Will open access undermine peer review?
SPARC Open Access Newsletter, issue #113
September 2, 2007
by Peter Suber

As soon as governments started contemplating policies to ensure open access to publicly-funded research, publisher trade associations and lobbying coalitions objected that the policies would undermine peer review.  Here are some recent examples from the US and EU:

From the Association of American Publishers (AAP) in May 2006, opposing the Federal Research Public Access Act (FRPAA):

If enacted, S.2695 could well have the unintended consequence of compromising or destroying the independent system of peer review that ensures the integrity of the very research the U.S. Government is trying to support and disseminate.

https://mx2.arl.org/Lists/SPARC-OAForum/Message/3006.html
http://www.earlham.edu/~peters/fos/2006_05_07_fosblogarchive.html#114726726169346460

From the DC Principles Coalition in February 2007, opposing FRPAA:

Subscriptions to journals with a high percentage of federally funded research would decline rapidly. Subscription revenues support the quality control system known as peer review....

http://www.dcprinciples.org/press/2.htm
http://www.earlham.edu/~peters/fos/2007_02_18_fosblogarchive.html#117202895951705271

From the Brussels Declaration in February 2007, organized by the Association of Learned and Professional Society Publishers (ALPSP) and International Association of Scientific, Technical & Medical Publishers (STM), opposing a proposed OA mandate for the EU:

Open deposit of accepted manuscripts risks destabilising subscription revenues and undermining peer review.

http://www.alpsp.org/ForceDownload.asp?id=304
http://www.earlham.edu/~peters/fos/2007_02_11_fosblogarchive.html#117138731546079566

From the AAP/PSP's new lobbying organization, PRISM, in August 2007, opposing government OA mandates:

[OA policies] would jeopardize the financial viability of the journals that conduct peer review, placing the entire scholarly communication process at risk.

https://mx2.arl.org/Lists/SPARC-OAForum/Message/3934.html
http://www.earlham.edu/~peters/fos/2007_08_19_fosblogarchive.html#365179758119288416

We don't know how many individual publishers share the objection.  Most have been silent, neither publicly endorsing nor publicly criticizing this claim by their trade associations and lobbyists.  A recent exception is Rockefeller University Press, which publicly dissociated itself from the AAP/PSP's PRISM campaign, including the claimed threat to peer review.
https://mx2.arl.org/Lists/SPARC-OAForum/Message/3941.html
http://www.earlham.edu/~peters/fos/2007_08_26_fosblogarchive.html#8048311513999344583

For convenience, I'll refer to those who raise this objection as the publisher trade associations or lobbyists.  I can't use the shorter term "publishers" because so few have taken a position on it.  Publishers are not monolithic, even if their lobbyists want to appear to speak for all of them.  While some publishers do support the objection, some reject it, and a growing number of publishers embrace both OA and peer review. 

Whenever I've seen publisher lobbyists raise the objection in public, I've blogged it along with a short, blog-length answer.  But for once I want to answer the objection at length.  I have three reasons.  First, the brevity of my blog responses hasn't allowed me to show the answer at full strength.  Second, the launch of PRISM, and its focus on this objection, suggest that publisher lobbyists are about to escalate their use of it.  Third, I'm tired of keying new blog responses to this old canard and don't want to have to do it every time lobbyists repeat it.  Writing new rebuttals even to old and unargued objections makes us vulnerable to anyone who wants to yank our chain.  It's much better to write a thorough rebuttal once and link to it as needed.  It strengthens the response, saves time, and cuts the chain.

(1) First note that the publisher associations who raise the objection usually have no problem with *low-volume* OA archiving.  They don't think it undermines subscriptions, revenue, or peer review, and most publishers have already adopted policies to permit it.  The objection does not target OA archiving as such, or spontaneous levels of author self-archiving, but only policies that would significantly increase the volume of OA archiving. 

(2) The objection is not that funder OA policies ask researchers to bypass peer review.  On the contrary, funder OA policies uniformly support peer review and encourage publication in peer-reviewed journals.  This fact is well-known among the friends and foes of OA, but I mention it for completeness in case newcomers to the debate did not know it. 

(3) Just as the funder OA policies do not ask authors by bypass peer review, they do not ask referees, editors, journals, or publishers to use any particular form of peer review, to lower their standard of peer review, or to change their methods of peer review.  The funder OA policies leave the methods and standards of peer review entirely up to the journals where authors/grantees choose to publish their work. 

