27 Is Open Access good?
Open access appears to be a great initiative that acknowledges that everything should be free to view. Neither scientists nor the public that fund them should be barred from accessing the knowledge that is produced. What could be wrong with Open Access? Many people have written a great deal on what has gone wrong, and what I provide here (and elsewhere in this book) is not the last word as this is a dynamic topic that is changing all the time.
27.1 So what is open access?
I have covered the many different kinds of Open Access (OA) elsewhere in this book (see part II). In another chapter, I have covered what the Article Processing Charge (APC) needs to cover. Here I concentrate on the fiscal implication of the APC aspect of Open Access, someone needs to pay for the work done. Who should pay and how?
The principle of Open Access is something that is easy to appreciate, and lies very much at the heart at the movement from closed to open science. Some history is appropriate when we discuss Open Access, because the movement started as a genuine attempt to change the publishing model for the better (see key texts including those by Poynder, 2019, 2020). Hopefully in the future this experiment will be seen as a success. In the meantime, Open Access has been employed by publishers as a way in which they can earn large amounts of additional income. The story of PLOS ONE is useful to see how this change came about.
27.1.1 PLOS ONE
The Public Library of Science (PLOS) started soon after the turn of the new millennium, primarily as a response to the problems with peer review that are highlighted elsewhere in this book. This really was a new experiment in publishing, that was to radically alter the publishing model in the 21st century, and in only ten years, PLOS ONE became the world’s biggest journal (Davies, 2019). The idea was to do away with all the normal waiting times in publishing, and to have articles published immediately online, and without any charges associated with accessing them: Open Access. At the time, this was all fairly radical, and it became very successful with a large amount of interest from scientists who submitted their articles to PLOS journals. But the founders were frustrated that they were rejecting a lot of papers that were technically good, but not selected as they were not novel, or did not significantly advance their field. In 2006, PLOS dreamt up a new journal to take all of these technically sound manuscripts, and this was the birth of PLOS ONE.
From the outset, PLOS ONE was not interested in acquiring any Impact Factor. Instead, it was mostly interested in taking technically sound manuscripts irrespective of their results. This meant that reviewers for PLOS ONE were (and still are) asked to assess the technical soundness of a manuscript, and not to judge the ‘value’ of the results. The original idea was that reviews should be post publication, through comments made on the website after the publications were posted. The initial review then, was simply meant as a preliminary check for technical soundness. This was again a radical departure from the publishing norm by the PLOS group, when many other journals were still rejecting manuscripts that had no faults other than they were not attractive to editors. PLOS ONE was opened to accept all such manuscripts, effectively doing away with publication bias. As you might imagine, at a time when academics were under pressure to ‘publish or perish’ rejections were plentiful and so manuscripts quickly found their way to PLOS ONE. Because PLOS ONE was also Open Access, its papers received an extra boost of citations through visibility, and ironically after five years it found that it had an Impact Factor to rival some of the journals whose rejections it was picking up. This meant that more and more scientists started submitting to PLOS ONE as their journal of choice, mostly because it was publishing Open Access.
PLOS ONE became the world’s biggest scientific journal, and in 2013 it published 32058 papers (Davies, 2019). It’s worth taking a moment to do the arithmetic on the income that PLOS received for these at > USD 1000 per article (so it seems inconceivable that they could make a loss). By this time, other academic publishers had noticed the rocketing ascent of PLOS ONE and had responded by starting their own ‘no impact factor’ journals that could be fed by the rejection piles of journals already in their stables. Notable among these was Scientific Reports from the Nature Publishing Group which started in 2011 with 208 papers, and after 7 years ramping up to nearly 20000 papers in 2018. For whatever reason, the Impact Factor of Scientific Reports is roughly double that of PLOS ONE for around the same Article Processing Charge (APC), drawing far more researchers in. During this same period, since 2013, a great many such examples of ‘no impact factor’ journals that accept any article that is technically sound began to appear. They are not only the general behemoths like PLOS ONE and Scientific Reports, but you will find that even niche journals that make significant numbers of rejections of technically sound papers have their own version that they will offer to you in the same email that informs you of a rejection from their flagship journal.
The bottom line, from this story of PLOS ONE, the highly innovative journal with an attempt to fundamentally change the publishing landscape, is that this is exactly what it did. However, mainstream publishers saw a massive and untapped market and decided not to call these new entities ‘no impact factor’ journals, or even the ‘rejection pile’, but placed them under the Open Access banner.
As a result of PLOS ONE success, there was a bandwagon movement to Open Access. Now the scientists pay for making their own content open for anyone to read. They pay a one-off fee to the publishers to typeset the manuscript and host it on their site without a paywall. Prices start from USD 1000 and go up to around USD 12,000. Prices often increase with the Impact Factor of the journal (Gray, 2020; Mekonnen et al., 2021), although the costs involved to the publisher remain static. Actual costs of publishing a research paper in 2019 have been estimated as around USD 200 (see full breakdown of these costs in Part II) and USD 1000 (for the most prestigious journals with a rejection rate >90%) (Grossmann & Brembs, 2021).
