2010 – 2016 APC journals comparison: attrition rate

In brief: our evidence suggests an average attrition rate of APC-charging journals of approximately 1.5% – 2% per year, depending on whether one includes or excludes the anomaly of a particular business model of a new publisher (Bentham Open) starting off with a very large number of journals and retaining only successful titles.


Solomon and Björk (2012) conducted a survey in 2010 of APCs based on a sample of 1,090 APC-charging journals that were listed in DOAJ at that time. Not all APC-charging journals in DOAJ as of 2010 were included; no journals listed in DOAJ as of 2010 that did not charge APCs were included. We compared this data with our 2016 APC survey (Morrison et al, 2017).

Of the 1,090 titles, 32 titles were published by publishers that have never been included in our longitudinal APC survey. Our survey is not limited to DOAJ, but is limited to fully OA journals that either have been listed in DOAJ in 2014, 2015, or 2016, or whose publishers have been listed in DOAJ in 2014, 2015, or 2016. Excluding the 32 titles, of the 1,058 titles studied, 849 are confirmed active in 2016, and 209 are discontinued or presumed discontinued for an attrition rate of 20% from 2010 – 2016. However it is important to note that Bentham Open is a significant anomaly, accounting for 123 of the discontinued titles from 2010 – 2016 and 4 of the discontinued titles from 2015 – 2016. Excluding these journals, the attrition rate is 82 / 935 journals or 9% over the 6 year period.

The attrition rate for 2015 – 2016 for journals studied in 2010 is 2%, or 19 of the 868 journals studied in 2015 found to be inactive in 2016. Note that 4 of these journals (25% of the total) are Bentham Open journals.

The Bentham Open anomaly reflects a particular business practice. This fairly new publisher has taken the approach of creating a large number of journals at once and appears to be retaining titles that are successful. That is, Bentham Open’s high numbers of discontinued journals need to be assessed in light of the total number of journals. In 2010, Bentham Open accounted for 211 of the 1,090 journals studied. Of these, 84 journals were active in 2016. In other words, the number of discontinued journals does not necessarily reflect an unsuccessful publisher.

In conclusion, the evidence appears to suggest an average attrition rate of about 1.5% – 2% of APC charging journals sampled, depending on whether we include or exclude the anomaly of a publisher model of starting with a large number of journals on speculation and retaining only successful titles.


2010 to 2016 comparison attrition


Morrison, Heather; Brutus, Widlyne; Dumais-Desrosier, Myriam; Laprade, Katherine; Merhi, Salah; Ouerghi, Arbia; Salhab, Jihane; Volkanova, Victoria; Wheatley, Sara, 2017, “Open access article processing charges 2016”, doi:10.5683/SP/KC2NBV, Scholars Portal Dataverse, V2

Solomon, D. J. & Björk, B.-C. (2012) A study of open access journals using article processing charges. Journal of the American Society for Information Science and Technology 2012, 63, 1485–1495. Available online: http://www.openaccesspublishing.org/apc2/preprint.pdf

APCs in DOAJ 2017: summary of 3 studies

The DOAJ application form requests information from journals and publishers about article processing charges: whether or not there are charges, if yes the amount and currency, and a URL for more information. Is DOAJ APC data sufficient for the purposes of our longitudinal study on APC charges? In brief, we compared the APCs on publisher website for 3 publishers and conclude that no, DOAJ APC is not sufficient. There are significant differences between APC data on the websites of publishers Hindawi, MDPI, and Taylor & Francis. Conclusion: APC details in DOAJ are not sufficient for the longitudinal study of APC.

Details in brief and links to substantive posts

We used the DOAJ metadata as of Jan. 31, 2017 (our DOAJ metadata set for the 2017 APC study) for these studies.

Widlyne Brutus compared journal and APC data for Hindawi on the Hindawi website and in DOAJ – not an easy task as the title lists in DOAJ and on the Hindawi website are quite different. DOAJ includes titles that Hindawi no longer publishes (a good practice but this makes research a challenge), but not all of the titles that Hindawi currently does publish. Hindawi titles have a high rate of APC listings in DOAJ, but only 9% of the titles have the same price in DOAJ and on the publisher’s website. 144 titles have higher prices on the publisher’s website, while 45 have a lower price on the publisher’s website. Details: https://sustainingknowledgecommons.org/2017/03/14/hindawi-apc-le-site-web-dhindawi-versu-le-repertoire-du-doaj/