Again, these facts are well-known to those familiar with the OA policies.  I mention them not only for newcomers, but because some statements by the publishing lobby have muddied the waters.  For example, just last week Rachel Deahl wrote in Publishers Weekly,

PRISM members are concerned that if government becomes involved in the publication of scientific and scholarly work, changing the standard peer review process that has long been in place, the work could lose its integrity. As Dr. Brian Crawford, chairman of the [Association of American Publishers'] Professional & Scholarly Publishing Division explained, changing the peer review process could ultimately open the gates for "agenda-driven research and bad science."

http://www.publishersweekly.com/article/CA6471211.html
http://www.earlham.edu/~peters/fos/2007_08_19_fosblogarchive.html#5309199826835837381

If Crawford is really talking about government policies to change the methods of peer review, then he's fighting without an antagonist.  There are no such policies and no policy-makers or OA advocates to support them.  If he's talking about policies to encourage or require OA archiving, then he's changing the subject and leaving a false impression, much as if an oil company opposed steps to reduce greenhouse gases on the ground that reducing automobile safety would ultimately open the gates for injury and death.

Since the NIH policy was proposed in 2004, publisher lobbyists have objected that it would "nationalize science", by which they seemed to mean that governments would take over the job of doing or managing peer review --and then corrupt it under political pressure from elected officials.  Crawford's statement echoes this older concern.  But the concern is easily answered by reading the policies and recognizing that nothing in them allows governments to take over or modify either of these jobs.  The policies regulate grantees, not publishers.  They concern OA repositories, not OA journals.  They focus on archiving work peer-reviewed and published elsewhere, not changing where or how work is peer-reviewed and published.  They leave every aspect of peer review up to independent (i.e. non-government) journals.  If peer review at independent journals is corrupted or politicized, it will be entirely the fault of the journals themselves.  The real-life examples of ideological control over government scientists by the Bush administration are not examples of OA policies at work and would even be prevented by effective OA policies. 

The only plausible form of the publishing lobby's objection is about defunding peer review, not corrupting, politicizing, or nationalizing peer review, and from here I'll focus on that form of the objection. 

(4) Publisher lobbyists who object that high-volume OA archiving will undermine peer review are never specific in explaining why or how.  But they seem to be thinking about a three-link chain of causation:  high-volume OA archiving will cause libraries to cancel journal subscriptions, which will in turn cause journals to lose revenue, which will in turn cause the journals to fold up and therefore to cease providing peer review.  At least that's the form of the objection to which I'll respond here.

(5) The first problem with the imagined chain of causation lies in the first link, the assumption that OA will trigger cancellations.  Maybe it will; maybe it won't. 

The evidence to date is that it won't.  Physics is the field with the highest level and longest history of OA archiving.  Physicists have been self-archiving in arXiv for 16 years, far longer than in any other field.  In some subfields, like particle physics, the OA archiving rate approaches 100%, far higher than in any other field.  If high-volume OA archiving caused journal cancellations, we should expect to see it first in physics.  But we don't see it at all.  Two leading publishers of physics journals, the American Physical Society (APS) and Institute of Physics (IOP), have publicly acknowledged that they've seen no cancellations attributable to OA archiving.  In fact, the APS and IOP have not only made peace with arXiv, but they now accept submissions from it and even host their own mirrors of it. 

arXiv
http://arxiv.org/

American Physical Society (APS)
http://www.aps.org/

Institute of Physics (IOP)
http://www.iop.org/

APS mirror of arXiv (launched December 1999)
http://aps.arxiv.org/

IOP mirror of arXiv (launched September 2006)
http://eprintweb.org/

Alma Swan's interview with the APS in IOP in which "both societies said they could not identify any losses of subscriptions" due to OA archiving.
http://eprints.ecs.soton.ac.uk/11006/

Other fields may have a different experience, however.  For fields that turn out to be like physics, toll-access (TA) journals will coexist with high-volume OA archiving and perhaps even become symbiotic with it.  For fields that turn out to differ from physics, high-volume OA archiving might threaten some TA journals. 

It would definitely help to understand why the experience in physics has gone as it has and how far it might predict the experience in other fields.  But, so far, it's fair to say that we don't know the variables and that publisher lobbyists are not among those showing a serious interest in them.  When publisher trade associations argue that high-volume archiving will undermine subscriptions, they don't acknowledge the countervailing evidence from physics, let alone rebut it or qualify their conclusions in light of it.  It would be more honest and helpful if they would acknowledge the evidence from physics and then argue, as well as they could, that they have identified the variables and can show that their members publish in fields that are not like physics in the relevant respects.  I'm still waiting to see such an argument.

We don't know how many fields will turn out to be like physics, but the real-world study that will give us some answers is already under way.  Five of the seven Research Councils UK have adopted OA mandates, and most of them took effect in October 2006.  Together they go well beyond physics to astronomy, biology, medicine, environmental science, economics, and the social sciences.  Answers won't come quickly though:  authors must receive their grants, do their research, write it up, and get it published; then we must wait for the OA embargoes to toll, for the volume of OA literature to grow, and for the new, larger volume of OA literature to have its effect, whatever it may be, on library renewal decisions.  But at least the wheels are already turning and on a large scale:  there are also multi-disciplinary OA mandates in Australia, Austria, Belgium, Canada, France, Germany, Scotland, and Switzerland. 

