Tag Archives: article processing charges

Hybrid open access – an analysis

Welcome to Open Access Week 2016. The Office of Scholarly Communication at Cambridge is celebrating with a series of blog posts, announcements and events. In today’s blog posts we revisit the issue of paying for hybrid open access. We have also published a related post “Who is paying for hybrid?” listing funder policies on hybrid.

Recent years have seen a proliferation of funder open access mandates, the terms of which can differ markedly, adding to the confusion of an already complex area. The Registry of Open Access Repository Mandates and Policies (ROARMAP) lists 80 funders with open access requirements, and the list continues to grow.

Within the UK, policies fall into three broad categories: those that mandate green Open Access without paying a fee, such as the HEFCE policy; those that prefer gold but make no additional funds available, such as the NIHR policy, and those that have a preference for gold and offer block grants to institutions to help cover the associated costs, such as the Research Councils UK (RCUK) and Charities Open Access Fund (COAF) policies.

Accompanying this expansion of mandates, unsurprisingly, has been an increase in the amount being spent to support Open Access. The Open Access Directory lists 179 funds for OA journal articles worldwide, compared with 81 in early 2014.

All this brings into sharper relief the question of how open access funds support hybrid publishing. But first a quick history lesson.

Hybrid origins

Hybrid journals provide open access to specific articles where an Article Processing Charge has been paid in an otherwise subscription journal. A few learned societies offered hybrid options in the early 2000s. Hybrid open access options were first offered by large publishers in 2004 with Springer’s Open Choice product charging USD3000 per article. This price has not changed in the past 12 years. In the UK the Springer Compact now pays for hybrid under a different model.

Wiley Online Open’s trial began the same year, charging USD2500. Today the price ranges from USD1,500 – 5,200. Oxford Open launched in 2005, and in 2006 Elsevier Open Access and Sage Choice began. In 2007, Taylor & Francis Open Select, Cambridge Open and Nature Publishing Group’s open access offering began.

The uptake of hybrid began slowly. It is very difficult to obtain statistics on what percentage of journals have hybrid Open Access content but in his 2012 analysis The hybrid model for open access publication of scholarly articles – a failed experiment?, (open access version here ) Bo-Christer Bjork found the number of hybrid journals had doubled in the previous couple of years to over 4,300, and the number of such articles was around 12,000 in 2011. This represented a small proportion of eligible authors (1-2 %).

That analysis was published the same year as the Finch Report which recommended a gold path to Open Access. The resulting RCUK Open Access Policy and RCUK Block Grants to fund Open Access APCs has dramatically increased the  expenditure on hybrid in the UK since 2013. According to a report published in 2015, “the UK’s profile of OA take-up is significantly different from the global averages: its use of OA in hybrid journals and of delayed OA journals is more than twice the world average in both cases, while its take-up of fully OA journals with no APC (Gold-no APC) is less than half the world average and falling.”

At Cambridge University we have spent literally millions of pounds on hybrid Open Access – which constitutes approximately 85% of our total APC spend. This is a higher percentage than estimates across the country, which are a 76% spend on hybrid Open Access.

Double dipping

Hybrid represents a second income stream to publishers and has raised questions about ‘double dipping’. Some publishers manage this by reducing the cost of subscriptions in proportion to the percentage of hybrid in a given journal, such as Nature Publishing Group. However ‘big deals’ for subscriptions can render this relatively ineffective, and the reduction is spread across all subscribers, regardless of who has paid the article processing charge. This means research intensive institutions (such as Cambridge) are contributing heavily to the system but not receiving a relative reduction.

To address this issue at a local level, several publishers have created offsetting arrangements, where discounts or refunds are provided in proportion to the contribution the institution has made in APC payments above subscriptions. However, each of these schemes operates differently and they can be complicated to administer, or have other preconditions such making large prepayments to publishers.

The biggest problem from an implementation perspective, however, is that they are by no means universal. By far the biggest publisher, Elsevier, for example, offers no form of offsetting at all, although they nevertheless assert that they do not double dip. The result is that in very many cases, institutions and authors continue to have to pay twice for material in hybrid journals, swelling publisher coffers at the expense of research funding.

Very expensive

One of the problems with hybrid is that even ignoring the added cost of subscriptions to the non Open Access material in those journals, hybrid Open Access charges are more expensive than those for fully Open Access journals.

In March last year both the Wellcome Trust and the RCUK undertook a review of their Open Access policies. The Reckoning: An Analysis of Wellcome Trust Open Access Spend 2013 – 14  noted: “The average APC levied by hybrid journals is 64% higher than the average APC charged by a fully OA title”.  In Wellcome’s data, the average APC for a hybrid article in 2014-15 was £2104, compared with only £1396 for fully OA journals. Worryingly, the data showed that fully OA APC costs had risen more than their hybrid counterparts since the previous year.

Similarly in the Research Councils UK 2014 Independent Review of Implementation the observation was that article processing charges for hybrid Open Access were “significantly more expensive” than fully OA journals, “despite the fact that hybrid journals still enjoyed a revenue stream through subscriptions”.

A Max Planck Digital Library Open Access Policy White Paper published on 28 April 2015 noted that The Wellcome Trust had a significantly higher average APC cost than German, Austrian and SCOAP3 figures. This was because the Wellcome Trust pays for hybrid APCs, “which are not only much higher than most pure open access costs but are also widely considered not to reflect a true market value. In Germany and many other countries, hybrid APCs are excluded from the central funding schemes.”

