email interview that emerged from our conversation.

The interview begins …

RP:  When RCUK’s new OA policy was published on 16th July you wrote that it had “shown sense and independence” in ignoring the recommendations in the Finch Report. Instead, you said, RCUK had “re-confirmed their policy of mandatory author self-archiving in Green OA repositories.” The same day you also said, “The publishing industry’s lobbying efforts have failed with RCUK and the EC; instead, the global research community’s self-help efforts to protect the interests of publicly funded research have triumphed”

I believe this is not your view now. What changed?

SH: No, I regret to have to say that this is no longer my view. As soon as I saw the new RCUK policy I spotted a potential bug in it, but I was so relieved that it still allowed Green OA self-archiving as a means of compliance, that I just hoped the bug would not be picked up on (given that there has been so much blindness to detail and subtlety already in OA, both pro and con).

But now I see that it has been picked up on — so far only by the author community and some neutral observers but, I am sure, it will soon be picked up on and acted upon, by the publishing community, and the consequences will be terrible.

RP:  What is the problem?

SH: The bug is this: RCUK, in a well-meaning effort to pressure journals to either offer a Gold option or a Green option (within an allowable embargo period), have used the oft-mooted strategy of a “pre-existing contract” with their funder that would bind RCUK fundees when they choose a journal to publish in and negotiate the copyright agreement with the journal.

So the RCUK policy became: RCUK fundees may only publish in journals that either offer Gold OA or Green OA (within an allowable embargo period).

Now think for a moment: If you were a journal publisher — including, and indeed especially one of the publishers of the 60% of journals that already endorse immediate, un-embargoed Green OA today — what would you do, when faced with a policy like that?

RP: What do you predict?

SH: The answer is obvious: You would offer to “allow” your authors to pay you for hybrid Gold OA (while continuing to collect your usual subscription revenues) and, for good measure, you would ratchet up the Green OA embargo length (up to the date your grand-children finished their university education!) to make sure your authors pay you for hybrid Gold rather than picking the cost-free option that you fear might eventually pose a risk to your subscription revenues!

(The endless embargo would almost not even be necessary, since it looks as if the RCUK policy even dictates that if the journal offers both Gold and Green, the fundee must pick Gold!)


RP: Your main concern, I assume, relates to Section 4 of the RCUK policy, which states

The Research Councils will continue to support a mixed approach to Open Access. The Research Councils will recognise a journal as being compliant with their policy on Open Access if:

1. The journal provides via its own website immediate and unrestricted access to the publisher’s final version of the paper (the Version of Record), and allows immediate deposit of the Version of Record in other repositories without restriction on re-use. This may involve payment of an ‘Article Processing Charge’ (APC) to the publisher. The CC-BY license should be used in this case.


2. Where a publisher does not offer option 1 above, the journal must allow deposit of Accepted Manuscripts that include all changes resulting from peer review (but not necessarily incorporating the publisher’s formatting) in other repositories, without restrictions on non-commercial re-use and within a defined period. In this option no ‘Article Processing Charge’ will be payable to the publisher. Research Councils will accept a delay of no more than six months between on-line publication and a research paper becoming Open Access, except in the case of research papers arising from research funded by the AHRC and the ESRC where the maximum embargo period is 12 months.

What in your view is the specific problem with this clause?

SH: The clause imagines that it will drive all publishers either to become “pure” Gold OA, CC-BY publishers or at least publishers that endorse Green (within an allowable embargo) otherwise they will face losing their RCUK authors.

But in reality it will simply drive subscription publishers to offer hybrid Gold and to jack their Green embargos up to unallowable lengths to make sure authors pick the hybrid Gold option (although the policy itself sounds as if RCUK authors must already pick the hybrid Gold option if it is offered, come what may).

And the consequence, besides the disastrous effects on publisher policy, will be that the RCUK policy will provoke an author revolt — not just because of the needless squandering of money diverted from scarce research funds to pay publishers extra for hybrid Gold OA, but because of the heavy-handed tampering with authors' choice of journals.

