Heather Morrison is an Associate Professor at the University of Ottawa’s School of Information Studies. Heather’s research focuses on scholarly communication and open access, and her teaching and advocacy includes information policy. She is a well-known advocate of open access and open approaches to research, which the Sustaining the Knowledge Commons project is intended to model.
Tanoh Laurent Kakou and Heather Morrison
Brief abstract (English)
This post presents highlights of preliminary results of our longitudinal study of the article processing charges (APCs) of MDPI, an open access publisher using the APC business model (exclusively), with new journals offering « free for now » publication. Thanks to Solomon and Björk (2012) we have data for a sample of 25 of MDPI’s journals that did charge APCs in 2011 (non-charging journals were excluded from their study). The APCs for these 25 journals ranged from 300 CHF to 1,600 CHF, with an average of 624 CHF. Only 1 journal (Algorithms) did not change in price ; all others increased in price from 2011 – 2016. The average APC for this group of journals in 2016 was 1,148 CHF, an average increase of 524 CHF or an 84% increase in 5 years. This compares with a compound U.S. interest rate from 2010 – 2015 of 8.7% as calculated by Sara Wheatley https://sustainingknowledgecommons.org/2016/04/13/comparison-of-biomed-central-apcs-from-2010-2016/. The EU compound interest rate would have been lower during this time frame, with some years in negative growth. One might say that MDPI’s APCs for established journals have increased by an order of magnitude greater than the overall inflation rate in the past 5 years.
As of March 2016, MDPI listed 155 journals on its website. Of these, nearly half (72 journals) are <>, that is they use the APC model but are not yet charging. The average APC of 359 CHF (662 CHF when non-charging journals are excluded) contrasts with the 1,148 average APC of the 25 journals that were included in the 2011 sample (established journals). From 2014 to 2015 APCs either stayed the same or decreased ; from 2015 to 2016, we see only stable or increasing prices with an average price increase from 2015 to 2016 of 18% (60 CHF increase on a 2015 average APC of 334 CHF).
This case study illustrates one approach to the evolution of a new commercial publisher committed to the APC business model, with new journals offering free publishing until journals are established. APC payers need to take this into account for budgeting purposes; journals that today either are free or have very low APCs may have substantial APCs in a few year’s time.
Following are the original abstract (in French) by data analyst Tanoh Laurence Kakou and a table with the APCs of the 25 journals for which we have APC data for 2011 and 2016.
Nous analysons des données de MDPI APC de 2011 à 2016. Notre recherche consiste à comprendre comment les frais de publication de MDPI ont évolué dans cette période. Nous avons analysé d’abord les données de chaque période. D’abord 2011, 2014, 2015, puis 2016. Puis nous avons comparé l’évolution des frais de 2011 à 2016, de 2014 à 2015 et de 2015 à 2016.
Les données de 2011 sont des prélèvements d’un échantillon effectué en 2011 par Solomon et Björk (2012) sur uniquement 25 revues avec des frais de publication. Ainsi, la moyenne des frais est 624 CHF, la médiane est de 500 CHF et le mode est de 300 CHF. Sensors a les frais de publication les plus élevés (1,600 CHF). Tandis que Molbank détient les moins élevés (200 CHF). En 2016, on a obtenu une moyenne de 345 CHF, une médiane de 300 CHF et un mode de 0 CHF. Sensors, Marine Drugs et Molecules ont les frais (1,800 CHF) les plus élevés. En ce qui a trait à l’évolution des frais entre 2011 et 2016, on obtient une moyenne de 115%, d’une médiane de 75% et d’un mode de 167%. S’agissant de l’évolution du nombre du montant des frais, nous obtenons une moyenne de 523 CHF, d’une médiane de 500 CHF et d’un mode de 200 CHF. Seule les frais de Algorithms n’ont pas changé. Ils sont restés à 300 CHF soit une évolution de 0%. Tandis que Polymers augmente de 367% qui représentent une valeur ajoutée de 1100 CHF.
