Tag: article processing charge

An odd paper about India’s gold OA fees

A paper about open-access fees in India published recently in the journal Current Science has repeatedly surfaced in my networks over some problems with it. The paper is entitled ‘Publications in gold open access and article processing charge expenditure: evidence from Indian scholarly output’ and is authored by Raj Kishor Kampa, Manoj Kumar Sa, and Mallikarjun Dora of Berhampur University, the Indian Maritime University, and IIM Ahmedabad respectively. This is the paper’s abstract:

Article processing charges (APCs) ensure the financial viability of open access (OA) scholarly journals.The present study analyses the number of gold OA articles published in the Web of Science (WoS)-indexed journals by Indian researchers during 2020, including subject categories that account for the highest APC in India. Besides, it evaluates the amount of APC expenditure incurred in India. The findings of this study reveal that Indian researchers published 26,127 gold OA articles across all subjects in WoS-indexed journals in 2020. Researchers in the field of health and medical sciences paid the highest APC, amounting to $7 million, followed by life and earth sciences ($6.9 million), multidisciplinary ($4.9 million), and chemistry and materials science ($4.8 million). The study also reveals that Indian researchers paid an estimated $17 million as APC in 2020. Furthermore, 81% of APCs went to commercial publishers, viz. MDPI, Springer-Nature, Elsevier and Frontier Media. As there is a growing number of OA publications from India, we suggest having a central and state-level single-window option for funding in OA journals and backing the Plan S initiative for OA publishing in India.

It’s unclear what the point of the study is. First, it seems to have attempted a value-neutral assessment of how much scientists in India are paying as article processing charges (APCs) to have their papers published in gold OA journals. It concludes with some large, and frankly off-putting, numbers – a significant drain on the resources India has availed its scholars to conduct research – yet it proceeds to “suggest having a central and state-level single-window system” so scientists can continue to pay these fees with less hassle, and for the Indian government (presumably) to back the Plan S initiative.

As far as I know, India has declined to join the Plan S initiative; this is a good thing for the reasons enumerated here (written when India was considering joining the initiative), one of which is that it enabled the same thing the authors of the paper have asked for but on an international scale: allowing gold OA journals to hike their APCs knowing that (often tax-funded) research funders will pay the bills. This paper also marks the first time I’ve known anyone to try to estimate the APCs paid by Indian scientists and, once estimated, deem the figures not worthy of condemnation.

Funnily enough, while the paper doesn’t concern itself with taking a position on gold OA in its abstract or in the bulk of its arguments, it does contain the following statements:

“Although there is constant growth in OA publications, there is also a barrier to publishing in quality OA journals, especially the Gold and Hybrid OA, which levies APC for publications.”

“However, the high APC charges have been an issue for low-income and underdeveloped countries. In the global south, the APC is a real obstacle to publishing in high-quality OA journals”

“Extant literature reveals a constant increase in APC by most publishers like BioMed Central (BMC), Frontiers Media, Multidisciplinary Digital Publishing Institute (MDPI), and Hindawi”

“One of the ideas of open access was to make equitable access and check the rampant commercialization of scholarly publications. Still, surprisingly, many established publishers have positioned themselves in the OA landscape.”

“formulation of national-level OA policies in India is the need of the hours since OA is inevitable as everyone focuses on equity and access to scholarly communications.”

But these statements only render the paper’s conclusion all the more odd.

Of course, this is my view and the views of some scholars in India’s OA advocacy community and the authors of the Current Science paper are free to disagree. The second issue is objectively frustrating.

Unlike the products of science communication and science journalism, a scientific paper may simply present a survey of some numbers of interest to a part of the research community, but the Current Science paper falls short on this count as well. Specifically, not once does its body mention the words “discount” and “waiver” (or their variations), which is strange because OA journals regularly specify discounted APCs – or waive them altogether – if certain conditions are met (including, in the case of some journals, if a paper’s authors are from a low- and middle-income country). Accounting for discounts, researchers Moumita Koley (IISc Bengaluru) and Achal Agrawal (independent) estimated the authors could have overestimated Indian scientists’ APC expenses by 47.7% – ranging from 4.8% when submitting manuscripts to the PLoS journals to 428.3% when submitting to journals of the American Chemical Society.

