Tuesday, March 17, 2009

Open Access: A publisher's perspective

In an article I posted on 10th March I discussed the issue of whether the Green and Gold roads to Open Access (OA) should be viewed as complementary or competitive strategies for the OA movement.

This generated some feedback on the Web: feedback, for instance, from OA advocates like Peter Suber and from Stevan Harnad; feedback from librarians like the University of California's Ivy Anderson; and feedback from consultants like Joe Esposito.

In the belief that further discussion of these issues might be helpful, and keen to get the perspective of a publisher, I was pleased when Steven Hall, former Commercial Director at Wiley-Blackwell, agreed to let me publish his thoughts on what I had written.

Prior to Wiley's acquisition of Blackwell Publishing Hall was Journal Sales and Marketing Director at Blackwell and on the Management Committee of the business. Before Blackwell, Hall was Senior Vice-President of Publishing and Publisher Relations for ProQuest, where he ran its higher education publishing, and before that Managing Director of the electronic humanities publisher Chadwyck-Healey before its acquisition by ProQuest.

Hall left Wiley-Blackwell last April and is now running his own publishing consultancy business. He specialises in areas like pricing, product development and business development, and strategy.


Steven Hall


In his article Open Access: whom would you back? Richard Poynder makes a number of assertions about access to journals, the impact of the ‘Big Deal’ and the costs of Gold OA publishing which repeat many of the canards of the OA debate, and which are simply not supported by the data.

He writes: This affordability problem means that each year librarians are forced to cancel more and more journals, depriving their researchers of access to a growing proportion of the refereed literature.

This claim appears to be based on an article by Stevan Harnad and others in which it is stated "Spiralling price rises mean research libraries can afford to subscribe to fewer journals and, in turn, users have access to fewer articles — whereas in the online age, we might have expected the opposite."

Well, libraries might be subscribing to fewer individual titles but their users have access to far more articles: the opposite really is the case. In 1992-1993, before electronic journals publishing really got underway, the median number of paid journal subscriptions across ARL university libraries was 15,016. By 1999-2000, with electronic journals publishing well established, the median figure had increased slightly to 16,121.

By 2005-2006, that median figure had grown by 73% on the 1992-1993 figure, to 25,967. (Click here for the statistics.) The 2006-2007 statistics have been compiled on a different basis, counting unique titles rather than subscriptions, but show a university median figure for titles purchased of 40,533.

This is a huge increase in access, and this is within the ARLs, the top research libraries in North America which historically have subscribed to the largest number of journals; if you look at less research-intensive institutions, then the growth in access is even greater, from a few thousand titles to tens of thousands. The UK statistics, which used to be compiled by LISU for the whole UK HE sector, tell the same story, as do the Australian statistics.

The Big Deal

This is, of course, as a result of the 'Big Deal'. It's perfectly possible to disagree with publishers' pricing policies, with the way they have sold the Big Deal, and to dislike some of its effects, but it cannot be argued that access to journal literature has been reduced. Researchers and students now have access to far more journals than at any point in the past, and this greater access is available at a much wider range of institutions.

On a related topic, Richard claims that: ...as scholarly journals were moved to the Web publishers began to bundle them into all-you-can-eat packages popularly known as the "Big Deal", and then sold large portfolios of journals rather than single subscriptions.

This is one of the assertions made in the Study on the Economic and Technical Evolution of the Scientific Publication Markets in Europe [pdf], which was commissioned by the Directorate-General for Research of the EC and published in 2006, and which in my view would never have made it past the peer-review process of any half-decent scholarly journal.

It may be true that a few publishers have tried to push customers towards taking a single package in which individual subscriptions ceased to be recorded, but most publishers have sold their Big Deal package as an add-on, a supplement to the customer's individual journal subscriptions. And, so far as I am aware, no publisher has offered only a Big Deal; all have continued to offer individual journal subscriptions as well as packages.

The Big Deal may have helped concentrate even more influence in the hands of a fairly small number of large publishers, but it was not imposed on libraries. The great majority of libraries have warmly embraced it, principally because of the huge increase in access to journals that it has provided, at a fairly small marginal cost, and partly because of the lower annual price increases that have come with it.