It’s one thing to mandate that the author must do a few extra keystrokes to provide Green OA, quite another to tell authors: “you may publish in this journal but not that.”

This extremely ill-thought-through new policy will sow confusion, resentment, and, ultimately, non-compliance among UK authors. More years of OA — and the opportunity to consolidate UK leadership in reaching it — squandered because of short-sightedness, gold fever, rights rapture, and, frankly, a goodly dose of stubbornness.

RP:  Can you expand on what you mean when you say that the RCUK policy will tamper with researchers' choice of journal?

SH: One could not possible tamper more directly! Researchers are accustomed to choosing which journal is the most appropriate for their paper, submitting there, and if successful, publishing there.

RCUK says you may no longer publish in any journal you choose. You may only publish in a journal that either offers a Gold option or endorses Green OA self-archiving within the allowable embargo interval (12 months for AHRC and ESRC, 6 months for the other RCs).

Moreover, if the journal offers Gold, you must choose gold (and pay for it).

It is not clear how many paid-Gold articles RCUK plans to subsidize for the author, but if it is not all of them, then that's yet another constraint on the author's choice.

Hybrid OA

RP:  So you believe that RCUK’s willingness to accept hybrid OA is what is problematic?

SH: Not only is it problematic, it is the heart of the problem: It is all too easy for a publisher to provide a pricey hybrid OA option as a sop for those minded to “mandate” OA and foolish enough to be ready to pay extra for it — while subscriptions are still paying in full for publication, and there is no alternative to subscription access.

(See the section on Hybrid OA and the Cheshire Cat’s grin below for further explanation of this point)

It’s precisely that alternative to subscription access that globally mandated Green OA would provide. And it’s precisely the availability of that alternative that would eventually induce the transition to universal Gold OA that RCUK desires — but at a far lower price, and having released the subscription cancelation funds to pay for it (instead of scarce research funds).

But that’s the difference between post-Green Gold OA and pre-emptive Fool’s Gold OA.

RP:  Am I right in thinking that the RCUK policy does not actually mention hybrid OA, but that you assume hybrid will be acceptable because of the wording, which simply says,  “This may involve payment of an ‘Article Processing Charge’ (APC) to the publisher”. The point is that this can encompass both gold and hybrid?

SH: You bet they don't mention it! They talk as if Gold meant pure Gold (i.e., a Gold-only journal). But it's clearly on the cards, since RCUK implies that if a journal offers either Gold OA *or* Green OA, the RCUK fundee must choose the Gold. Only a hybrid Gold journal can offer either/or.

My guess is that RCUK both remembered and didn't remember that hybrid Gold was at issue.

RP: However, it is the case, is it not, that currently few if any hybrid options offer CC-BY, which the policy requires for gold OA. If so, I guess you are assuming that publishers will start to offer CC-BY for their hybrid options so that they can take best advantage of the policy?

SH: Offering CC-BY with the hybrid Gold option would be no skin off the publisher's nose: well worth all the extra dosh. (How many users really need it, and whether all authors would want to provide it is another matter).

RP:  If the implications of the policy will be as you predict, is David Arnold therefore right to say, as he has on the Economist web site, “Since the vast majority of journals now offer a gold route, the green option is essentially redundant”?

SH: It’s certainly not yet true that the vast majority of journals already offer Gold, but it is alas very likely that they soon will.

The only hope now — for OA and the rest of the world — is to try to contain this, as a piece of UK folly alone. This is especially painful for me, because the UK was the first and undisputed leader in OA until now.

The UK only generates 6% of worldwide research. If it chooses to squander its scarce research funds on paying extra for hybrid Gold OA instead of mandating cost-free Green OA perhaps the global economic crunch will discourage the rest of the world from following this folly.

RP:  Based on what you say, how then would you characterise the difference between the Finch recommendations and the RCUK policy?

SH: Finch says Green is inadequate and should be downgraded to data-archiving and digital preservation; all UK journal articles should be paid Gold.