Gold OA’s publishing fees are not in proportion to the amount of work and resources required to make a published paper open-access, and often extortionate, and that while discounts and waivers are available, they don’t spare research-funders in other parts of the world the expense, continue to maintain large profit margins at the expense of governments’ allocations for research, and – has scientist Karishma Kaushik wrote for The Hindu – the process of availing these concessions can be embarrassing to researchers.

Issue #1: the Current Science paper erects a flawed argument both in favour of and in opposition to APCs by potentially overestimating them! Issue #2: In their correspondence, Koley and Agrawal write:

“A possible reason for their error could be that DOAJ, which forms their primary source, does not mention discounts usually given to authors from lower-income countries. Another important error is that while the authors claim that they filtered the articles. Page 1058: ‘Extant literature suggests that the corresponding author most likely pays the APCs’. Following the corresponding author criterion, APC expenditure incurred by Indian researchers was estimated; they have not actually done so. Table 2 shows the discrepancy if one applies the filter. Also, Table 1 shows the estimated error in calculation if this criterion is included in calculation.”

To this, the authors of the Current Science paper responded thus:

“We wish to clarify any misunderstanding that may have arisen. We analysed the APC expenditure incurred in India without calculating the discounts or waivers received by authors as there is no specific single source to find all discounts, for example, an author-level or institute-level discount; hence, it would be difficult to provide an actual amount that Indian researchers spent on APC. Additionally, discounts or any publisher-provided waivers are recent developments, and discounts/waivers given to authors from LMIC countries were not mentioned in DOAJ, which is the primary source of the present study. Hence, it was not analysed in the current study. These factors may be considered as limitations of the study.”

This is such a blah exchange. To the accusation that the authors failed to account for discounts and waivers, the authors admit – not in their paper but in their response to a rebuttal – they didn’t, and that it’s a shortcoming. The authors also write that four publishers they identified as receiving 53% of APCs out of India – MDPI, Springer-Nature, Elsevier, and Frontiers Media – don’t offer “country-level discounts/waivers to authors” from LMICs and that this invalidates the concerns of Koley and Agrawal that APCs have been overestimated too much. However, they don’t address the following possibilities:

  1. The identification of these four publishers itself was founded on APC estimates that have been called into question;
  2. “Country-level” concessions aren’t the only kind of concessions; and
  3. The decision to downplay the extent of overestimation doesn’t account for the publishers that received the other 47% of the APCs.

It’s not clear, in sum, what value the Current Science paper claims to have, and perhaps this is a question better directed at Current Science itself, which published the original paper, two rebuttals – the second by Jitendra Narayan Dash of NISER Bhubaneswar – the authors’ unsatisfactory replies to them, and, since we’re on the topic, doesn’t seem to have edited the first correspondence before publishing it.

Preference for OA research by income group

Two researchers from Rwanda performed a “systematic computational analysis of the biomedical literature” and concluded in their paper that:

… papers with authors based in sub-Saharan Africa, papers with authors based in low income countries, and papers resulting from international collaboration are all much more likely to be made openly accessible than papers that don’t have these properties.

They analysed 547,404 papers indexed in PubMed, which is:

… a free resource developed and maintained by the National Center for Biotechnology Information (NCBI) at the National Library of Medicine (NLM). PubMed PubMed provides free access to MEDLINE, NLM’s database of citations and abstracts in the fields of medicine, nursing, dentistry, veterinary medicine, health care systems, and preclinical sciences.

Source

The researchers also found that after scientists from low-income countries, those in high-income countries exhibited the next highest preference for publishing in open-access (OA) journals and that scientists from lower and upper middle-income countries – such as India – came last. It is important to acknowledge here that while there exists a marked (inverse) correlation between GDP per capita and number of publications in OA journals, a causation might be harder to pin down because GDP figures are influenced by a large array of factors.

