Growth of OA

Graphs on any topic related to OA in the past five years will show large-scale growth. This might be growth of the number of journals in the Directory of Open Access Journals (DOAJ), growth in the number of and size of subject and institutional repositories, or growth in the number of OA publishers. Virtually every aspect of the burgeoning OA publishing industry has expanded, and at rapid rates. For example, Figure 9.1 shows the number of OA papers published by a selection of major publishers between 2000 and 2011. It is clear that the engine of growth here has been the new OA publishers – BioMed Central (BMC), Hindawi and the Public Library of Science (PLOS) – all of whose growth far outstrips that at Springer and OUP. Hindawi and PLOS in particular have seen huge growth, largely due to the impact of PLOS ONE in the latter’s case, and due to rapid and aggressive expansion and acquisition in the case of the former. It’s telling that PLOS published 16,264 papers in 2011 compared to BMC’s 19,996, but that PLOS only publishes seven journals in comparison to BMC’s 233. Anyway, the growth of these three major OA publishers is abundantly clear.

Moving on from articles to journals, Figure 9.2 shows the number of OA journals listed in the DOAJ. As of 30 May 2012, there were 7815 journals in the DOAJ, considerable growth even since the end of 2011, and 76 per cent more than at the end of 2009.

It is not just journals that are growing either – Figure 9.3 shows the growth of PMC (formerly PubMed Central), a vast resource of scientific and medical full-text versions of record, and Figure 9.4 shows the growth of arXiv, a huge pre-print server favoured in the physical sciences and mathematics.

We could labour this point here for a while, but we won’t. Basically, the key is that the growth of OA has been quick, is getting quicker, and is here to stay.

The mega and cascade journal concepts

We have already described PLOS ONE as the driver of PLOS’s growth. PLOS ONE was the first ‘mega journal’ and has sparked a succession of clones. PLOS describes the journal’s scope as follows on their website:

It is this model of accepting ‘methodologically sound’ papers which has allowed PLOS ONE to grow at such a fast rate. By not limiting themselves, as most journals traditionally have done and as PLOS had done with its first six journals (e.g., PLOS Genetics, PLOS Pathogens), PLOS created a journal with no size barrier. There are no print costs to worry about. There is no limitation on scope, and no limitation on submissions. As long as the science is sound, you’re in. This simple and tremendously scalable idea has enabled PLOS ONE to grow at such a rate that PLOS predicts the journal will account for 3 per cent of all STM publishing in 2012. Table 9.1 shows the growth of PLOS ONE submissions and published papers more clearly.

So what’s the draw of PLOS ONE to authors? There are several factors which make PLOS ONE a desirable place to publish. Firstly, speed of publication – rightly or wrongly (and this isn’t the place to debate this issue – visit the Scholarly Kitchen if you’d like to read more on this discussion¹¹). PLOS ONE’s methodological peer review and high acceptance rate has been perceived as a speedier route to publication than the traditional journals route. Secondly, the article processing charge (APC) is cheaper than for most major OA publishers at US$1350,¹² and, thirdly, PLOS ONE’s size begins to bring it an aggregational desirability all of its own – the PLOS ONE database is now virtually the size of a small publisher, and it is the largest journal in the world.

One thing PLOS has been very keen to emphasize is that it is not particularly concerned about its impact factor (IF): ‘At PLOS, we believe that articles in all journals should be assessed on their own merits rather than on the basis of the journal in which they were published.’ PLOS ONE’s basic scope and remit is a rejection of the sort of limited analysis which the IF is predicated on, and the huge size and generous acceptance policy of the journal would seem to count gainst PLOS ONE’s chances of doing well in Thomson Reuters’ calculations. Despite all this, PLOS did apply for an IF and in 2010 PLOS ONE achieved an IF of 4.411, which is very healthy.¹³ Figure 9.5 shows the marked upturn in PLOS ONE publications which coincided with the announcement that the journal would be receiving an IF. Whatever the opinions of PLOS on the matter, it is a reality of journals publishing, or at least of STM journals publishing, that the IF is important and so the receipt of an IF stands as another step in the realization of PLOS ONE’s acceptance into the mainstream.

PLOS ONE’s success has also been financial – turning round PLOS’s finances from a position of loss-making to strongly profitable (see Table 9.2).