(Note to the EU, US, and other jurisdictions considering an OA mandate:  there are at least nine national-level, multi-disciplinary OA mandates and even more softer OA policies that encourage OA without requiring it.  These constitute a large, ongoing natural experiment.  By all means look at the evidence, but don't fall for the argument that we must delay the adoption of new OA policies in order to launch another, later, smaller study of the effect of OA archiving on journal subscriptions.)

Although the publishing lobby ignores the experience in physics, that experience isn't unambiguously favorable for either friends or foes of OA.  It cuts both ways for both camps.  The good news for OA, and bad news for publishers, is that the publishing lobby's favorite argument is very weak.  So far, fears that high-volume OA archiving will kill journal subscriptions are groundless outside physics and contradicted inside physics --and physics is by far the strongest test case.  The bad news for OA, and good news for publishers, is that even the very high levels of OA archiving inside physics does nothing to justify libraries in cancelling TA journal subscriptions.  So far, the money needed to support peer-reviewed OA journals on a large scale is still tied up in TA journal subscriptions.

Whenever I point out the evidence from physics, I also argue that even if OA archiving does threaten TA journals in other fields, OA mandates are still justified.  I'll make some of those arguments below.  But most of them are beyond the scope of this note, which is limited to the effect of OA mandates on subscriptions and peer review.

(6) Another reason to think that OA mandates will not kill subscriptions is that they leave standing at least four library incentives to subscribe.

First, all OA mandates include an embargo period to protect publishers.  For example, the OA mandates at the Research Councils UK require OA within six months after publication.  The bill now before Congress that would strengthen the NIH policy from a request to a requirement would allow an embargo period of 12 months.  Libraries that want to provide immediate access will still have an incentive to subscribe.

Second, OA mandates only apply to the final version of the author's peer-reviewed manuscript, not to the published version.  If the journal provides copy editing after peer review, then the policies do not apply to the copy-edited version, let alone to the published PDF.  Libraries that want to provide access to published edition, or the published version of the text, will still have an incentive to subscribe.

Note what happens when we put the two previous points together.  The OA mandate, and its associated embargo period, only apply to the author's peer-reviewed manuscript.  Publishers retain life-of-copyright (virtually permanent) exclusivity on the published edition.  Hence, the funder-mandated OA copies of the author's peer-reviewed manuscript won't compete with subscriptions for 6-12 months, and OA copies of the more desirable published edition need never compete with subscriptions.  Even if OA archiving does have harmful effects on subscriptions outside physics, publishers have longer and better protection from these effects than their lobbyists ever acknowledge. 

Third, OA mandates only apply to research articles, not to the many other kinds of content published in scholarly journals, such as letters, editorials, review articles, book reviews, announcements, news, conference information, and so on.  Libraries that want to provide access to these other contents will still have an incentive to subscribe.

Fourth, funder OA mandates only apply to articles arising from research funded by the mandating agency.  Very few journals publish nothing but articles from a single funder or even from a set of funders all of whom have OA mandates.  Libraries that want to provide access to all the research articles in a journal, regardless of the source of funding, will still have an incentive to subscribe.  (This incentive will weaken as more and more funders adopt OA mandates; but we're very far from universal funder mandates; unfunded research, which predominates in many fields, will still fall outside this category; and the other incentives above will still stand.)

Here's how the Association of College and Research Libraries put the point in a November 2004 open letter on the NIH policy: "We wish to emphasize, above all, that academic libraries will not cancel journal subscriptions as a result of this plan....Even if libraries wished to consider the availability of NIH-funded articles when making journal cancellation decisions, they would have no reasonable way of determining what articles in specific journals would become openly accessible after the embargo period."
http://www.ala.org/ala/acrl/acrlissues/scholarlycomm/nihsupport.cfm

These four reasons do not guarantee that high-volume OA archiving won't cause some cancellations.  They are merely reasons or incentives for libraries to renew their subscriptions, and libraries will evaluate them in light of the reasons or incentives to cancel, such as eventual OA to a subset of the articles, rising prices, oppressive licensing terms, and competition for limited budget dollars from journals with better terms, higher impact, or greater local usage.  They aren't a decisive "no" to the publishing lobby's slippery slope, just cautions against its glib and oversimple "yes".

(7) Some studies bear on the question whether the growing volume of OA archiving will trigger journal cancellations.

Publisher associations like to cite the Publishing Research Consortium (PRC) study by Chris Beckett and Simon Inger, Self-Archiving and Journal Subscriptions: Co-existence or Competition? An international Survey of Librariansí Preferences, October 26, 2006.
http://www.publishingresearch.net/documents/Self-archiving_report.pdf
http://www.earlham.edu/~peters/fos/2006_11_05_fosblogarchive.html#116311103707504219

The PRC asked 400 librarians about the relative weight of different factors in their decisions to cancel subscriptions.  Other things being equal, the librarians preferred free content to priced content, and short embargoes to longer ones.  Publisher lobbyists interpret this to mean that the rise of OA archiving will cause a decline in subscriptions.  The chief flaw with the study, however, is its artificiality.  The authors' attempt to eliminate bias also eliminated realism.  For example, they did not ask about named journals but only about resources with abstractly stipulated levels of quality and reliability.  They also disregarded faculty input on cancellation decisions when all librarians admit that faculty input is decisive.  Other things are never equal, and result was a study of hypothetical preferences, not actual cancellation decisions. 