RCUK says you can still provide cost-free Green instead, if your journal does not offer Gold. (But it is obvious that all it needs is the stroke of a pen for a subscription publisher, and then they too offer hybrid Gold.)

Step backwards

RP:  RCUK has had an OA policy since 2006. I guess you would say that the new policy is a step backwards?

SH: It is a huge, foolish, wasteful and counterproductive step backwards.

In the naive hope of forcing all journals to convert to affordable pure-Gold OA with CC-BY, with Green OA (within allowable embargo limits) as a back-up, RCUK has instead given publishers an irresistible incentive to offer hybrid Gold, along with unallowable Green embargoes to ensure that authors take up the pricey option.

The result will be confusion, wasted resources, and understandable and justified resentment and non-compliance from RCUK fundees. All I can do is hope that exposing the UK OA fiasco will prevent it from being emulated in the rest of the world.

My slimmer and fonder hope, of course, is that RCUK will have the sense and integrity to recognize its mistake, once the unintended negative consequences are pointed out, and will promptly correct it. It can still be corrected completely with 2 simple patches.

RP:  What patches do you have in mind?

SH: RCUK should:

(1)           Drop the implication that if a journal offers Green and Gold, RCUK fundees must pick Gold


(2)           Downgrade to a request the requirement that the Green option must be within the allowable embargo interval.

(The deposit of the refereed final draft would still have to be done immediately upon publication, but the repository’s “email-eprint-request” Button could be used to tide over user needs by providing “Almost-OA” during the embargo.)

RP: Should they not consider prohibiting the use of hybrid OA?

SH: There is no way to resurrect the current RCUK policy in such a way as to rule out hybrid Gold: It would have to be re-conceived and re-written completely. If that were done, all of the fatal bugs of the present draft would be gone:

“(1)You must provide at least gratis OA (within the allowable embargo). This can be done either by paying for pure Gold OA (not hybrid) — but then the OA must be libre and immediate and unembargoed (and the paper should be immediately deposited in the fundee's repository anyway). Or (2) you can provide Gratis Green OA to the refereed final draft within the allowable embargo (but the deposit must be made OA immediately upon acceptance for publication).”

That would be a fine policy, especially if beefed up with a link to submission to HEFCE [Higher Education Funding Council for England] for REF.

RP:  How and why do you think RCUK came up with the form of wording it did: did those drafting the policy fail to see the implications of the wording, or was it (as you have suggested of the Finch Report) worded in this way in order to placate scholarly journal publishers?

SH: I am now pretty sure that the RCUK policy, at least, was not just the result of successful publisher lobbying: Premature gold fever and “rights rapture” are just as responsible.

Rights rapture

RP:  Can you explain this in more detail?

SH: As you know, one of the most frequent and flagrant errors about OA is to assume that OA is synonymous with OA publishing (i.e., Gold OA). The other is to assume that OA is not “fully” OA unless it is “Libre OA” (i.e., free online access, also called “Gratis OA,” plus various additional re-use rights, such as the right to republish and re-sell articles, online as well as in print, and the right to create derivative works, including re-mixes and mash-ups, with such licenses as CC-BY).

Free online access (Gratis OA) means that research is accessible to all its potential users, not just those whose institutions can afford to subscribe to the journal in which it was published, so that the research can be read, used, applied, and built upon in further research and applications, to the full benefit of research progress and the tax-paying public that funded it. Gratis OA includes the user’s capability of screen-viewing, down-loading, linking, printing-off, storing, and data-mining the text; in addition, harvesters like Google Scholar invert and index it for navigability and search.

A conversion of journals from subscriptions to affordable Gold OA as well as the licensing of some further re-use rights (if/when users need them and authors want to provide them) are both desirable, but not even faintly as important or urgent as free online access itself, which is the revolutionary new possibility that the advent of the online medium had opened up for researchers, launching the OA movement. But neither Libre OA nor Gold OA is worth paying extra for today, nor worth further delaying Green Gratis OA for.