Arbia Ouerghi found that 107 MDPI titles listed in DOAJ have an APC according to the MDPI website. DOAJ has APC data for only 21 of these journals, and only 3 have the same price on the MDPI website and in DOAJ. APCs on the MDPI website are higher than those listed in DOAJ. Details: https://sustainingknowledgecommons.org/2017/03/02/comparaison-doaj-et-mdpi-titres-et-apc/

Katherine Laprade found 150 fully OA titles on the Taylor and Francis website. 77% of these titles have an APC according to the Taylor & Francis website, but only 1% have an APC listed in DOAJ. Of these titles, 54% have a different amount in DOAJ as compared with the publisher’s website. https://sustainingknowledgecommons.org/2017/03/28/taylor-francis-2017-data/

Comment (Heather Morrison): the question about whether journals charge an APC or not is a useful one, however the answer based on our research is more complex than a simple yes / no. DOAJ used to have a “conditional” charges option that I recommend re-instating. As we noted in 2014 (Morrison et al, 2015), the vast majority of fully OA journals (over 90%) have variations in pricing based on such factors as the status of the author (society member, editing contributions to the journal, location, perceived ability to pay), and the nature of the work (length, quality, type of article). A single flat-fee approach to APC does exist but is not that common. Collecting specific APC information and keeping it up to date will depend on publishers updating DOAJ every time there is a price change. It seems likely that DOAJ’s APC information will become more and more outdated over time. The yes / no / conditional information and link to where to find the information seem likely to be stable and useful, but DOAJ and its user community might want to reconsider the costs and benefits of capturing specific APC amounts.


Morrison, H.; Salhab, J.; Calvé-Genest, A.; Horava, T. Open Access Article Processing Charges: DOAJ Survey May 2014. Publications 2015, 3, 1-16. doi:10.3390/publications3010001

Taylor & Francis 2017 Data

Abstract: Taylor & Francis is one of several major, traditionally commercial, scholarly publishers that is moving to the Open access space. Taylor & Francis publishes 2,550 journals and now has a major part of them available for Gold Open Access. 150 journal published by Taylor & Francis are in fully open access.

Around 77% of fully open access journals have an APC, but only 1% of those have an APC in DOAJ. Of the APCs available in DOAJ, 54% varies from the APCs found in Taylor & Francis. Less than 3% of fully open access journals have an APPC.

At least 7 journals published by Taylor & Francis do not have publications fees.

The APC average for fully open is 794.55 GBP (including only the 106 journals with APCs), for hybrid is 1574.38 GBP and for subscription was not calculated because only three journals had APCs.

These averages are higher than the APC average for 2016 of 702.74 GBP (converted with the Bank of Canada currency converter (on March 21st, 2017)). (Excluding journals with an APC of $0).

T&F Pie Chart 2017

See full text here

Ottawa open access workshop April 1

If you are in Ottawa on Saturday, April 1, please join the open access class for an all-day workshop on OA. Registration is free, but appreciated if you plan to join us for lunch. Details and a link to register can be found here: https://oa20171avril.wordpress.com/2017/03/18/all-day-open-access-workshop-at-university-of-ottawa/

Thanks to students of ISI 6300, a new master’s-level special topics course on open access, for organizing the event. Sustaining the Knowledge Commons is one of the sponsors.

MPDI : suivi des APCs

Résumé :

En comparant les frais de publication des différentes revues mentionnées sur le site MDPI entre mars 2016 et mars 2017 nous remarquons quelques différences entre les 2 années.

Ceci dit que les APCs ont augmenté dans 92 % des revues. Cette augmentation oscille entre 6% et 350 %.

Par exemple, les APCs pour la revue Coalitions est passé de 350 CHF en 2016 a 850 CHF en 2017.

Selon le site MDPI, environ 60 revues en libre accès n’exigent pas de frais de publication en 2017.

En observant de près les APCs des différentes revues, nous remarquons qu’il n’y a aucune baisse des frais de publication et que seulement 8% des revues ont gardé les mêmes frais de publication 2016-2017.


Texte intégral de l’article

Scientific Research Publishing 2017: not currently in DOAJ but apparently stable and active

Scientific Research Publishing

2017 sample (5 journals)

In brief: Scientific Research Publishing is not currently listed in DOAJ, but is included in the APC study as a publisher previously included in DOAJ and the longitudinal APC study. The publisher currently has 247 titles. A sample of the 65 titles for which we have data from previous years shows stable pricing from 2016 – 2017. All journals are active, publishing 4 – 14 issues and 13 – over 100 articles in 2016. Conclusion: this publisher appears active and stable. Our sampling for 2017 is limited to these 5 journals.


Full explanation and methods notes in PDF:  Scientific Research Publishing

Download data in excel format: SRP_2017_plus_OA_APC_Main_2016sample