Editorial: Where is scientific publishing heading?

Many of us researchers probably do not spend much time reflecting on the structure of, by far the most important common infrastructure for science, namely scientific publishing. Quite naturally, we have our focus on traditional academic activities such as teaching, mentoring graduate students and post docs, and the next exciting experiments that will allow us to advance our understanding of what we are studying. Even though we daily depend on reading research presented by our colleagues in scientific papers and depend on the same journals to handle our manuscripts that present results from our research. We surely all agree that it is extremely important for our work, and it is therefore to some extent strange that we as scientists have lost control of the majority of this infrastructure, which today is dominantly controlled by a few large multinational companies. Traditionally scientific publishing was controlled by learned societies such as Royal Society and National Academy of Science in USA and publishers associated with key universities, e.g. Oxford University Press and MIT Press (that started to publish Cell), but just like multinational companies such as Sigma-Aldrich, Roche and Agilent have evolved through mergers and acquisition to dominate the provision of chemicals, research equipment, and various services to researchers, the publication of scientific results in peer-reviewed journals has evolved to become a highly consolidated and very profitable industry controlled by for-profit companies. Thus, the three largest publishers Elsevier, Springer and Wiley-Blackwell now represent about half of the 10 billion GDP scientific publication industry, with Elsevier being by far the largest with a 24% market share and the two others having a market share of about 12% (https://www.theguardian.com/science/2017/jun/27/profitable-business-scientific-publishing-bad-for-science) (I strongly recommend reading this excellent article in the Guardian describing in more detail the history of the scientific publication industry). This dominance has followed years of consolidation in the industry, a process that is still ongoing as illustrated by the recent merger of Nature Publishing Group and Springer, resulting in formation of what is now called Springer Nature having an annual revenue of about 1.5 billion EUR. Not only does the scientific publication industry have large revenues, it is also extremely profitable. Thus, in 2010 Elsevier reported profits of 724 million GBP corresponding to a margin of 36%. These are margins that not even pharmaceutical companies can compete with and higher than Apple, Google and Amazon reported that year (https://www.theguardian.com/science/2017/jun/27/profitable-business-scientific-publishing-bad-for-science). Even though typical journals have a 12%–15% profit margin (https://www.theguardian.com/science/2017/jun/27/profitable-business-scientific-publishing-bad-for-science), we as scientists contribute significantly with work for free! For many journals, scientists work as editors and handle manuscripts for free (or a small annual fee) and we all contribute to the success of journals by serving as reviewers.

The scientific publication industry is, however, undergoing dramatic changes these years. Thus, there is continuous expansion in the number of journals that compete for submission of papers. This is clearly manifested by the large number of invitations to submit papers, even in some cases personalized with reference to some of our previous work, that we all receive in our inbox daily. Still, most journals are the traditional type offered to scientists via library subscriptions, but the publishers are constantly increasing the subscription rate of these resulting in an increasing constraint on the library budgets. However, most of the new journals are so-called open access journals, where the authors are paying for a publication fee to have their paper published. Following acceptance (or after an embargo period), the paper is then made freely available for all. The idea of open access journals is appealing for many reasons, and it started as an opposition against the traditional publication industry. Probably the first serious initiative of open access publishing was made by BioMed Central (BMC). BMC launched its first journals, Genome Biology and Journal of Biology, in 1999 as open access journals. BMC was very successful in rapidly expanding with other journals, resulting in now more than 180 journals (many not with the BMC brand, but using their platform, e.g. Microbial Cell Factories). This expansion was followed by the establishment of the not-for-profit Public Library of Science, that first launched PLOS Biology in 2003 and later other journals such as PLOS Genetics and PLOS Computational Biology, and who received 10 million USD in start-up funding from the Gordon and Betty Moore Foundation for establishment. PLOS was launched by Patrick Brown, a biochemist at Stanford, and Michael Eisen, a computer scientist at UC Berkeley and at Lawrence Berkeley National Laboratory, and as the name indicates it was established as an initiative to provide researchers with free access to published papers. In 2007, Frontiers Media SA launched their first journal and they now have published more than 65 000 papers in their various journals. PLOS is run by non-for-profit organizations, whereas BMC was a for-profit organization that in 2008 was acquired by Springer. Frontiers started as a non-for-profit organization with the objective to have no (or little) costs for authors, but rapidly turned into a for-profit organization and in 2013 it received significant financing from Holtzbrinck Publishing Group, the then owner of Nature Publishing Group, which at this time took an equity share in the company. Publication fees for papers in the Frontiers journals now vary between 1100 and 2500 USD/paper.