The romance of OA also inspired many important new possibilities, beyond free online access, including creative new forms of online use and re-use as well as a possible solution to the journal affordability problem.

The Gold OA model (which I, surely not the first, innocently mooted in 1994 as the obvious way to cover publishing costs once research was all OA) particularly captured the imagination and allegiance of OA advocates, including beleaguered serials librarians and authors aggrieved with rapacious publishers.

But what the Gold OA and CC-BY enthusiasts forgot — as they went on and on about the potential for gilded, mashed-up splendours over and above free online access was that the years kept going by and we still did not even have free online access (Gratis OA).

And the reason is very simple: Green Gratis OA is fully within reach today, because with institutional subscriptions already covering the costs of publication (in full, and generously!) it is entirely within the research community’s hands to provide Green Gratis OA, today, cost-free: All we have to do is mandate it. Whereas Gold Libre OA is not within reach today, because it is in the publishing community’s hands to provide, and costs extra money — money that is currently locked into journal subscriptions.

So instead of grasping what is already entirely within reach, cost-free, by mandating Green OA, RCUK wanted even more: It wanted all publishers to convert to Gold OA, which would give immediate Libre OA, free of re-use restrictions and embargoes. But not free of extra cost; nor of restrictions on author journal choice.

So RCUK thought they could use the widely mooted rationale that if authors are already bound by a pre-existing contractual obligation to their funder or institution at the time they negotiate the copyright agreement with their publishers, the publishers have to respect the prior agreement, or lose their authors. So publishers have to offer Gold OA, or at least Green OA (within the allowable embargo limit).

RP:  Why was this not obvious?

SH: The three possibilities RCUK did not reckon on in advance were:

(1)           That publishers could simply offer RCUK authors pricey hybrid Gold (rather than converting to “pure” Gold publishing, as RCUK had anticipated) and jack up embargoes to unallowable limits to make sure RCUK authors have to pick the hybrid Gold option

(2)           That authors would resent the diversion and spending of scarce research money to pay publishers for Gold OA, when Green OA can be provided for free, and

(3)           That authors would resent the restrictions imposed on their choice of journal, when Green OA (or at least 60% immediate-OA and 40% “Almost OA”) can be provided without needlessly renouncing their journals of choice — or being forced to renounce publishing further articles at all, once the allotted Gold OA funds have been exhausted.

In short, RCUK did not take into account the confusion, resentment and resistance their heavy-handed efforts to force publishers into converting to Libre Gold OA would elicit from their fundees.


RP:  Do you have reservations about any other aspects of RCUK’s policy?

SH: Yes. It is remarkably autistic (or solipsistic), imagining that a local RCUK OA “solution” could be implemented without taking into account the question of compatibility and scalability for OA in the rest of the world.

Unlike the 2004 UK Select Committee’s timely, realistic, and eminently scalable recommendation, which was that UK institutions and funders should mandate Green OA and merely experiment with some funding for Gold OA, RCUK are acting as if the experiment has been conducted, proved successful, and now just needs to be implemented — by throwing money at Gold and forcing publishers to offer it and authors to pick and pay for it.

RCUK should now admit it did not think it through fully, and promptly apply the two patches sketched above. Then the UK will again have a realistic, scalable solution for the rest of the world to follow.

If not, then UK OA is doomed to years more of confusion and non-compliance.

RP:  If you are right, then what puzzles me is that OA advocates also failed to see the implications. Not only did you initially applaud the RCUK policy, but so too apparently did SPARC Europe (which “warmly welcomed” the new policy, describing it as a stronger policy). Likewise, Peter Suber wrote on 16th July, “Instead of favouring gold over green, and even disparaging green, the new RCUK policy favours green over gold”. You are saying that in reality the reverse is actually the case I believe.