Also see Steve Hitchcock's collection of other objections to the PRC study and replies from Beckett and Inger.
http://www.eprints.org/community/blog/index.php?/archives/163-Self-Archiving-and-Journal-Subscriptions-Co-existence-or-Competition.html
http://www.earlham.edu/~peters/fos/2006_12_03_fosblogarchive.html#116541411621829751

A less hypothetical study was commissioned by ALPSP and conducted by Mark Ware, ALPSP survey of librarians on factors in journal cancellation, March 2006.
http://www.alpsp.org/ngen_public/article.asp?id=200&did=47&aid=157&st=&oaid=-1
http://www.earlham.edu/~peters/fos/2006_03_19_fosblogarchive.html#114294607786647567
The study is not free, but ALPSP has published a free summary.
http://www.alpsp.org/ForceDownload.asp?id=53

Quoting the summary:  "The three most important factors used to determine journals for cancellation, in declining order of importance, are that the faculty no longer require it (i.e. relevance to research or teaching programme), usage and price.  Next, availability of the content via open access (OA) archives and availability via aggregators were ranked equal fourth, but some way behind the first three factors.  The journal's impact factor and availability via delayed OA were ranked relatively unimportant....With regard to OA archives, there was a great deal of support for the idea that they would not directly impact journal subscriptions." 

Bottom line:  journals have much more to fear from their own price increases than from OA archiving.  If raising the risk of cancellations can be blamed for undermining peer review, then publishers are far more guilty than funding agencies with OA mandates.

The studies that support the publishing lobby and the studies that contradict it both show that there are many factors behind journal cancellations and many reasons to cancel subscriptions that are entirely unrelated to the rise of OA.  In fact, most journals were already suffering 5-10% attrition per year when OA was new and negligible.  If subscriptions continue to fall as the volume of OA archiving continues to rise, then it will be very difficult to disentangle the factors and decide what part of the attrition is attributable to OA.  This difficulty is aggravated by the fact that, contrary to supply and demand in a healthy market, journals respond to cancellations by raising their prices, and high prices cause many more cancellations than OA.  But publisher lobbyists are already blaming OA, with mounting desperation, while systematically diverting attention from their members' own hyperinflationary price increases and their cardinal role in causing journal cancellations.  As the difficult "disentangling problem" gets more difficult, we should expect to see even less care with evidence and even more OA scapegoating.

(8) There is evidence that OA archiving decreases *downloads* from publishers' web sites.  For example, the ALPSP describes this phenomenon in its April 2006 comments to the UK Gowers Commission.
http://www.alpsp.org/ngen_public/article.asp?id=0&did=0&aid=347&st=gowers&oaid=0
http://www.earlham.edu/~peters/fos/2006_04_23_fosblogarchive.html#114588756205545546

The effect on downloads is understandable:  many users who know about both the OA and TA editions, will prefer to click through to the OA edition (e.g. because they aren't affiliated with a subscribing institution or, even when they are, because authentication can be a hassle), and many users who only know about the OA edition will stop looking.  But it's important not to confuse decreased downloads with decreased subscriptions.

I haven't seen any evidence that OA leads to decreased downloads overall, that is, fewer readers and less reading.  On the contrary, the same evidence that OA leads to more citation impact shows that it leads to more readers and more reading. 
http://opcit.eprints.org/oacitation-biblio.html

But while shifting readership from the publisher web site to an OA repository is compatible with continued subscriptions and revenue, it can cause other problems for journals.  For example, it could throw off the journal's traffic metrics, which it uses to set advertising rates.

Insofar as publishers need accurate traffic data, this problem can be solved if OA repositories agree to share traffic data with publishers (indeed, with everyone).  Insofar as journals need eyeballs on their own site to retain advertisers, and not just data on eyeballs elsewhere, this problem can be solved, or at least mitigated, if publishers agree to post OA copies to their own sites.  After all, if there are OA copies in repositories, publishers have much to gain and nothing to lose by hosting their own.  If they host OA copies of the published editions, and not just the final versions of the authors' peer-reviewed manuscripts, they could even steal traffic away from the OA repositories.

(9) Some subscription journals have found that OA after an embargo period, even a very short one like two months, actually *increases* submissions and subscriptions. 