Richard also claims that: Since this [the Big Deal] reinforced the market power of the big players it allowed them to ramp up their prices even further.

He presents no evidence for this. Publishers were ramping up their journal prices (for all sorts of reasons, including the growth in research output) long before the Big Deal came along. Again, the ARL Statistics and the UK data support this.

Journal prices have been increasing at a rate above inflation for years and the advent of the Big Deal has not increased them further; in fact, price increases at title level have moderated in the last few years and the Big Deal has in many cases further moderated the annual increases in total costs that any library pays for a particular publisher's journals.This is not to argue that annual increases in journal costs well beyond the increases in library budgets are sustainable; it's to present facts that are frequently conveniently ignored in this simplistic debate.

Publishing costs

Moving on to the costs of Gold OA publishing, Richard makes two more statements in his article which require comment: In an online world scholarly publishing should surely be cheaper: there are no print costs... and ... Springer has never demonstrated how or why it costs $3000 to peer review a scholarly paper.

On the first, we are not yet in an online world. We remain in a hybrid world of print and digital, partly because of a certain conservatism on the part of librarians and academics in some parts of the world and in some disciplines; and partly because of the absurd situation that VAT is payable on electronic journals but not on print journals, or at a higher rate than on print journals, within the EU.

We will get to a largely electronic environment, but we are still a good number of years away from it, and in the meantime publishers have to bear both sets of costs.

On the second, Richard is careful to say that PLoS charged an APC of roughly half of Springer's charge of $3,000 at the time Springer's service was launched. This is disingenuous. PLOS now charges $2,850 for PLoS Biology and PLoS Medicine and $2,200 for all its other journals with the exception of the much more lightly reviewed PLoS ONE.

PLoS itself says openly that there are real and substantial costs to peer-review, editing and hosting of journal articles. PLoS has been increasing its APCs faster than any large commercial publisher and when I last looked at its accounts it was still a long way from breaking even on an annual basis and it may never break even on a cumulative basis if you take into account the very large grants and donations with which its launch was supported.

PLOS now charges more than some commercial publishers for its APC, has no print costs, aims to make no profit and may still be unable to balance its books. It seems that it costs at least $2,850 to peer review, edit and host an online article in a high-quality journal.

Green v. Gold

As for the whole issue of Green v. Gold OA, it's easy to see why Gold OA is acceptable to publishers (it's just another business model) and why Green OA isn't — it relies on the publisher giving away its investment in editing and peer review, and no publisher is going to do so without a fight.

When a researcher gets a grant to research — well, let's say a Mayan archaeological site in Mexico — he doesn't expect the airline to fly him there for free, nor the porters to provide their services for free, nor Dell to provide him with a free laptop. Why, then, should he expect the publisher to provide its editing and reviewing services for free? Publication is a legitimate part of the research process and it has genuine costs; why shouldn't these be recouped?

Green OA also ignores one other important element that the publisher brings to the scholarly process, and that is the brand of the journal which gives the imprimatur to the research article. This is something that academics and librarians, and even Green OA proponents, value very highly but which the latter believe should simply be given away by the publisher.

Researchers are, of course, able to make their research freely available — before it is reviewed, edited and has the journal brand attached.

The Green v. Gold debate is ultimately a dead end if the arguments are based on assumptions and assertion rather than hard facts. The EC study mentioned above and the JISC’s recent report on Economic Implications of Alternative Scholarly Publishing Models: Exploring the Costs and Benefits both lack reliable data to support their claims (they are both based on theoretical models) and neither seems to have involved publishers.

In time the PEER project may provide the data we need to assess the potential impact of Green OA. As for Gold OA, if its supporters believe that the current APC s are too high, from both commercial and not-for-profit publishers, then perhaps they should set up their own alternative peer review process and compete with the publishers.

Steven Hall


Dana Roth said...