Concernant les données de 2014 de Morrison et al, (2015) elles ont été effectuées sur toutes les 124 revues au site web de MDPI. On obtient dans cet échantillon, une moyenne de 372 CHF, d’une médiane de 300 CHF et d’un mode 0 CHF. 3 revues : Sensors, Marine Drugs et Molecules détiennent les frais les plus élevés. En 2015, Morrison et al. (2016) ont étudié 141 revues. 70 revues n’avaient pas de frais de publication. Sensors, Marine Drugs et Molecules gardent les mêmes frais (1800 CHF). Par rapport à l’évolution des frais de publication de 2014 à 2015, deux tendances s’observent.
Au niveau de l’évolution des frais des 124 revues de 2014 à 2015, les frais de 84 revues n’ont pas changé. 40 revues ont baissé leurs frais.
En mars 2016, nous avons trouvé 155 revues au site web de MDPI. Presque la moitié (72 revues) sont <>. La moyenne FDP est 359 CHF (662 CHF si on exclut les revues sans frais) la médiane 300 CHF (500 CHF si on exclut les revues sans frais et le mode 0 (300 CHF si on exclut les revues sans frais).
Au niveau de l’évolution 2015 et 2016, la moyenne augmentation est 18%, (60 CHF), une médiane de 0 et un mode de 0.
L’analyse des données de MDPI APC de 2011 à 2016 montre que les frais de publication de cette période ont augmenté en moyenne 84%, allant jusqu’ à 367% pour la revue Polymers. Cependant entre 2014 et 2015, on a constaté une stabilité des frais et même une baisse de 100% de 6 revues : Sports, Systems, Technologies, Toxics, Universe, Veterinary Sciences. Entre 2015 et 2016, on a remarqué en augmentation en moyenne de 18% (60 CHF).
|JOURNAL NAME||APC 2011 (CHF)||APC 2016 (CHF)||Change 2016 – 2011|
|International Journal of Environmental Research and Public Health||1,000||1,600||600|
|International Journal of Molecular Sciences||1,400||1,600||200|
Solomon, D.J. & Björk, B.-C. (2012) A study of open access journals using article processing charges. J. Am. Soc. Inf. Sci. Technol. 2012, 63, pp. 1485–1495. Retrieved from http://www.openaccesspublishing.org/apc2/preprint.pdf February 2, 2016.
Sustaining the Knowledge Commons (SKC) is pleased to offer 1 – 2 open access research oriented 2018 summer internships for international undergraduate students through the competitive Mitacs Globalinks Research Internship program. The call for 2018 applications is now open. Following is a brief description of the Mitacs Globalink Research Internship and the SKC internship project.
The Mitacs Globalink Research Internship
From the Mitacs website: “The Mitacs Globalink Research Internship is a competitive initiative for international undergraduates from Australia, Brazil, China, France, India, Germany, Mexico, Saudi Arabia, Tunisia, and Ukraine. From May to September of each year, top-ranked applicants participate in a 12-week research internship under the supervision of Canadian university faculty members in a variety of academic disciplines, from science, engineering and mathematics to the humanities and social sciences.
Mitacs Globalink Research Internships are currently available at over 45 universities across Canada. For more information, please click on “Information for Students” to see the list of universities hosting Globalink interns in 2017.
Globalink Research Internship alumni interested in returning to Canada for graduate studies are eligible for the Globalink Graduate Fellowship.
For further information or to submit an application, go to:
The Sustaining the Knowledge Commons internership: details
The central idea driving this research is that humanity collectively shares all of our knowledge. Everyone should thus be able to access scholarly works free of charge and everyone capable should be welcome to contribute to the knowledge commons. We study alternative economic approaches, for example the commons and the gift. We do quantitative research on costs related to producing a scholarly article. We also do qualitative research, talking with the directors of scholarly journals about the resources required for journals to be open access. The approach is one of open research; to learn more: sustainingknowledgecommons.org
There is a broad consensus among researchers, librarians, research funding agencies, non-governmental organizations and even most scholarly publishers that today the optimum method of sharing the results of research is open access, that is, articles, books and data that are freely available through the internet. The dilemma is how to transform a system based on demand (subscriptions and purchases) to a system based on production in order to sustain free access for the reader. Sustaining the knowledge commons targets the means of transformation by studying the models used by open access scholarly journals (for example, article processing fees or subsidies).