PLOS ONE’s success has been so spectacular that it has led to the entrance of virtually all publishers into the mega journal market. The theory is simple – launch a journal which covers everything, don’t limit yourself, charge a low-ish APC, and wait for the money to roll in. In the past two years new launches have come from Springer, BMJ, SAGE, Nature, and many societies including the Genetics Society of America and the Company of Biologists. However, there is some evidence to suggest a slightly diminished return – as you might expect. Figure 9.6 shows that BMJ, SAGE, and so on, did not see the same levels of publication as PLOS in the early months of their mega journal. There are multiple reasons for this on each title of course (e.g., SAGE is working in the social sciences with SAGE Open) but, generally, we can make the assumption that PLOS ONE continues to benefit from being the first mega journal in the field and the most well-known.

For all its obvious benefits, the mega journal concept is not a panacea for all ills and it does present some problems, particularly for publishers who base their brand strongly on high quality. Doubts persist over the quality control of journals with methodology-based acceptance criteria. It is notable and clear that Nature branded its mega journal Scientific Reports and not ‘Nature Scientific Reports’ for this reason expressly. It is difficult to place a stamp of excellence on a title when you explicitly agree to let the reader be the final arbiter of quality. However, as with anything, there are counter arguments to this. A lot of the work published in PLOS ONE is material rejected from other journals with up to 95 per cent rejection rates; clearly not all of this is going to be bad science – any journal turning away 19 in every 20 papers is turning away good material. PLOS ONE and the other mega journals are simply providing the opportunity to authors not to have to submit to journal A, be rejected, submit to journal B, be rejected, submit to journal C, be rejected, and eventually pitch up in journal D. They are giving authors the opportunity to get through to publication right away, and if that article is of publishable quality then they are also granting a greater efficiency in the system, which brings us on to the cascade.

The theory of the cascade

Neither the mega journal concept nor the cascade have to be OA, but they are both best fitted to OA publication models and have been taken advantage of to the greatest degree by OA publications. The cascade is a simple theory based around the inefficiency of wasting peer review. The theory goes that if your top journal, journal A, is rejecting 90 per cent of papers, why allow all that work in review and recommendation to go elsewhere; why not use some of those papers in another journal. BMC have been using the cascade for many years (see Figure 9.7).

The simple efficiency of the cascade raises the question of why it had not been more widely adopted previously, and why it took the entrance of the new OA players into the market to herald its widespread adoption. Cascade publishing makes the entire publishing ecosystem more efficient. If journal B can accept an article based on the recommendation of the editors of journal A, this means journal B gets a paper without bespoke review, the author gets published more quickly (as per the mega journal), and the wider ecology of peer review benefits from the same paper not needing another set of reviewers to tell everyone what they probably already knew: that the article is good enough for publication in journal B, just not quite up to the standard of journal A. Incidentally, it doesn’t have to be a quality-based decision either – journal A’s editors/reviewers could quite easily decide that the paper is simply a better fit with the editorial direction of journal B.

Thus, it is no surprise to see a major publisher like Wiley follow BMC down the cascade route. Wiley Open Access, launched at the tail end of 2010, uses the cascade from established and successful, usually society-owned, journals.¹⁷ Editors and reviewers of these titles are presented with a ‘reject and refer’ option (as well as the standard options of accept, reject, revise, and so on) and authors are then given a choice: to take their paper elsewhere or to choose to publish in the OA cascade journal. And therein lies the rub. The Wiley Open Access titles are, as you might wisely surmise, open access, but, more importantly, they charge an APC for all papers which the parent title does not. While there is nothing wrong with this, and there is absolutely no compulsion on the author to publish, and therefore pay, in the new journal, there is a brand problem here. In reality, a publisher launching a suite of journals like this in the current climate, certainly in STM, does not really have a choice but to launch OA titles, as subscription journals would be very unlikely to earn enough subscriptions to survive. However, as ever, cascading into OA could lead to a suggestion that the publisher is exploiting the rejected papers for profit. There is some truth in this of course – quite rightly, the publisher is looking to gain value from the time and work spent by their editors and reviewers, but as discussed above there are many other benefits to the cascade, including benefits to authors. Nevertheless, it is a tricky brand proposition: how do you market your cascade journal? You could have a cascade journal of very high quality – for example if journal A is accepting only 5 per cent of papers the cascade journal could still be taking only the next 10 per cent (probably better than many pre-existing titles on the market) – but what to call it? ‘Journal A Lite’? ‘Journal A but a little bit worse’? There is also a brand issue concerning whether this is a publisher or a journal initiative. Over the years, most studies have shown that the journal name is far more important to authors and readers than the publisher name¹⁸ – yet in the mega journal market we see SAGE Open, Wiley Open Access, Taylor & Francis Open, SpringerPlus, and so on. Is it an attractive proposition to publish in a publisher-branded journal? PLOS ONE has avoided this problem because PLOS itself is a name, a non-profit enterprise, and one which has branded all its journals with its name, in the same vein as Nature (until Scientific Reports).