It is, however, expensive to run high-impact specialty journals like the initial set of PLOS journals, where there is a full-time editorial staff assisting the editors. Even the standard publication fee of 2000–3000 USD for papers to be published in a PLOS journal does not cover all the costs due to the relatively low number of papers published each year, and even though editors and reviewers still work for free. PLOS has overcome this problem by launching the generic journal PLOS ONE in 2006, which has become a large success. PLOS ONE was an interesting addition to scientific publishing as a requirement for acceptance is based only on technical correctness, and following its launch the submission rate increased dramatically reaching its peak of more than 30 000 papers published annually in 2013 and 2014. Many papers presenting exciting new results were also sent to PLOS ONE, as many authors liked the fast and transparent review process, and its journal impact factor (JIF) was therefore initially relatively high (about 4.4 in 2009 and 2010), but following the surge in submissions the JIF dropped below 3 in 2016. This drop has been associated with a decrease in the number of published papers, which in 2016 was about 22 000. However, at this level and a publication fee of 1495 USD/paper PLOS ONE is still providing PLOS with an annual revenue of about 33 million USD, which assist with financing the other PLOS journals and hereby has made PLOS independent of subsidies from foundations.

The success of PLOS ONE probably inspired Nature Publishing Group to launch Scientific Reports, which has the same concept of solely evaluating papers based on technical correctness and not a strict novelty criteria as for most journals and at a relatively low publication fee (currently 1675 USD/paper). With a JIF > 5 in 2013, where the journal had published <1000 papers, there was a surge of submissions to this journal. Thus, in 2016 more than 20 000 papers were published in Scientific Reports (up from about 10 000 in 2015). This rapid increase in number of papers has, however, coincided with a decrease of the JIF to about 4.2 in 2016 from an all-time high of 5.6 in 2014. The success of open access journals also led Nature Publishing Group to launch Nature Communications as a high-end journal publishing papers in all fields, but which benefits much from taking an outlet of high-quality papers being rejected following peer review by other Nature journals (Science recently initiated Science Advances as a competitor to Nature Communications). This journal has become a large success, having a JIF above 12 and it published >3500 papers in 2016, which has now surpassed the number of papers published by PNAS (about 3200 papers in 2016). With such a volume of papers, it is still probably possible to finance the costs for full-time editors working on the journal, as illustrated by the fact that the journal recently switched from a subscription model (combined with the requirement for payment of publication fees) to a full-open access mode (the publication fee is 5200 USD/paper in Nature Communications). The importance of volume in keeping publication fees at a reasonable level for high-end journals is in line with statements from Diane Sullenberger, Executive Editor ofPNAS, that if PNAS had to change from a subscription model to an open access model they would have to charge 3700 USD/paper (van Noorden 2013). For journals that have editorial teams working on not only handling papers, but also providing a lot of so-called front matter, e.g. News and Views, Perspectives, Research Highlights, Book Reviews, and News in Focus, as seen in Nature and Science (and many other Nature journals), it would, however, be difficult to cover all these costs if there is switch to an open access model. Thus, Peter Campbell, Editor in Chief of Nature noted: ‘if Nature was to make such a shift it would result in publication fees of 20–30 000 GBP/paper’ (van Noorden 2013). If the industry will eventually turn into a completely open access model, we as scientists are therefore likely going to lose the various news coverages, perspectives, etc. provided by many of the high-end journals such as Science, Nature and various Nature journals.

With this rapid development of open access journals, it seems like the publication industry is going in the direction of open access and away from the traditional library subscription model, and as pointed out in a recent analysis in Nature, open access journals have lower costs as they can reduce costs for marketing, rights management, online user managements, etc. and therefore do provide some benefits for the scientific community (van Noorden 2013). This move toward open access journals is supported by many funders who today require that research papers are made open access. Furthermore, the European University Association recently published a document entitled ‘Towards Full Open Access in 2020’ where they recommend all their member universities to install policies that ensure a reduction in publication costs, that authors maintain all publication rights (as is generally the case in open access journals), and that all research papers are open access (http://www.eua.be/Libraries/publications-homepage-list/towards-full-open-access-in-2020-aims-and-recommendations-for-university-leaders-and-national-rectors-conferences). Even though many library subscription journals offer an open access opportunity (the so-called hybrid model), it will become problematic for the journals to maintain income from library subscriptions if too many journal papers become open access. This development may exert additional pressure on the library subscription model in the future. However, for the scientific community open access publishing can also have disadvantages. Some research groups with small research budgets may have difficulties in paying the publication fee. Even though some publishers, e.g. PLOS, offer to waive the publication fee based on application, this may make it more difficult in the future for researchers with smaller research budgets to publish in respectable journals. Concern has also been raised that switching to complete electronic publishing may result in narrowing science and scholarship, as researchers will primarily look into highly downloaded and highly cited papers which will bring researchers in touch with prevailing opinions and thereby narrow the range of findings on which future research ideas are built upon (Evans 2008).