SH: I can’t speak for others, but in my case it was a combination of:

(1)   shock at the Finch recommendation, which was to phase out Green altogether, and just fund Gold, exactly as publishers had been urging for the past several years (once they had realized that the clamour for OA was not going to go away, and had to be placated somehow)

(2)   relief that RCUK quickly announced that it would continue to “allow” Green as an option, and, frankly,

(3)   some conscious wishful thinking (if not self-delusion), and the desire to put a more positive spin on the new RCUK policy than its own wording quite warranted, especially to limit the damage that mindless emulation of the Finch recommendations could do to the global OA movement.

I am ashamed to say that I even told Richard van Noorden, a journalist for Nature, that I hoped he would not mention the awful contingency that the RCUK might inspire (hybrid Gold plus hyper-embargoes), in the hope that publishers would not notice it.

In a word, a combination of stupefaction and stupidity on my part.

RP:  Do you anticipate that the RCUK policy will be changed, or does the research community now have to live with it?

SH: I know for sure that the policy can easily be fixed so that it works and provides a scalable model for the rest of the world to emulate. I was shocked by the Finch recommendations, and am now equally appalled by the RCUK policy.

But I have not forgotten the immense surprise and joy I felt when you, Richard, contacted me in Barcelona in 2004 to ask me for my reaction to the UK Select Committee’s recommendation.

That was the first I heard of that outcome which, against all odds, resisted the urgings of both the anti-OA publishing lobby and the nascent pro-Gold-OA publishing lobby (as reported in your interview of BMC’s Vitek Tracz), and instead recommended mandating Green OA.

Good sense and good judgment can again prevail in the UK, and the RCUK can make the two patches that will not only salvage its policy but accelerate its global emulation. There is no shame in admitting an error, and only kudos for fixing it forthwith.

Green OA is clear

RP:  The other major source of research funding in the UK comes via HEFCE. HEFCE has announced that it plans to introduce an OA policy, and it is expected that this will have a green component to it. Speaking to me recently Martin Hall, Vice-Chancellor at the University of Salford, said of the proposed HEFCE policy:

It will be important to see exactly what HEFCE means by “green”. For the purposes of the 2020  Research Excellence Framework (or its equivalent), HEFCE could merely require that the author’s last version is made available via a repository (a condition that can, of course, be met at present).

Alternatively, they may require open access to a version of record, which will be a big push towards full and upfront APCs. HEFCE (in contrast with the Research Councils) is also going to have to work out what to require for research outputs that can (and must) be submisable, but which are for research not supported with public funds. The details will be important here.

What is your view on this? What can HEFCE do to ameliorate the weaknesses you see in the RCUK policy?

SH: As I will be shortly stating on your blog, I must say with profound regret that Martin Hall, who I had thought was a comrade-in-arms in the quest for OA, has turned out not to have a clear understanding of OA at all.

What is meant by “green” is very clear: The deposit of the author’s refereed final draft in an OA repository immediately upon publication (unembargoed). That is Green OA.

A Green OA mandate, however, has to live within a constraint, namely, the fact that 40% of publishers still demand an embargo of various lengths before the deposit is made OA. The compromise is to mandate immediate deposit and urge (but not require) immediate OA. The rest will take care of itself naturally, of its own accord (with the help of the Button, the increasingly palpable benefits of OA, and human nature).

HEFCE should without a shadow of a doubt require that the author’s refereed final draft be deposited in the author’s institutional repository.

This will reinforce institutional and funder Green OA mandate adoption as well as facilitating both author compliance and the institutional monitoring and assurance of compliance along the lines that Bernard Rentier, the rector of U. Li├Ęge has successfully implemented within the university and with the Belgian funding council: designating repository deposit as the mechanism for submitting papers for research assessment.

If HEFCE insisted instead on the deposit of the publisher’s version of record (with its far more restrictive publisher constraints) that would be another gratuitous and costly mistake, gratuitously insisting on more than what is necessary — when the author’s draft is enough, and already able to do so much more good in reinforcing Green OA mandates and hence OA — at the cost of a lost opportunity to accelerate Green OA growth.


RP:  You have always maintained that OA is a no-brainer,and have characterised it as "Raincoat Science". I wonder if the misapprehensions over the RCUK policy do not cast doubt on your assertion. The devil, it seems, is in the details. 