For example, this has been the experience of the American Society for Cell Biology and its journal, Molecular Biology of the Cell,
http://www.biomedcentral.com/openaccess/archive/?page=features&issue=6
...and the experience of the MedKnow journals (with TA print editions, full-text OA editions, and no embargoes),
http://openmed.nic.in/1599/01/Open_Access_in_India.pdf

In the face of the publishing lobby's unqualified prediction of harm, it should be enough to point out documented counterexamples of harmlessness (APS, IOP, and physics generally) and benefit (ASCB and MedKnow) to subscription journals.  When we trouble ourselves to look at the evidence, we find that different journals in different research niches will have different experiences.  All the examples I've cited have been cited many times in the debate and yet the most common response from the publisher lobbyists and trade associations (for example, the response from PRISM last week) is a mindless repetition of unargued confidence in a self-serving prediction.

(10) There are two strong signs that many publishers don't believe that OA mandates will defund peer review.  Their trade associations are either out of touch with their members or deliberately distorting the evidence.

First, many publishers have tried to head off an OA mandate at the NIH by encouraging grantees to comply with the current, voluntary policy.  They understand that the very low compliance rate, hovering at about 5%, is one of the strongest arguments to adopt a mandate, and they have tried to neutralize that argument by boosting compliance.  But either publishers don't expect these efforts to succeed or they don't seriously believe that increased compliance will defund them.
http://www.prismcoalition.org/docs/Letter_to_Zerhouni_7_31_06.pdf
http://www.reed-elsevier.com/index.cfm?Articleid=1681
http://www.earlham.edu/~peters/fos/2006_02_26_fosblogarchive.html#114139238210263579

Second, many publishers provide OA to their back issues, voluntarily, sometimes on the same timetable as the OA mandates they oppose.  Sometimes they even provide OA to a superior version (the published edition as opposed to the peer-reviewed but unedited author's manuscript) and to a wider scope of content (all the articles in the journal, not just the research articles or those funded by a certain agency).  If they seriously thought this would defund them, they wouldn't do it.

I wish I could point to a comprehensive list of publishers who provide voluntary OA to their journals on the same timetable as the OA mandates.  But there isn't one.  However, Highwire gives us good data for Highwire-hosted journals, mostly from society publishers.

As of August 30, 2007, here are the tallies of Highwire-hosted subscription-based journals that provide free online access after an embargo period or moving wall:
http://highwire.stanford.edu/lists/freeart.dtl

after 2 months = 1 journal
after 3 months = 6 journals
after 4 months = 9 journals
after 6 months = 18 journals
after 12 months = 197 journals
(Omitting those with longer embargoes)

Publishers who provide free online access on the same timetable as the OA mandates, or even sooner, don't object to OA itself or even high-volume OA.  And they certainly don't object to the effect of delayed OA on revenue and peer review.  If they don't object to OA mandates either, then their trade associations should qualify their message and acknowledge the dissenting voices.  If they do object to OA mandates, then the trade associations should still adjust their message, though in a different way, for the real concern of this group is not OA threats to revenue and peer review, but OA threats to publisher control over access.
http://www.earlham.edu/~peters/fos/newsletter/11-02-04.htm#control

(11) But even if publishers don't seriously believe that OA mandates will defund them, perhaps they will defund them anyway.  So let's continue the analysis and suppose that high-volume OA archiving will definitely cause journal cancellations, at least in some fields outside physics.  It just doesn't follow that defunding TA journals will defund peer review.

The DOAJ lists more than 2,800 peer-reviewed OA journals.  Together they account for a bit more than 10% of the total number of peer-reviewed journals in all fields and languages.  It's true that 10% is a small minority, but it's also true that it's impressive progress after only a decade of growth, more or less, especially when most of the money to pay for OA journals is still locked up in subscriptions.   

Even if the rapid rise of OA archiving undermines TA journals, it will not undermine the OA journals.  Even if libraries cancel subscriptions, peer-reviewed OA journals don't use subscriptions to pay their bills.  Hence even under the publishers' most catastrophic scenario, peer-reviewed journals will survive.  More importantly, the new mutant strain of journals is adapted to an environment of high-volume OA archiving and is not threatened by the further growth of OA. 

Just as the publishing lobby rarely acknowledges that TA physics journals thrive alongside high-volume OA archiving, it rarely acknowledges that OA journals provide peer review, are not threatened by OA archiving, and often practice OA archiving themselves.  It's much easier to argue that OA will kill peer review as such (or "to equate traditional publishing models with peer review", as Eric Dezenhall allegedly recommended to the AAP) than to argue that, at worst, OA will only kill peer-reviewed journals with a print-era business model.

But 10% is not 100%, and not even close.  If all or most TA journals fail, then could OA journals meet the demand?  For this, see the following three points.

(12) Here's where we have to focus on the weakness in the third link of the publishing lobby's causal chain:  the claim that if rising OA does cut journal revenue, then the journals will fold up and stop providing peer review.  It's much more likely that, if they could no longer sustain themselves on subscription revenue, they would convert to OA and continue providing peer review.  Survival as an OA journal would be vastly preferable to folding up, and most would try it at least experimentally before turning out the lights and closing the door.