The mega and cascade journals also present a problem to publishers which have positioned themselves in the market more specifically on quality – working with small to medium-sized lists; publishers like university presses and societies. If one of the central themes behind mega journals is to get research out into the system and leave the audience to judge its importance, how do you justify publishing this material as a publisher only publishing ‘high-quality’ material? Or is there a certain inevitability, a revolution in the journals publishing industry, which makes that point of view slightly archaic? Can a ‘high-quality’ publisher add quality in other ways – quality of presentation, of copy-editing, or editorial teams and innovation?

We are still in the early phases of the mega journal era, and there is no reason to believe that the entire publishing world will be a mega journal in the imminent future, but it is clear that the phenomenal success of PLOS ONE has led to a whole gamut of imitations (see Table 9.3). It remains to be seen how these develop over time, and whether anyone can replicate the success of PLOS. At PLOS, it will also be informative to see if its other titles eventually fold into PLOS ONE. PLOS has given no indication of this as of yet, but it would seem logical. It certainly has not launched any titles since PLOS ONE. This cuts to the question – are mega journals to become the focus of all publishing, or just an addition to it?

Table 9.3

Publishers currently in the mega journal market

Journal name	Publisher	Content from	APC	Subject area
PLOS ONE	PLOS	2006	US$1350	All disciplines within science and medicine
Sage Open	Sage	April 2011	US$395 (introductory rate)	Social sciences/humanities
SpringerPlus	Springer	March 2012	£690	All science disciplines
BMJ Open	BMJ	February 2011	£1200	Medicine
BMC Research Notes	BioMed Central	February 2008	£675	Biology and medicine
Scientific Reports	NPG	June 2011	$1350	All science disciplines
AIP Advances	AIP	March 2011	£1350	Applied physics
G3	Genome Society of America	June 2011	$1950 (non-member charge)	Genetics
Biology Open	Company of Biologists	Autumn 2011	$1350	Biological sciences
RSC Advances	RSC	2011	£1600	Chemistry
Physical Review X	APS	August 2011	$1500	Physics
Open Biology	The Royal Society	Autumn 2011	£1200	Biological sciences
Cell Reports	Cell Press/Elsevier	2011	$5000	Cell sciences

If the success of PLOS ONE is the carrot driving publishers towards OA, government regulation and funder mandates are the very hefty sticks. In the past five years there has been almost never-ending legislation and counter legislation in the US, with the most notable result so far being the National Institutes of Health (NIH) mandate¹⁹ requiring all authors funded by the NIH in the US (one of the world’s biggest funders of medical research) to ‘submit or have submitted for them to the National Library of Medicine’s PubMed Central an electronic version of their final, peer-reviewed manuscripts upon acceptance for publication, to be made publicly available no later than 12 months after the official date of publication’.²⁰ The NIH mandate’s adoption has not proved too frightening for publishers, who have usually co-operated with PMC to deposit material directly rather than leaving it at the discretion of authors.²¹ Rather more threateningly, the Federal Research Public Access Act (FRPAA)²² proposes to make all US Government-funded research publicly available within six months of publication, which certain studies have shown bears significant risks to the sustainability of scholarly publishing.²³ Also in the US, we have had COMPETES,²⁴ one provision of which called for the White House Office of Science and Technology Policy (OSTP) to create a Public Access Committee to look at the possibility of expanding the federal OA mandate,²⁵ and, perhaps most controversially, the Research Works Act (RWA), which contained provisions to prohibit OA mandates for federally funded research.²⁶ The RWA was put forward by Elsevier-backed congressmen and women, and was strongly supported by, among others, Elsevier²⁷ and the Association of American Publishers.²⁸ Unfortunately for all those backing the RWA, though, it did not act as a silver bullet against the OA movement; instead, it had rather the opposite effect. The RWA can be perceived to have caused Tim Gowers’ Elsevier boycott,²⁹ and subsequently the rather daftly and emotively labelled ‘Academic Spring’.³⁰ It is important to note that far from all publishers were behind the RWA – most notably Nature, which labelled it a ‘ridiculous distraction’.³¹

Daftly named or not, the ‘Academic Spring’ has proved a major signal post in the development of OA. Elsevier’s climbdown on the RWA³² and the subsequent mainstream coverage in the UK and US media³³ have heralded the emergence of OA as a significant issue in mainstream news. FRPAA continues to move slowly through US legislation, but in the UK the mood is more active.