A deplorable development that followed the introduction of open access journals is that there are a lot of so-called open access predatory journals. The worst of these are fake journals that simply request submission for a charge and then never publish the paper. Many others have a fake editorial process with no real peer review. The seriousness of this problem was well illustrated in a study published in Science earlier this year (Sorokowski, Kulczycki and Sorokowska 2017). The authors created the identity of a ‘fake scientist’. Impersonating this fabricated professional, they sent a fictitious CV to several open access journals, some even not matching the research area of the ‘fake scientist’ and requested to be enrolled as an editor. Many of the journals welcomed the ‘fake scientist’ without any further quality check, but requested that the ‘fake scientist’ ensured submission of new papers, e.g. in connection with a conference organized by the ‘fake scientist’, and in some cases even proposed to split the revenue generated (Sorokowski, Kulczycki and Sorokowska 2017). This paper in Science clearly indicates the fact scientific publishing has turned into an opportunity to make money in a criminal way, and it shows that we as scientists have to be careful where we are sending our papers. As a warning against these predatory journals, Jeffrey Beall, a librarian at University of Colorado, Denver, established a list of predatory journals that was regularly updated based on input from scientists (http://beallslist.weebly.com) (the list does unfortunately not seem to be updated since 2016). This list contains surprisingly many publishers, and it even at some time included Frontiers due to several critical voices raised over the editorial policy of some of the Frontiers journals.

In this landscape of a continuously increasing number of journals, some serious, some predatory, it is becoming harder for our researchers to choose where to publish our papers. On his webpage, Jeffrey Beall offers a quick guidance (http://beallslist.weebly.com) that basically states that we should check the following: who is publishing the paper (I will get back to this at the end), are the editors and/or members of the editorial boards of the journal well-respected scientists, does the journal have a clear editorial policy, are publication fees clearly stated, is the journal indexed in e.g. PubMed, and does the journal publish other papers that are presenting similar type of work as the one you are presenting in your paper. Here I would particularly like to emphasize the role of editors (and the associated editorial board), as they are the ones who are responsible for handling your paper, i.e. deciding whether your paper will be sent out for peer review and how the results of the peer-review process will influence decision whether to publish or not. The document referred to above from the European University Association (http://www.eua.be/Libraries/publications-homepage-list/towards-full-open-access-in-2020-aims-and-recommendations-for-university-leaders-and-national-rectors-conferences) also mentions the necessity of universities to check carefully the editorial boards of journals in order to evaluate their quality. As mentioned by Randy Schekman, Nobel Laureate, Professor at UC Berkeley, and founder of the open access journal eLife, the editors of the top-level journals basically decide on the career of many young researchers who say ‘I will not get a job if I do not have a CNS’ (Cell, Nature or Science) (https://www.theguardian.com/science/2017/jun/27/profitable-business-scientific-publishing-bad-for-science). With this important role of editors, it is indeed a ‘A Glaring Paradox’ that we leave this responsibility to inactive researchers in many of the top-level journals, such as the Nature and Cell journals, as pointed out by Mark Johnston, a yeast geneticist at University of Colorado, Denver and Editor-in-Chief of Genetics (a journal published by the Genetics Society of America) (Johnston 2015). He did an interesting analysis of the editorial board of several journals and found that the average H-index for the editors, what he referred to as the Journal Authority Index (JAF), is very low for Nature and Cell journals having high JIFs, whereas the JAF is high for society/academy-driven journals such as Genetics, Science, Molecular Biology of the Cell, Molecular and Cellular Biology and PNAS. However, most of these journals have much lower JIFs than their counterpart Nature and Cell journals.

