Open access for scholarly publications, part 7 – Publication costs

The National Library of Sweden has held the appropriation directive from the Swedish Government to produce recommendations for the change to an open access scholarly publication system. These recommendations are presented by Henrik Schmidt, librarian here at KIB, in a series of blog posts.

The cost for the scientific publication system have risen substantially in the last two decades. Even if the conditions of publication and the costs agreed upon between the different parties partially are, and have been, confidential, it is nonetheless a well-known fact that the costs have risen far above other consumer prices. Studies have pointed out the publishers’ gain a high percentage of profit and researchers have gathered to criticize the high cost of knowledge.

Now that we are about to undergo a transition from a subscription-based model to an open access model, the costs of publication will be distributed among various actors: e.g. authors, educational institutions and/or research funding agencies. Mapping projects, such as Open APC Sweden, help to clarify the picture regarding the cumulative costs for Swedish educational institutions, both as regards to the subscription fees and the cost of publications.

At KIB, we are trying to map the cumulative costs of scientific publication borne by Karolinska Institutet (KI). Naturally, we can follow and ledger the subscription costs. However, individual researchers who are paying publication fees, so-called article processing charges (APC), present greater difficulties. If one considers that the list of journals most frequently used for publication by KI researchers include Plos One, Scientific Reports, BMJ Open and Nature Communications, one understands that the money for the publishers goes both via us at KIB, but also via other parts of KI.

In an attempt to get a better overview of APC costs, KI introduced a specific account in Agresso in April 2016, known as number 5726 ("Cost of publication, Open Access"). SUHF recommends (only in Swedish) that all educational institutions do this. Unfortunately, as we show here (only in Swedish), the account is not always used, and it is sometimes incorrectly used. The roughly 15 million entered into the account in 2017 does not give an accurate picture of the total cost of OA-publication.

As a kind of amalgamate between open access and subscription-based publication systems, many publishers have developed a model whereby individual researchers can pay a publication fee to make their articles open and freely available, even if the journal is otherwise subscription-based. This is called hybrid publication. The number of scientific journals that "offer" hybrid publication has increased from 2.000, (2009) to closer to 10.000, (2016).

This phenomenon has arisen in part as a response to demands for open access publication by several research funding agencies. The publishers are paid twice (so called “double dipping”); both in the form of subscription fees and then in APCs. The publishing fee is on average higher in hybrid journals than in fully open access journals. Available data in Sweden (only in Swedish) suggests that the average cost of hybrid publication is 2.173 EUR, while publishing in fully open access journals is 1.298 EUR.

Both private and public research funding agencies struggle with the question of hybrid publication. The many research funding agencies which are behind the so-called Plan S (among them Swedish Forte, Formas and Vinnova), oppose hybrid publication. The European Comission's proposal for Horizon Europe, also makes clear that publishing costs are not covered in the case of hybrid journals. The same attitude can be found among universities, such as Harvard and Columbia in the US, and Imperial College London and the University of St Andrews in the UK. Grants issued through these universities do not include means for publishing in hybrid journals. 

It is of great significance that the transition to an open access publication system is implemented and stabilized without costs increasing uncontrollably. This is emphasized both by the European Commission’s recommendation from April 2018 and by National Library of Sweden (KB) in March 2019. KB estimates that the cost of making all publications open in a pure open access journals would have been approximately 330 million SEK in 2017 (where the main authors are affiliated to a Swedish research institution). This sum can be compared to the approximately 468 million paid in 2017 for both subscription costs and publishing costs in Sweden. 

The investigation conducted by KB was commissioned by the government and contains sixteen recommendations. The seventh is about the cost of publication.

Recommendation 7 reads as follows: In order to limit the costs of scientific publication, one or more relevant authorities should be given the responsibility for continuously following-up and analyzing the national costs of publication in the transition to an open access publication system.

Find out more about the description of and argument in favour of this recommendation in the report: Finansiering av omställningen från ett prenumerationsbaserat till ett öppet tillgängligt publiceringssystem (only in Swedish).

All blog posts in this series.

Henrik Schmidt

Librarian engaged in research support in various forms. Teaches doctoral students and researchers in areas related to literature search, publishing strategies and publication analysis.