Thanks to a generous grant from the Social Sciences and Humanities Research Council, we are a team in the process of growth consisting of the principal investigator and several students, most of whom are at the doctoral level. We are conducting a longitudinal study using an innovative open research approach. We publish our data and results as part of the work process. The idea is to facilitate research in this area (other researchers use our data, for example) and to speed up knowledge translation, because there are organizations that want to accelerate the transformation process by using their funds, like the signatories of OA 2020 http://oa2020.org/
Preferred: knowledge of open access and/or open science, experience with academic writing, social media, research processes, teamwork. Attention to detail (e.g. research data gathering and management, analysis of large datasets with tools like the open source data statistical package JASP or PSPP) is desirable. French / English bilingualism, and/or other languages, would be an asset.
(was 12 Copernicus journals tripled or quadrupled page charges in 2016)
Update July 7, 2016: thanks to Copernicus’ Silke Hartmann (S.H.) for adding a substantive comment. Following is a brief data update about 2 journals formerly free of charge that now charge APCs, clarification about my category “cost not specified”, and highlights worth repeating of S.H.’s clarification about the meaning of “currently waived”.
Data update: in brief, two journals with conflicting information about APPCs as of June 15, 2016 now post APPC details on their webpages.
Earth Surface Dynamics (ESurf) and Geoscientific Instrumentation, Methods and Data Systems now charge from €50 to €120 per printed page as of July 1, 2016.
Both journals posted conflicting information about APCs as of June 15, 2016. Their APC pages stated: “If the discussion paper is published after 1 January 2016 the APCs are levied for the publication of the final revised paper in the journal. The publication in the discussion forum is free of charge”. Under the price per journal page, the cost is indicated as “currently waived”. The June 15 version of the ESurf APC page is attached here in order to illustrate: ESurf – Article processing charges 2016 06 15
This is a good illustration of why I have a category called “cost not specified” and how this is used. The journal APC page clearly indicates that “APCs are levied [emphasis added]” along with conflicting information that APCs are “currently waived”. It would be consistent with this explanation to provide free publication to authors, or to inform them that fees are now in place at some point between submission and publication. I do not mean to suggest that these journals have any intention of deceiving authors, merely to point out that this is conflicting information and authors and payers reading the website information do not have a clear indication as to whether or not there is a charge. Two other journals S.H. indicates have appropriate APC information continue to post this conflicting information, that APCs are levied but currently waived.
Highlights from S.H. of Copernicus’ comments: “Currently waived” journals either plan to institute APCs or that have not yet decided whether to use this model or not and wish to keep their options open. “Cost not found” in the case of Copernicus journals (a category I use when I cannot find any information indicating whether there is or is not a cost) means that publishing in the journal is free of charge [Earth System Science Data and Soil].
Data will be updated before it is included in the main spreadsheet (to be completed and released later this year).
Update July 6, 2016: I retract my statement on tripling of page charges thanks to new evidence indicating that the difference reflects a change in the stage at which papers are assessed (now final publication stage generally one-third the pages of discussion paper stage). The change is intended to be revenue-neutral but more data from APC payers would be needed to confirm this. 2015 data from the Open APC project includes values for 2 journals with papers at both stages, and prices paid are 17-24% higher for papers at the final publication stage. I re-affirm my assessment of the volatility of the APC market. I found 6 journals with APCs indicated “currently waived”, presumably journals that will charge APCs of unknown quantities in future. I found 4 journals that referred to APCs without specifying the cost and 2 journals with no indication of whether or not there is a cost. This is a very substantial percentage of Copernicus’ journals for which the answer to the questions “is there an APC or APPC, and if so, how much is it?” is not available on the Copernicus’ website. I regard Copernicus as a model OA publisher. It is likely that this situation reflects journals that would rather not charge APCs, feel they must charge APCs but are not sure how much to charge, etc., rather than deliberate obfuscation. New evidence and the original post follow.