This brings me to my last point, namely whether JIF is a good measure of a journals quality or not? The analysis performed by Mark Johnston indicates that it is not. To illustrate this further, I also often refer to the Journal of Biological Chemistry, a highly respectable journal publishing a large number of papers annually (even though this has been dropping from more than 6000 papers in 2004 to about 2000 papers in 2016), which has seen a decline in JIF from above 6 in 2004 to around 4 in 2016. Has this decline reduced the quality of the papers in this journal? Not as far as I know—this journal still publishes really high-quality papers! The poor measure of quality by the JIF is also illustrated by the fact that this measure is often driven by a few highly cited papers. When looking at the citations to each paper in a journal, they all have a distribution function with a very long tail (this also holds for FEMS Yeast Research), and about 70% of papers published in a journal (including Nature and Science) are cited less than the JIF (Callaway 2016). Furthermore, all journals (also Nature and Science) have papers that are not cited at all in the so-called citation window used to calculate the JIF. These observations have resulted in several societies turning against the JIF and to removal of JIF references on their journal web pages. The American Society for Microbiology (ASM) is a prominent proponent for this policy. ASM publishes several key journals in the microbiology space—traditional library subscription journals as well as open access journals. More publishers will likely follow this trend. However, the JIF is still going to be used in the scientific community, and in some cases is even used for deciding on hiring and promotion of faculty (particular in China there is much focus on the JIF and there are even paid bonuses for publishing in high-impact journals like CNS). However, hopefully we as scientists can assist in bringing some sense back into these evaluations by using more than the JIF for judging the quality of a journal. Here a journal's JAF, as introduced by Mark Johnston, may indeed be a very good additional measure of quality as this illustrates that highly cited researchers are spending time on evaluating your papers when they are being submitted to the journal. A counter argument can, however, well be that the full-time editors of Cell and Nature journals have more time to perform a careful evaluation of your paper, compared with very active researchers that are busy with many other tasks. But still, my recommendation is to always take a careful evaluation of the list of editors who are going to handle your manuscript before choosing the journal you are submitting to.

Let me end with the question of the title: where is scientific publishing heading? As Niels Bohr said ‘prediction is very difficult, especially about the future’, so of course it is hard to answer this question. However, even though papers published in open access only accounted for about 15% in 2010 (van Noorden 2013), there is a clear trend towards open access publishing. I think it is likely that the traditional library subscription model will eventually disappear, which will result in a significant disruption of the market. Maybe this will be good for our scientists and the society at large as it may result in overall lower costs. In this process, it is, however, extremely important that we as scientists make sure that we provide support for publishers that are driven by non-for-profit organizations, either in the form of learned societies, academies or other organizations (like PLOS and Annual Reviews, another non-for-profit publisher that started in 1932 with Annual Reviews in Biochemistry and now publishes 46 review series), who have a clear objective of supporting the scientific community. So besides checking on the list of editors, I encourage you to submit your papers to journals that have an association with academies such as PNAS (NAS), EMBO journals, Royal Society journals or learned societies such as Science (AAAS), FEBS journals, FEMS journals, GSA journals and ASM journals. We as scientists have an interest that journals published by these societies survive, and this is far more important than having a discussion of open access versus non-open access. Why publish in a journal, where the profit is going to the investors behind it, when you can send the same paper to a society journal where the profit may go to a scholarship for your next post doc or to a scholarship for your PhD student to join an international conference?

FEMS Yeast Research belongs to this last category of journals. In our journal, all papers are handled by scientific leaders of the yeast research community (an estimated JAF for FEMS Yeast Research is well above 40). It is free to publish, but with an open access option. If you do not choose open access you are still guaranteed that your paper is widely accessible as all FEMS journals are available through a very large number of libraries in the world, including in developing countries. Besides this, we aim to be the yeast community journal (Nielsen 2015) with publication of not only high-quality research papers, but also minireviews, which are often collected in thematic issues that allow you to get an excellent overview of a whole research topic. We have also recently added retrospectives of many of the leaders of the yeast research community as well as publishing interviews with prominent yeast researchers. The yeast research community has traditionally been very strong (Nielsen 2014), but in order to maintain this community it is important you as researchers continue to support FEMS Yeast Research, so we can support you back with sponsoring of research fellowships and conferences. I therefore encourage all editors, reviewers, authors and readers of the journal to share your thoughts on the journal policy, either through engagement with the editors or through commentaries to this editorial.

REFERENCES