If subscription journals find themselves losing a critical number of subscriptions, and the growth of OA archiving seems to be part of the cause, then they could try to stop the growth of OA or they could stop betting against the internet.  They could stop using a business model that makes cost recovery into an access barrier that limits audience and impact.  They could look for alternatives to the subscription model, and find that many are already being tested, and some proved, in the highly diverse world of OA journals.  There's a lot to examine and try out, and perhaps a lot to reject, before giving up and folding. 

At such a dinosaur moment, some will adapt rather than die, and most will at least try.  We know that adaptation is possible because many are already adapting.  Every year since 2005 the number of TA journals converting to OA has increased significantly.

Publishers don't like this scenario because OA journals have lower profit margins and "unproved" business models.  But it's one thing to argue that TA journals might be forced to adapt to a changing world and survive in a less profitable form, and quite another to argue that they will not survive at all.  However, publisher lobbyists prefer the Chicken Little objection and invariably fail to draw this distinction.

The objection that business models for OA journals are not proven should be taken seriously, and can be seriously answered.  Indeed, taking it seriously puts a different color on lobbyist predictions that publishers would fold up their TA journals rather than convert to OA.  Suddenly that kind of dramatic ending looks less like the result of inexorable causation than a business judgment made in advance of the facts. 

(To answer the "unproven business model" objection quickly:  At least two OA publishers using two different business models are now profitable.  Hindawi, with 51 OA journals, is the world's second largest OA publisher, and its OA journal program is profitable; all of its OA journals charge publication fees.  MedKnow, with 43 OA journals, is probably the third largest OA publisher, is also profitable, and none of its OA journals charge publication fees.  Apart from the health of these OA publishers, and the proof of concept they provide for fee-based and no-fee OA business models, there is the proven unhealth of the subscription model itself.  Even if OA models were riskier than they are, they would be worth exploring if the subscription model, as the University of California put it, is "incontrovertibly unsustainable".)

Hindawi
http://www.hindawi.com/

MedKnow
http://www.medknow.com

University of California, Letter to faculty from Lawrence Pitts, Chair of the Academic Senate, and the head librarians of the 11 UC campuses, January 7, 2004
http://libraries.universityofcalifornia.edu/news/facmemoscholcomm_010704.pdf

(13) Most importantly, the publishing lobby never acknowledges that if subscription-based journals did fail, the money formerly spent on subscriptions would be freed up to pay for peer-reviewed OA journals.

Peer review could survive, could be funded at the same levels as today, and could be funded through a business model immune to the continuing "threat" of OA. 

The new generation of peer-reviewed OA journals needn't be new journals.  They could be the existing TA journals, with their existing brands, standards, reputations, editors, referees, readership, and impact.  The money now spent on subscriptions could keep existing TA journals afloat as OA journals if they couldn't sustain themselves as TA journals. 

For the same reason, the new generation of OA journals could be the equals of their TA predecessors in quality.  There are reasons to think they could, on average, be even better.
http://www.earlham.edu/~peters/fos/newsletter/10-02-06.htm#quality

In short, it's false that the subscription model is the only business model that can support peer review.  It's false that the hypothetical collapse of subscription journals would mean a defunding of peer review.  On the contrary, it would mean a windfall replenishing of university library budgets that could fund any successor to subscription journals that we chose. 

But first a qualification:  We don't know what would happen to the money freed up by TA journal subscriptions.  It will be allocated by human beings thinking about a complicated range of opportunities and constraints.  The decisions will be made either by librarians, in spending the library budget, or by university administrators, in setting the library budget.  There will be many attractive uses for the money and many differences in policy and priority from one institution to another.

I've often argued that the money should go first to peer-reviewed OA journals, in order to replace the departing peer-reviewed TA journals, and second to the book budgets long depressed by rising TA journal prices.  But that's a normative argument, not a prediction.  The money might go instead to surviving TA journals (who might even raise their prices to grab some of the new library bounty) or to needs outside the library like new faculty, new administrators, higher salaries, and financial aid.

The point is that it will be decision by academic leaders.  If we want to preserve peer-reviewed journals after the hypothetical failure of TA journal subscriptions, we could do it without raising extra money, without curbing OA, and without propping up TA journals and encouraging them to make their financial viability depend on access barriers or artificial scarcity to published knowledge.  The question is not about the inevitable consequences of OA, but about the will of the academic community.  If we find ourselves with windfall savings from the cancellation, conversion, or demise of peer-reviewed TA journals, and we don't spend it on peer-reviewed OA alternatives, and if peer review withers and dies as a result, the fault will lie with university administrators, not with OA. 

When publishers assert that threats to TA journal subscriptions automatically translate into threats to peer review, like the collision of billiard balls, they only insult their present customers, whose intelligent decisions will determine the outcome.