Thanks to Copernicus’ Xenia van Edig for a clarification posted to the Global Open Access List and inthe comments section below. According to Xenia, the tripling of per-page pricing does not impact the article pricing as this represents a change in timing of assessment of charges. Previously, charges were assessed at the discussion paper stage, now they are assessed at the final publication stage, where articles are typically one third of the length.This makes sense and my 2015 data does refer to costing at the discussion stage. It is clear to me that Copernicus aims to be fully open and transparent about article processing charges. I recommend that Copernicus and its partners review the website information for the journals listed below (second chart) indicated “cost not found” or “cost not specified”.
Dirk Pieper on the GOAL list points to an open APC server for server payers of APCs and indicates that their data does not confirm tripling of prices. A review of the OLAC server data for Copernicus for 2015 and 2016 does suggest that the primary difference in pricing is paper stage (final v. discussion). However, only 16 Copernicus articles are listed for 2016. I have not checked to see if all are from the journals with this change in pricing; of those with this change in pricing, any submitted before December 31, 2015 would have been assessed at the discussion paper stage. This data is far from conclusive. From the 2015 data, there are 2 journals for which mean APCs are available for both Discussion Papers and the final publication stage. The mean value for Atmospheric Chemistry and Physics is €1.543 compared to Atmospheric Chemistry and Physics Discussions at € 1.32. There is a 17% difference, with the 2016 final paper model representing the higher value. Atmospheric Measurement Techniques’ mean value is € 1.34 compared to Atmospheric Measurement Techniques Discussions at € 1.09, a 24% difference. This data suggests that the change in timing of assessment of charges may not be price neutral, although this is not enough data to come to any conclusions about direction or net impact of the change.
Original post and update
Update July 5, 2016: thanks to astute readers for asking questions about the calculations. It may not be easy to follow as we’re used to much smaller price increase – a 3% increase from year to the next, not 300%! I’ve added more detail below which should make this easier to understand. Further questions are welcome.
On June 1, 2016, German-based commercial open access publisher Copernicus endorsed OA2020, an “an international initiative to support the swift, smooth, and scholarly oriented transformation of journals from print subscription to open-access publishing”. Half of Copernicus’ journals (12) for which we have numeric data for both 2015 and 2016 tripled or in some cases quadrupled their page charges from 2015 to 2016. Coincidence? The rest of this group of journals (10 journals) showed no change in price or modest increases. There were 18 journals for which I found no data with which to compare. Of these, 1 is clearly sponsored and free of charge; 6 are “currently waived”; 4 are “cost not specified”, i.e. the website indicates a charge will be applied but the amount is not given; 3 are “no cost found”, i.e. presumably free but no clear language to confirm; 3 are “title not found” and 1 title that was free last year began to implement charges in 2016. Following are comments: in brief, the price changes illustrate what I call the volatility of the market suggesting it may be too early for OA2020 as pricing is not stable, and I suggest that the topics of many of the journals which touch on things like drinking water quality, climate change and ocean science, would support an argument for public sponsorship of research dissemination in these areas that are important public priorities in the present and near future. Full data is available in charts below.
Copernicus is a highly regarded and innovative open access publisher with an innovative model where prices reflect the work involved (pricing varies with size of article, format of submission and whether the author uses the publishers’ template). Even after the tripling of prices most of the fees for these journals are fairly modest, e.g. a 10-page article in the preferred format using the publisher’s template adds up to less than a thousand Euros. However, this data illustrates what I describe as the continuing volatility of this market. Libraries and others that might be payers of article page processing charges tend to have fixed budgets that do not easily accommodate tripling of prices or sudden institution of pricing for “currently waived” journals. To illustrate: if a library budgeted based on 2015 prices and then prices on average triple in 2016, the fund will run out 1/3 of the way through the year.