SH: But the details are trivial and obvious. Policy makers just have to listen: not to ideology blinkered by gold dust and rights rapture but to simple, practical, cause-effect evidence and reasons.

RP:  But if the end result of the RCUK policy is to increase the number of research papers that are freely available on the Internet does it matter so much?

SH: Let me make this perfectly clear: My own goal is and always has been OA — by which I mean, first and foremost, online access to peer-reviewed research articles, free for all would-be users, not just those at institutions that can afford subscription access to the journal in which it was published (“Gratis OA”).

I do not care in the least how that Gratis OA is provided — green, gold, or day-glo — as long as it is provided, today.

But even just this Gratis OA (Libre OA is virtually zero) is only being provided today to the tune of about 20% worldwide, and this has been true for years now. Research usage, impact and progress are consequently being needlessly lost for years.

Yes, every increase in OA is welcome, regardless of its colour — but not if it comes at the expense of preventing or delaying even greater increase in OA.

And that’s precisely what the UK’s latest turn toward a fool’s gold rush is doing: providing a modest (though needlessly and immoderately expensive) “increase [in] the number of research papers that are freely available on the Internet” at the considerable expense of a far greater increase that could be provided by adopting a sensible, scalable, cost-free Green OA mandate for UK research, suitable for emulation by the rest of the world.

If the UK first does the latter — clearly and unambiguously mandates Green OA for all UK research output — then it is welcome to throw all the cash it has to spare on also subsidizing Gold OA if it so wishes.

But not instead.

RP:  Thanks for sharing your thoughts with me.



Harnad explains below why he believes publishers will seek to cash in on the new RCUK policy by expanding hybrid OA:

Suppose you're a subscription journal. Hybrid gold means you just keep selling subscriptions and — on top of that — you can charge (whatever you like) extra for selling single-article hybrid gold.

How much do you charge? Well, if you publish 100 articles per year and your total annual revenue is £XXX, you charge 1% of £XXX for hybrid Gold OA per article.

Once you've got that (plus your unaltered subscription revenue of £XXX) you've earned £XXX + 1% for that year.

Good business.

And if the UK publishes 6% of the world's articles yearly, then on average 6% of the articles in any journal will be paid-up hybrid Gold OA, thanks to Finch and RCUK. That means worldwide publisher revenue — let's say it's £XXX per year — will increase from £XXX per year to:

£XXX + 6% per year

Not bad.

Publishers are not too dense to do the above arithmetic. They've already done it. That is what hybrid Gold is predicated upon. And that is why publishers are so pleased with Finch/RCUK: “The world purports to want OA. Fine. We’re ready to sell it to them — on top of what we're selling them already.”

In the UK, Finch and RCUK have obligingly eliminated hybrid Gold OA's only real competition (Green OA) — Finch by ignoring it completely, and RCUK by forcing fundees to pay for Gold rather than provide Green whenever the publisher has the sense to offer Gold.

Of course publishers will say (and sometimes even mean it ) that they are not really trying to inflate their income even further. As the uptake of hybrid Gold increases, they will proportionately lower the cost of subscriptions — until subscriptions are gone and all that's left, like the Cheshire Cat’s grin, is Gold OA revenue (now no longer hybrid but "pure") — and at the same bloated levels as today's subscriptions.

So what? The goal was always OA, not Green OA or Gold OA. Who cares if all that money is being wasted?

I don't.

I care about all the time (and with it all that OA and research progress) that has been wasted, and that will continue to be wasted, as the thrall of Gold Fever and Rights Rapture keep the research community from mandating the cost-free Green OA that would bring them 100% OA globally in next to no time, and leave them instead chasing along the CC-Byways after gold dust year upon year upon year, at unaffordable, unnecessary and unscalable cost.

Image: Judith Economos; license: CC-BY.


Mike Taylor said...

Oh, Stevan.

Richard Poynder said...

A conversation about this interview has been taking place on Google+ here.