Funders

Beyond governments, independent funding bodies have driven the growth of OA. One of the most notable examples of these has been the Wellcome Trust in the UK, which for many years has required the following:

and has provided funds expressly to allow authors to pay APCs.⁵⁰ In 2011, however, the Trust moved in another direction – rejecting what they perceived as the slow movement of existing publishers to OA – and announced the launch of eLife.⁵¹ eLife is a joint initiative of the Wellcome Trust, the Howard Hughes Medical Institute and the Max Planck Society, and it marks the direct entrance of funding bodies into the publishing system. Fully OA, the journal will have no APCs initially,⁵² will pay peer reviewers and promises to be truly disruptive. Wellcome has justified the initiative on several fronts: that the publishing industry is not moving to OA quickly enough; that the spend required for eLife amounts to a tiny fraction of their overall budgets; and, most contentiously, that eLife is a means of reclaiming science for scientists, a perceived dig at Nature and Science.⁵³ Whatever anyone’s opinion of eLife (and its less than invigorating name), its progress will be fascinating to watch. All three funders are very powerful and have been able to generate mainstream press,⁵⁴ and the suspicion remains that eLife is too big to fail. The analogy that comes to mind is of the oligarch purchasing a football club and pumping endless money into it. The amount of money the funders bring to eLife dwarfs that available to most journals, and considering that there will be no APCs and that reviewers will be paid – luxuries not anticipated or received at most other journals – it would be strange not to expect eLife to have quite an impact. The journal published its first content late in 2012.

Is the hybrid journal dying?

Back in the middle of the last decade, most traditional publishers did not launch whole suites of fully OA journals. That was left to the likes of BMC, with the larger publishers going the hybrid route instead. Publishers might have had one or two fully OA titles, but for the rest of their list they would offer an OA option. In this scenario, the default option for the author is subscription access, but they could choose to pay a fee and publish OA. At the time, this hybrid option was seen as a good compromise – a way of offering OA within established journals – avoiding some of the problems of brand quality associated with start-ups (which most OA journals are). There was probably an expectation there that over time hybrid uptake would grow, but largely this hasn’t materialized, and notable problems presented by the hybrid approach have led to a certain discontent. Figure 9.8 shows the percentage of OA uptake at Oxford University Press for each discipline.

As can be seen from Figure 9.8, at OUP overall figures for hybrid OA have actually declined over the period. There are numerous factors that play into this, such as journals joining the initiative which have seen lower uptake than early adopters,⁵⁵ but, regardless, since 2009 there has been actual year-on-year decline even allowing for this affect. Hybrid uptake has also been low at other publishers, with Springer, Wiley and CUP all reporting uptake under 10 per cent across their lists (see Table 9.4). Nature bucked this trend with Nature Communications, which has seen uptake of over 40 per cent, but this has proved to be very much the exception.⁵⁷

There are real problems with hybrid which have led to dissatisfaction. For authors, publishing in a hybrid journal can be confusing: it is not as clear whether you can publish OA, there is a choice of licence, etc. For librarians and aggregators there are frequently problems if their federation systems cannot define access control at an article level. For funders, it is not always so clear whether their mandate has been fulfilled, and so the universities that have signed up to COPE (Compact for Open-Access Publishing Equity) specifically decline to support hybrid.⁵⁸ Finally, for publishers hybrid presents numerous difficulties, with perhaps the thorniest being the issue of ‘double dipping’. Quite rightly, publishers are expected by the library and OA community to discount subscription prices based on OA uptake within a hybrid journal. In the simplest terms, if a journal publishes 10 per cent of its content OA – i.e., free to anyone – those who have paid for a subscription want 10 per cent off its cost. Most publishers do this, but it is not that straightforward an equation for a number of reasons. Firstly, because prices are set and subscriptions are sold in advance, a publisher can’t discount a subscription until two years after the fact. Thus, a 2013 subscriber will derive the benefit of there being OA uptake in 2011. Secondly, a whole host of other factors affect subscription pricing, including – but not limited to – inflation, changes in page extent, changes in frequency, competitiveness with other titles on the market, and so on. So any OA discount can often be obfuscated in that overall price change, or at least can prove harder to explain. Throw in the fact that the vast majority of journal sales come through bundles, and you have a very complicated picture.