The topics of some of the journals suggest to me the advisability of public sponsorship of at least some of these journals that are publishing in key public good priorities for the present and near future, for example Drinking Water Engineering and Science, Climate of the Past, and Ocean Science (all journals that tripled in price this year). It is great to see that these journals are open access, but if anyone has the ability and inclination to grow our knowledge on how to care for our water, climate, or oceans, let’s not let page charges stand in the way of disseminating the research.
The following chart illustrates APPC (article page processing charges) for all the journals for which we have numeric data for both 2015 and 2016. In reading the % change column, 100% means no price change, 300% means the price tripled and 400% means the price quadrupled.
Update July 5: I’ve added a column to the chart below on the right hand side which explains in plain language the difference in price between 2015 and 2016. A 400% change is a quadrupling of price. If your rent was $1,000 last year and increased to $4,000 this year, that’s 4 times higher. This is easiest to see and understanding with the journals that tripled in price from 25 EUR per page to 75 EUR per page. 25 X 3 = 75.
For many of these journals, to their credit Copernicus has posted pricing for articles submitted in both 2015 and 2016. To illustrate let’s look at one of these journals, Atmospheric Chemistry and Physics. The APCs are currently posted here and a PDF copy can be found here ACP – Article processing charges 2016 06 15 (created because Copernicus is likely to update their APC page at some point in the future).
Here are the APC details, an extract copied from the ACP website as of July 5, 2016 (without formatting, extracted to facilitate comparison). The following illustrates that the price for every category is precisely 3 times higher after January 2016 than it was before December 2016. Category 1 jumps from €25 to €75. Category 2 jumps from €30 to €90. Category 3 jumps from €35 to €105. Category 4 jumps from €40 to €120. Not every journal follows this pattern precisely, but it is a repeated pattern. I encourage readers to look at the APCs on the Copernicus journal sites to see for yourselves.
Discussion papers published before 31 December 2015
Price per discussion page
Category 1 €25 net LaTeX submissions using the Copernicus Publications LaTeX Package
Category 2 €30 net Word submissions using the Copernicus Publications Word Template
Category 3 €35 net LaTeX submissions causing higher expenditure of work due to any LaTeX style and/or difficulties with tables/figures
Category 4 €40 net Word submissions causing higher expenditure of work due to any Word
style and/or difficulties with tables/figures
Discussion papers published after 1 January 2016
Price per journal page
Category 1 €75 net LaTeX submissions using the Copernicus Publications LaTeX Package
Category 2 €90 net Word submissions using the Copernicus Publications Word Template
Category 3 €105 net LaTeX submissions causing higher expenditure of work…
Category 4 €120 net Word submissions causing higher expenditure of work…
|Copernicus APPC (*category 1 where multiple prices exist) 2016 / 2015|
|Title||2016 APPC as of June 15, 2016 (EUR) *||2015 05 15 (APPC) (EUR) *||Price increase in EUR 2016 – 2015||Price ratio 2016/2015 in %||Plain language price ratio from 2015 to 2016|
|Nonlinear Processes in Geophysics||75||17||58||441%||more than quadrupled|
|Solid Earth||75||17||58||441%||more than quadrupled|
|Drinking Water Engineering and Science||69||20||49||345%||more than tripled|
|Atmospheric Chemistry and Physics||75||25||50||300%||tripled|
|Atmospheric Measurement Techniques||75||25||50||300%||tripled|
|Climate of the Past||75||25||50||300%||tripled|
|Geoscientific Model Development||75||25||50||300%||tripled|
|Hydrology and Earth System Sciences||75||25||50||300%||tripled|
|Natural Hazards and Earth System Sciences||75||25||50||300%||tripled|
|Ocean Science (OS)||75||25||50||300%||tripled|
|Advances in Radio Science : Kleinheubacher Berichte||26||25||1||104%||slight increase|
|Advances in Science and Research||26||25||1||104%||slight increase|
|ASTRA Proceedings||31||30||1||103%||slight increase|
|Advances in Geosciences||31||31||0||100%||no change|
|Annales Geophysicae||45||45||0||100%||no change|
|Fossil Record||50||50||0||100%||no change|
|Mechanical Sciences||45||45||0||100%||no change|
|Primate Biology||55||55||0||100%||no change|
|Proceedings of the International Association of Hydrological Sciences||35||35||0||100%||no change|
|Stephan Mueller Special Publication Series||25||25||0||100%||no change|
|Geothermal Energy Science||60||57.5||2.5||104%||no change|
|Average percentage change||219%|
The following chart provides data for journals for which we do not have numeric data for both 2015 and 2016. Note that Copernicus has a number of journals which have associated journals that are called Discussions; these are actually early release of the journals with the same name and are not included here.