At the same time, given the strength of the correlation, their conclusion – about scientists from middle-income countries being associated with the fewest OA papers in their sample – seems curious. The article processing charge (APC) levied by some journals to make a paper openly accessible immediately after publishing is only marginally more affordable in middle-income countries than it is in low-income countries. However, the effects of technology and initiative seem to allay some of this confusion.

There are two popular ways, or routes, to publish OA papers. In the ‘gold’ route, the authors of a paper pay the APC to the journal, which in turn makes the paper openly accessible once it is published. A common example is PLOS One, whose APC is at the lower end, $1,595 (Rs 1.13 lakh). On the other hand Nature Communications charges a stunning EUR 4,290 (Rs 3.4 lakh) per paper for submissions from India. In the ‘green route’, the authors or publishers upload the paper to a publicly accessible repository apart from formally publishing it; common example: the arXiv preprints server, which is moderated by volunteers.

There is also ‘hybrid’ OA, whereby a part of the journal’s contents are openly available and the rest is behind a paywall. In one review published in February 2018, researchers also pointed out a ‘bronze’ route: “articles made free-to-read on the publisher website” but “without an explicit [OA] license”.

The authors of the current paper reason that researchers from high-income countries might be ranking higher in their preference for OA papers because the “‘green’ route of OA has been encouraged by an enormous growth in the number of OA repositories, particularly in Europe and North America”; they also note that Africa was home to only 4% of such repositories in 2018. In the same vein, they continue, “the vast majority of funding organizations with OA policies as of 2018 were based in Europe and North America, with less than 3% of total OA policies originating from organizations based in Africa”.

Additionally, many journals frequently waive APCs for submissions from authors in low-income countries, whereas those from lower- and upper-middle income countries – again, including India – do not qualify as frequently to have their papers published without a fee. A very conservative, back-of-the-envelope estimate suggests India spends at least Rs 600 crore every year as APCs.

It was to reduce this burden that K. VijayRaghavan, the principal scientific adviser to the Government of India, announced earlier this year that India was joining the Plan S coalition of research-funders, which aims to have all research funded by them openly accessible to the public by 2021. As a result, researchers funded by Plan S members will have to submit to journals that offer gold/green routes and/or journals will have to make exceptions for publishing research funded by Plan S members.

This is going to take a bit of hammering out because the Plan S concept has many problems. Perhaps the most frustrating among them is its Eurocentric priorities. Other commentators have acknowledged that this limits Plan S’s ability to serve meaningfully the interests of researchers from South/Southeast Asia, Africa and Latin America. In July, two Argentinian researchers lambasted just this aspect and accused Plan S of ignoring “the reality of Latin America”. They wrote that Plan S views “scientific publishing and scholarly publications … as a commodity prone to commercialization” whereas in Latin America, they “are conceived as the community sharing of public goods”.

The latter is more in line with the interests of the developing world as well as with the spirit of knowledge-sharing more generally. At present, a little over 50% of research articles are not openly accessible, although this is changing thanks to the increasing recognition of OA’s merits, including the debatable citation advantage. Research-funders devised Plan S to “accelerate this transition”, as Jon Tennant wrote, but its implementation guidelines need tweaking.

Another problem with Plan S is that it keeps the focus on the ‘gold’ OA route and does little to address many researchers’ bias against less prestigious, but no less credible, journals. For example, while Plan S specifies that it will have gold-OA journals cap their APCs, scientists have said that this would be unenforceable. So, as I wrote in February:

… if Plan S has to work, researcher-funders also have to help reform scientists’ and administrators’ attitude towards notions like prestige. A top-down mandate to publish only in certain journals won’t work if the institutions aren’t equipped, for example, to evaluate research based on factors other than ‘prestige’.

To this end, the study by the researchers in Rwanda offers a useful suggestion: that the presence or absence of policies might not be the real problem.

There was no clear relationship between the number of open access policies in a region and the percentage of open access publications in that region. … The finding that open access publication rates are highest in sub-Saharan Africa and low income countries suggests that factors other than open access policy strongly influence authors’ decisions to make their work openly accessible.