In a 2005 essay on OA-TA coexistence, I made the point this way:  "[T]his is nothing like predicting a force of nature.  We're talking about the actions of interested human beings, including ourselves."
http://www.earlham.edu/~peters/fos/newsletter/03-02-05.htm#coexistence

The OA debate could be more productive and even collaborative if it could be regrounded in this reality:  even with OA mandates, human beings will decide the future of peer review.  Even with OA mandates, we face a policy decision, not a row of falling dominoes.  Even if OA causes mass cancellations of TA journals, we'll have choices and money.  If OA mandates threaten the current system of scholarly communication, they do not threaten the money to support peer review, but only the business model supporting peer review from the age of print.  If the future of peer review is uncertain, it's not because OA makes its existence more difficult or less likely, or because it reduces funding, but because we don't know who will spend the money or on what.  However, we don't have to guess the future when we can make it.  If we want OA mandates and peer reviewed journals, we can have them.

(14) If the money to support peer review will survive the failure of TA journals, so will the people who provide peer review.

Scholarly journals don't pay authors for articles and they don't pay referees for participating in peer review.  Business setbacks at TA journals, including total failure, will not have the slightest effect on the willingness of scholars to write and referee research articles.  The same talent will be available, at the same price, motivated by the same interests, to any peer-reviewed OA journals that may arise to take the place of failed TA journals. 

Publishers of TA journals like to present themselves as peer review providers --indeed, the only peer review providers-- but they merely organize unpaid volunteers to provide peer review.  This organization is nontrivial and costs money, and at TA journals the money is paid by publishers from subscription revenue.  But if subscription journals fail, the same money will be available to organize the same people to provide the same service. 

The rub for publishers is that the money won't be in the hands of publishers, but in the hands of universities and other former subscribers.  This explains publisher resistance, but it doesn't change the facts about peer review.  Peer reviewers aren't paid by TA journals and will live on, with unchanged interests and motives, if TA journals die off.  Peer reviewers are paid by their employers, usually universities, and this subsidy to peer-reviewed journals will be available to any journals, TA or OA, that want to make use of it.

If TA journals do fold up rather than convert, and we do have to redirect the money formerly spent on them to support peer review at new journals, will that be turbulent and messy?  Almost surely, and also exciting and hopeful, like any period of flux in adapting to fundamentally new technologies and taking advantages of the opportunities they offer.  (A big topic for another day.)  Will it be impossible or unlikely?  I don't see any reason to think so, not even with the help of the publisher lobbyists.

(15) In conclusion, publishing lobbyists claim that the growth of OA archiving will undermine peer review, but they don't connect the dots.  Insofar as they hint at a theory, they assume a causal chain of events not borne out by evidence and contradicted by the evidence to date.

The first step in evaluating their objection to OA mandates is to distinguish the effect on subscriptions from the effect on peer review.  The publishing lobby's attempt to blur this distinction is self-serving FUD.  Moreover, both effects are much weaker than they claim.

There's strong evidence that OA archiving has not caused cancellations in physics and will not cause cancellations in those fields that turn out to be like physics in the relevant respects.  At the same, it might eventually cause cancellations in fields that turn out to differ from physics in the relevant respects.  For now, nobody knows which fields, or how many, will turn out to be like physics, and we don't even have a good grasp of the relevant variables.

There are attested cases in which immediate OA and OA after a short embargo actually increase subscriptions (and submissions) at TA journals.  Nobody knows how widely this pattern would hold if other journals adopted the same policies.

OA mandates will increase one library incentive to cancel subscriptions (eventual OA to a subset of articles) but won't affect four incentives to renew subscriptions:  to provide immediate access; to provide access to the published editions; to provide access to all the research articles in a journal; and to provide access to all the contents in a journal beyond research articles.

Publishers themselves have uncovered good evidence that high journal prices far surpass OA archiving as a cause of cancellations.  At the same time, many publishers voluntarily provide OA to back issues on the same timetable as the OA mandates, and encourage compliance with voluntary OA policies as a tactic to head off mandates, without finding that these practices undermine their subscriptions.

The connection to peer review is even more tenuous than the connection to subscriptions.  One reason is the existence and viability of peer-reviewed OA journals, which would survive unharmed if libraries cancelled all their journal subscriptions. 

It's true that most peer-reviewed journals today are TA.  But we cannot even say that *to the extent* OA archiving undermines TA journal subscriptions, then *to that extent* it will undermine peer review.  That claim makes the two events appear to be connected by rigid causation, when in fact they are connected, if at all, by human decisions.  There are at least two decision points along the way that can sever the connection.

The first decision point arises for journals losing a critical number of subscriptions:  they must decide whether to fold up or convert to OA.  The second arises for the academic community when TA journals do fold up or convert to OA:  it must decide whether to use the money freed up from peer-reviewed TA journals on peer-reviewed OA journals.  It must decide, in other words, whether to continue supporting peer review with the funds that formerly supported peer review.  TA journals could all die and strong peer review live and thrive, with no greater overall cost, if we want it to.  If peer review dies, it will not be from the failure of peer-reviewed TA journals but own our failure to fund their successors with the same money.