|Title||2016 APPC as of June 15, 2016 (EUR) *|
|Earth System Dynamics||50|
|Geographical Research Abstracts||abstracts of conference papers – no cost found|
|Earth Surface Dynamics||cost not specified|
|Earth System Science Data||cost not specified|
|Geoscientific Instrumentation, Methods and Data Systems||cost not specified|
|SOIL||cost not specified|
|Advances in Statistical Climatology, Meteorology and Oceanography||currently waived|
|Archives Animal Breeding||currently waived|
|Geographica Helvetica||currently waived|
|Journal of Sensors and Sensor Systems||currently waived|
|Web Ecology||currently waived|
|Wind Energy Science||currently waived|
|Scientific Drilling||free of charge|
|History of Geo- and Space Sciences||no cost found|
|ISPRS Archives and Annuals||no cost found|
|Animal Breeding||title not found|
|Social Geography (SG)||title not found|
|Astrophysics and Space Sciences Transactions (ASTRA)||title not found – see ASTRA|
Walt Crawford has released his data and book Gold Open Access Journals 2011-2015. For explanations on the different versions and links to download, go to Crawford’s website http://waltcrawford.name/goaj.html
Selected key findings:
- Of the 10,324 fully open access journals listed in DOAJ studied by Crawford as of December 31, 2015, 71% do not charge article processing fees. This is an important contribution to our understanding as previous information from DOAJ only went up to 2013.
- These 10,324 journals collectively published 566,922 articles in 2015 alone. This data was derived from counting the articles on each journal’s website, and is another major contribution of this work. This count is more accurate than relying on DOAJ metadata for article count which reflects the limitations of not all journals being able to supplement metadata and not all journals supplying metadata are able to supply correct metadata for year of publication.
- 44% of the articles published in 2015 were published by no-fee journals.
This selection of key findings reflects the summary data most pertinent to the SKC OA APC project; I encourage readers to go to the book and download the data for answers to other questions that may more relevant for you.
Kudos and thanks to Walt Crawford for doing this work and freely sharing both the data and the book and to SPARC for supporting this work. Disclosure: Crawford provided us with an early version of his spreadsheet which we are incorporating into our own longitudinal study. One illustration of the value of this is that having the ability to identify free-to-publish journals as of 2015 will enable comparisons of free and APC journals using our longitudinal data and the DOAJ metadata once our 2016 spreadsheet is complete (likely this fall). Also although there are important difference between our data, there are enough similarities that we will be able to use a considerable amount of Crawford’s data in lieu of doing the work ourselves, for example in the case of fairly stable publishers with straightforward APCs in USD. This behind-the-scenes collaboration is one of the reasons our team has been able to forge ahead with individual publisher case studies this year and even include selected open/hybrid comparisons. If anyone is looking for examples of the benefits of collaboration and open sharing, you might consider adding this one to your list.
The slides from my presentation at the Society for Scholarly Publishing AGM June 3, 2016 in Vancouver are available here: SSP presentation 2016 06 03 hgm
Please see the references slides at the ends for citations and links to further information. Comments and questions are welcome, here or via email to Heather dot Morrison at uottawa dot ca. If you would like to comment on the blog but prefer anonymity please contact me via e-mail