Consolidation

Back in 2010 Rhodri Jackson wrote an article for Serials E-News (now UKSG eNews) predicting the consolidation of the OA industry,⁵⁹ but to a large extent this still has not come to pass. It is a curious anomaly, for example, that the major subscription agents still have not entered the OA space. With APCs, we see a huge array of micro-payments being made by authors and their universities/funders, just crying out for an organization to centralize them. One early mover into this space (if you can call it ‘early’) has been a start-up called Open Access Key (OAK). Established in 2011, OAK describes itself as ‘a unique, new financial platform to manage, consolidate and process publication fees incurred in open access publishing’.⁶⁰ It’s an obvious move. Just as subscription agents such as Swets and EBSCO act as intermediaries between publishers and universities for the purchase of subscription products, OAK seeks to fill that role for the payment of APCs. And if we can anticipate that more and more APCs will be paid in the future, the OA ‘agent’ role is one which requires filling. It is notable that the Finch Report makes numerous allusions to simplifying both funding and payment.⁶¹ In that case Finch is discussing a multitude of factors – including universities providing clarity to their faculty on what is available to them; but Finch is also referring to the need for publishers to make payment mechanisms as easy as possible – and one of the easiest ways of doing this is to create industry standards. We can only assume that over the next few years the larger agents will step into the space currently occupied only by OAK (which hasn’t as yet happened) and simplify the process for those attempting to pay OA charges.⁶² Consolidation will not come in this one form only though – in some subject areas other initiatives are in hand.

SCOAP³

We’ve already mentioned the at times patchy development of OA across different subject areas, playing out through different levels of uptake of gold OA journals, and different levels of development of subject repositories, and so on. However, one area which has not been slow to embrace OA in different ways is High Energy Physics (HEP). arXiv remains one of the largest pre-print servers in the world, and now HEP has the Sponsoring Consortium for Open Access Publishing in Particle Physics (SCOAP³). SCOAP³ aims to do the following:

SCOAP³ is probably the most ambitious attempt in any subject area to invert the system. It has taken CERN (European Organization for Nuclear Research), which is driving the project, a period of several years of fundraising, campaigning and advocacy to reach the stage it is at now – where it has enough journals and funding on board to have felt confident enough to issue a tender for SCOAP³ funding for publishers to complete in early 2012. SCOAP³ is now expected to start in 2014. SCOAP³ is a hugely interesting initiative and its success or failure will directly motivate or demotivate potential replications in other areas. Theoretically, it sounds good – all journals in that field are OA, with no need for the complicated administration of individual author payments. There are potential problems, however – going beyond the initial complexities of establishing exactly which articles will be covered in journals that are not completely HEP, and so on. The risk for those journals covered completely by SCOAP³ funding is very high – for how long is the funding secured? All subscriptions will presumably drop off, so if SCOAP³’s funding disappears or decreases, those journals will find themselves starting from scratch. There is also the question of whether SCOAP³ is anti-competitive. One of the aims enshrined in the Finch Report is to build a more competitive journals market through the establishment of APCs which can presumably be determined better by author choice than the subscription prices before them. Ignoring any flaws in that ideal, SCOAP³ is due to determine a set of journals which will essentially become the gold standard in HEP – authors will be able to publish in those titles freely, with OA. It is hard to envisage too pleasant a future for those titles outside the SCOAP³ bubble – and it is also hard to see how those outside SCOAP³ will be able to get in during future tenders, or how any that drop out can survive.

Regardless of these concerns, SCOAP³, when it comes to pass, will be watched very closely.

Conclusion

Over the past five years OA has moved firmly into the mainstream of academic publishing, yet to some it still remains the bête noire of the industry. PLOS, BMC and others have shown that OA can be profitable, although probably less on a paper-by-paper basis than subscription journals, but there is no compelling evidence to suggest that OA works for the humanities or social sciences. There’s not much doubt, though, that fuelled by funder and governmental pressure, the percentage of OA material published will only increase in future years. It is to be hoped that this growth is managed sensibly and in a way that does not disrupt the existing ecology so greatly as to be detrimental. That is the vision espoused in the Finch Report, and it is commendable. It is important that the moral arguments for OA are set aside and the actual cost/benefit to universities, authors, readers, publishers and communities as a whole be considered. In particular, it would be beneficial if the sometimes extreme ignorance of the different demands of humanities and social science journals could be addressed. As with anything, there can be no one-size-fits-all approach.

The next few years promise to be fascinating; consolidation has already started – through PLOS ONE and the other mega journals, through SCOAP³, through OAK – and it will continue. Gazing into the crystal ball it is easy to see more mega journals, more cascades, more repositories, and a larger conglomeration of papers in huge sources, rather than the previously existing proliferation of titles. Publishers and learned societies will have difficult choices to make, as will universities navigating the new waters of the APC. It’s difficult to envisage anything other than a mixed model of funding for the next few years, but some areas will probably advance faster than others. Going back to where we started, OA may well be the future of the academic journal – but it is important that the process is managed so that it is a bright future, and not the end of the line.