In short, and with some simplification, will the rise of OA archiving cause cancellations of TA journals?  Either it will or it won't.  The evidence, as opposed to the fears, is that it won't.  But if it will in some fields other than physics, then we'll lose some peer-reviewed TA journals.  As we do, we'll reap savings that we can redirect to peer-reviewed OA journals (which might be the same journals under different business models).  To make this scenario frightening, the most the publishing lobby can say is that spending the money to support peer review will not be automatic.  On the other side, to refute the publishing lobby's current claim, all we have to say is that spending the money on something else will also not be automatic.

If we do find that some fields are not like physics, and the growth of OA archiving causes journal cancellations, then we will have discovered an incompatibility between OA to publicly-funded research and the subscription business model, at least for those fields.  That would a problem with the subscription business model, not with OA.  It would emphatically not be an incompatibility between OA and peer review.  If we do discover an incompatibility between OA mandates and the subscription business model, then the OA mandates would still be justified, especially at public funding agencies, by the principle that puts the public interest ahead of special interests.  If we reach that point (repeat, if), then it would be perverse and backwards to compromise the public interest in accelerating research, sharing knowledge, and serving taxpayers in order to prop up publishers who are not able to adapt to good public policy.  And if we reach that point and choose OA, we would not be limiting ourselves to OA minus peer review.  For we can have universal OA archiving and adequately funded, independent peer review, if we want them both.  We won't have to find new money, but we will have spend the money in new ways.

If we take this course and observe that OA archiving and journal cancellations are rising in tandem, then the best response for funding agencies would be to stand fast rather than retreat to today's dysfunctional system of artificial scarcity.  The best response for universities would be to redirect the savings from cancelled subscriptions to peer-reviewed OA journals (through journal subsidies and/or publication fees) rather than cut the library budget.  The best response for publishers would be to convert their journals to OA, or launch new OA journals, and accept the new revenue.

I am not saying that we should deliberately defund TA journals in order to fund OA journals, and I am not saying that TA journals should die.  I'm saying that publishers have not even come close to making good on either of their claims, that OA archiving would kill TA journals or that killing TA journals would kill peer review.  I am saying that funding agencies should mandate OA archiving without fear.  (So should universities, but I'm omitting that argument here.)  Either peer-reviewed TA journals will survive the transition, as they have in physics, or they won't and we'll face the decision whether to re-fund peer review by spending the savings on peer-reviewed OA journals or defund it by spending it elsewhere.

Academic publishers defending peer review should set an example of reasoned argument and sensitivity to evidence.  The AAP/PSP and other publisher trade associations have not lived up to this responsibility, and should be called on it by their members.  A good example is the public statement by Rockefeller University Press, dissociating itself from the hand-waving of the PRISM campaign.  For another, see the editorial in The Lancet (an Elsevier journal) from October 16, 2004, in response to AAP lobbying against an OA policy at the NIH:

[A]s editors of a journal that publishes research funded by the NIH, we disagree with [AAP President Patricia] Schroeder's central claim.  Widening access to research is unlikely to bring the edifice of scientific publishing crashing down. Schroeder provides no evidence that it would do so; she merely asserts the threat.  This style of rebuttal will not do.  Indeed, the aggressive rhetorical line taken by the AAP unnecessarily pits publishers against the interests of science and the public....

http://www.thelancet.com/journals/lancet/article/PIIS0140673604172322/fulltext
http://www.earlham.edu/~peters/fos/2004_10_10_fosblogarchive.html#a109794429081763100

----------

Read this issue online
http://www.earlham.edu/~peters/fos/newsletter/09-02-07.htm

SOAN is published and sponsored by the Scholarly Publishing and Academic Resources Coalition (SPARC).
http://www.arl.org/sparc/

Additional support is provided by Data Conversion Laboratory (DCL), experts in converting research documents to XML.
http://www.dclab.com/public_access.asp

==========

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 <SPARC-OANews-off@arl.org>.

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
http://www.arl.org/sparc/publications/soan

Peter Suber's page of related information, including the newsletter editorial position
http://www.earlham.edu/~peters/fos/index.htm

Newsletter, archived back issues
http://www.earlham.edu/~peters/fos/newsletter/archive.htm

Forum, archived postings
https://mx2.arl.org/Lists/SOA-Forum/List.html

Conferences Related to the Open Access Movement
http://www.earlham.edu/~peters/fos/conf.htm

Timeline of the Open Access Movement
http://www.earlham.edu/~peters/fos/timeline.htm

Open Access Overview
http://www.earlham.edu/~peters/fos/overview.htm

Open Access News blog
http://www.earlham.edu/~peters/fos/fosblog.html

Peter Suber
http://www.earlham.edu/~peters
peter.suber@earlham.edu

SOAN is an open-access publication under the terms of the Creative Commons Attribution License.
http://creativecommons.org/licenses/by/1.0/


Return to the Newsletter archive