Leaving aside concerns about signing up for multi-year Big Deals, the PLOS example appears to be a straw man. "PLOS now charges more than some commercial publishers for its APC, has no print costs, aims to make no profit and may still be unable to balance its books. It seems that it costs at least $2,850 to peer review, edit and host an online article in a high-quality journal."
This seems a little disengeneous in that a better example is the charges of the APS 'Free to Read' initiative under which "The current FREE TO READ fees will be $975 for articles in Physical Review A-E and $1300 for Letters in Physical Review Letters."

Stevan Harnad said...


I'll leave it to those who enjoy publisher-baiting (I don't) to spar with Steven Hall on the subject of journal pricing and degree of accessibility across the years.

My own view is that publishers can try to charge whatever they like and can get for subscriptions, whether singly or in bundled deals, and that institutions should pay whatever they want and can afford for the subscriptions they want and can afford, singly or in bundled deals. (That's pretty much the way things are today.)

But all authors should self-archive all their refereed final drafts in their institutional repositories, immediately upon acceptance for publication, so that all would-be users can access them online, not just those whose institutions can afford to subscribe to the journal in which they are published. This is called providing Green OA.

And all institutions and all research funders should mandate that all their authors and fundees must provide Green OA.

Apart from a certain degree of vagueness, hedging and double-talk, most of the journals of the three Wileys -- John, Blackwell, and Berlin -- have all already given their formal green light to their authors to go ahead and provide immediate Green OA. (The exceptions are journals that ask for an embargo of 6 or 12 [and occasionally 24] months.)

So, apart from the few embargoing Wiley journals (for which there is a solution too, see below), Wiley cannot be faulted for its concrete actions regarding Green OA. What about the whinging about Green OA?

Should we feel sympathy for publishers who claim that Green OA is free-riding on their hard-earned investment? Hardly. The hard work is mostly the authors'; the investment is mostly that of their institutions and funders; even the peer reviewing is all done for free by the peer-reviewers. The publisher's investment is in providing the service of managing the peer review -- and of course in providing the print and online edition.

Today, as we speak, the publisher's investment is fully repaid by the fact that the authors, their institutions and their funders allow the publisher to be the sole vendor entitled to sell and earn profits on the sale of their writings. The author -- and the author's institution and funder -- charge nothing for giving the publisher the exclusive right to sell their writings, and ask for no royalties in return.

Hence it borders on the absurd for the publisher to complain if the authors also give away their writings for free online. As long as the publisher is making ends meet through subscriptions, there isn't even anything to talk about. If and when the availability of the Green OA version causes subscription cancellations, it is time to think of phasing out the print and online edition and their associated expenses, and offloading access-provision on the network of institutional repositories containing the refereed final drafts. Then all that will be left of journal publishing is to provide the peer review service (and certify its outcome with the journal's name and track-record), a service whose far lower costs the institutions can pay out of just a portion of their annual subscription cancellation savings. (Note that no Gold OA publisher, including PLoS, has yet done this, but that is because Green OA is not yet universal!)

But until and unless Green OA makes subscriptions unsustainable, it is sophistical to say that authors self-archiving their refereed drafts (and their institutions and funders mandating it) amounts to free-riding on publishers' hard-earned content, like using airplanes and computers without paying for them.

(Now I promised a solution for the case of the articles published in journals that embargo open-access provision for 6 months or more: All authors deposit (and their institutions and funders mandate deposit) anyway, immediately upon acceptance for publication; but access to the deposit may be set as Closed Access instead of Open Access. This provides almost-immediate, almost-OA, because whenever a would-be user reaches the metadata of a Closed Access deposit, the Institutional Repository has an automated request-a-copy Button. One click sends an automatic email eprint request to the author, requesting a copy for research purposes; the email allows the author to authorize the automatic emailing of the copy with one click. Almost-OA will tide over research needs during embargo periods, until all embargoes die their inevitable and well-deserved deaths under mounting, unstoppable pressure for immediate OA and its increasingly apparent benefits.)

Richard Poynder said...

Steven Hall has asked me to stress that the views expressed in his commentary above are his personal views alone, and not those of any publisher for which he has worked.