27.2 So does that mean that OA journals are now free?
Mostly no. The majority OA model (hybrid OA) means that a minority of articles in these journals are free, but the universities are expected to subscribe to those same journals at ever increasing prices because much of the rest of the content is still behind the paywall. This is because most authors cannot afford to pay the fees charged by the journals (although some countries now have this payment as mandatory - cOAlition S, they and their scientists are still in a minority). There are some journals that are entirely open access (gold OA and diamond OA). These are (almost) exclusively online and have never been part of traditional packages that university libraries spend so much of their budget on. Hence the fact that they are entirely free does not impact library budgets. ‘Transformative publishing agreements’ are a new publishing model (Janicke Hinchliffe, 2019) discussed later in the book, which are estimated to cost up two to three times as much current models (Poynder, 2020). Hence paying for open access has not reduced the cost of access to scientific journals for libraries. This cost constantly goes up. For Hybrid OA we pay for much of this content not twice but thrice (Van Noorden, 2013; Buranyi, 2017)!
You, as an Early Career Researcher, are in one of the best positions to do something about this change from closed to open science. Currently, diamond OA journals are a very small and quite unusual component of the publishing scene. But it is totally possible for these to become the mainstream using initiatives like the overlay journal system. The current dynamic will only change when academics submitting their manuscripts change from for-profit publishers to other models. The power is literally in your choice of where to send your manuscript.
27.3 The wicked problem
Publishing has become very expensive for scientists and their funders. In parts of the world these costs are preventing some scientists from publishing where they want to. In richer nations, funders are now allocating increasing resources away from science and toward publishing. Publishing metrics are driving the hiring and promotion of scientists globally (Part IV), and competition for these are increasingly associated with fraud and misconduct. Choice of study subjects are leaning more towards publishing content with important metrics, than the acquisition of knowledge for the societies that are funding it. There is also the suggestion that some publishers have begun to capture the academic work-flow with a view to selling associated data (Brembs et al., 2021). We, as scientists, are allowing this to happen, and even paying for the privilege. Together with our institutions, funders and governments, we perpetuate the dominance of publishing in our scientific domain.
A wicked problem is one that is not only complex, but lacks clarity (with respect to solutions) or a way to scientifically test and study it. According to Rittel and Webber (1973), there are ten important characteristics, all of which are met by the ongoing situation in academic publishing. There are multiple stakeholders in our wicked problem, and they include our employers, our funders (political and societal), our peers as well as the gatekeepers and societies that police the publishing system. It is up to you to be aware of the options and become part of a solution that ultimately leads us to Open Science.
As scientists, we need remember our core concerns around the scientific method: rigour, independence, transparency and reproducibility. We need to persuade our institutions that our core concerns are what we should be measured by, and that the new range of Open Science tools be used to determine the extent to which we live up to our scientific values. Getting jobs, tenure and grants should be based on our Open Science credentials instead of Impact Factor and other publishing metrics.
27.3.1 Choosing open source for Open Science
These days ‘printing’ really means hosting electronic pdfs only, as there’s very little paper that’s actually printed. The layout from the manuscript (most often written in a word processing document) into a formatted pdf takes skill and talent. Today it is possible to use free software, like R Markdown (Xie, Allaire & Grolemund, 2018), to write papers that can quickly and easily be made into any sophisticated layout using LaTeX, the same language used by the publishers. Current models suggest that this way of formatting costs as little as USD 10 per paper (Grossmann & Brembs, 2021). Many journals allow submission of articles already formatted this way. Some publishers are buying these tools (e.g. Overleaf, Mendeley, Peerwith, Authorea, etc.) as part of the academic workflow that could be used as spyware (Brembs et al., 2021).
There are large costs associated with placing journals on platforms that allow for the dissemination, peer review and archiving essential for academic journals. Currently, there is no open source equivalent to a big publishing company that hosts hundreds or thousands of academic journals. But this does not mean that it is not possible. Once the investment has been made to set up such a platform, adding another 10 or 20 journals comes at practically no cost.
27.4 Making the change
My message, throughout this book, has been that the tools to make the shift from closed to open science are available to us now (Brembs et al., 2021). In my view, scholarly publishing is incompatible with all OA models except diamond OA, which necessitates a movement away from current publishers and back to the academic domain (see Fuchs & Sandoval, 2013; Brembs et al., 2021). There is nothing to stop us changing the system other than the need to act collectively. Indeed, there is an imperative to change the system as soon as we can to avoid academic capture.
Scholarly societies offer an excellent opportunity to organise. Some societies still dictate terms to publishers (for example, The British Ecological Society), and they have the power to finish their contracts and move their journals to new diamond Open Access formats, using tools such as overlay journals. In turn, they have to give up the income that their for-profit publishers supply. This does not mean that they would lose all income, they would still have membership funds and conferences. But there would be a loss of some of the comfort that they have become used to, making giving up hard to do. But societies need to carry out the wish of their members, and I would suggest that there are more members that would benefit from diamond Open Access to society journals than currently see any benefit in the payouts from for-profit publishers.