A study undertaken last year considered APCs in the five-year period between 2010 and 2014 found the mean for fully-OA journals published by non-subscription publishers was£1,136 compared with £1,849 for hybrid journals. The same study also found that traditional subscription publishers are capturing most of the APC market. The top-10 publishers in terms of numbers of APCs received from participant institutions (who received 76% of the total APCs paid from the sample) “only included two fully-OA publishers (PLOS and BMC). The others were established publishers (Elsevier, Wiley, Springer and so on) who are mostly gaining APC income from hybrid journals.”

The 2014 report Developing an effective market for open access article processing charges was written for a consortium of research funders comprising Jisc, Research Libraries UK, Research Councils UK, the Wellcome Trust, the Austrian Science Fund, the Luxembourg National Research Fund and the Max Planck Institute for Gravitational Physics. The authors noted of the hybrid journal market that it is “highly dysfunctional, with very low uptake for most hybrid journals and a relatively uniform price in most cases without regard to factors such as discipline or impact“.

Value for money?

A second issue which has become apparent as open access mandates have expanded is the extent to which publishers – mostly of hybrid journals – do not deliver the Open Access option that has been paid for. In many cases, the ‘immediate’ Open Access for which an author or institution has paid an APC may take months or even years to be made Open Access; some articles are never made Open Access at all. Even when articles are made available, there is no guarantee that it will have the appropriate licence. It is by no means uncommon for articles to carry more restrictive licences than those requested, or for the appropriate licence to appear on a journal website while the PDF of the article itself bears only a publisher copyright notice and a prominent ‘All rights reserved’.

In March 2016 the Wellcome Trust published a report into compliance among its paid-for articles in 2014-15, concluding:

The good news is that we have seen an improvement in correct and programmatically identifiable licences (from 61% of papers in ’13-‘14, to 70% in ’14-‘15) and a similar increase in overall compliance from 61% to 70%.  The bad news, however, is that in 30% of cases we are not getting what we are paying for.

The source of this non-compliance was overwhelmingly hybrid journals, and the largest publishers were the worst offenders: in the Wellcome data, 31% of Elsevier hybrid articles (and 26% of their ‘fully OA’ articles!) were non-compliant, as were 54% of Wiley’s.

One might conclude, then, that hybrid Open Access represents a bad deal for funders and institutions, with poor service and double-dipping.

Other hybrid issues

To further complicate matters, some have argued that the open access/hybrid dichotomy is too stark. Some journals, particularly coming from learned societies, (e.g. Plant Physiology, from the American Society of Plant Biologists) make all articles open access after a certain period, but charge an optional APC to make them available sooner. This would generally be considered hybrid publishing, but could be seen as a rather different category from the majority of corporate hybrid journals, in which articles never become Open Access unless an APC is paid. There is a possibility that strict funder mandates against hybrid could close off such journals to researchers, exacerbating the anxieties regarding open access felt by many learned societies.

Where does this leave authors and institutions? It’s clear that the situation remains very much in flux. The problems that have existed with hybrid since the beginnings of Open Access are far from resolved, despite the expansion of journal offsetting schemes. Meanwhile, prices continue to rise and while many funders have taken the step of allowing their funds to be used only for fully Open Access journals, it is still a minority of the largest and most powerful funding bodies.

The result is confusion for researchers and an increased administrative burden for institutions, who have to manage and advise on a proliferation of divergent funder and publisher policies, as well as conducting regular and extremely resource-intensive compliance-checking of hybrid publications to ensure publishers have delivered what has been paid for. As numbers of Open Access publications increase, it is questionable how sustainable this will be.

Published 24 October 2016
Written by Dr Philip Boyes and Dr Danny Kingsley 
Creative Commons License

Cambridge University spend on Open Access 2009-2016

Today is the deadline for those universities in receipt of an RCUK grant to submit their reports on the spend. We have just submitted the Cambridge University 2015-2016 report to the RCUK and have also made it available as a dataset in our repository.

Compliance

Cambridge had an estimated overall compliance rate of 76% with 46% of all RCUK funded papers  available through the gold route and 30% of all RCUK funded papers available through the green route.

The RCUK Open Access Policy indicates that at the end of the fifth transition year of the policy (March 2018) they expect 75% of Open Access papers from the research they fund will be delivered through immediate, unrestricted, on‐line access with maximum opportunities for re‐use (‘gold’). Because Cambridge takes the position that if there is a green option that is compliant we do not pay for gold, our gold compliance number is below this, although our overall compliance level is higher, at 76%.

Compliance caveats

The total number of publications arising from research council funding was estimated by searching Web of Science for papers published by the University of Cambridge in 2015, and then filtered by funding acknowledgements made to the research councils. The number of papers (articles, reviews and proceedings papers) returned in 2015 was 2080. This is almost certainly an underestimate of the total number of publications produced by the University of Cambridge with research council funding. The analysis was performed on 15/09/2016.

Expenditure

The APC spend we have reported is only counting papers submitted to the University of Cambridge Open Access Team between 1 August 2015 and 31 July 2016. The ‘OA grant spent’ numbers provided are the actual spend out of the finance system. The delay between submission of an article, the commitment of the funds and the subsequent publication and payment of the invoice means that we have paid for invoices during the reporting period that were submitted outside the reporting period. This meant reconciliation of the amounts was impossible. This funding discrepancy was given in ‘Non-staff costs’, and represents unallocated APC payments not described in the report (i.e. they were received before or after the reporting period but incurred on the current 2015-16 OA grant).

The breakdown of costs indicates we have spent 4.6% of the year’s allocation on staff costs and 5.1% on systems support. We noted in the report that the staff time paid for out of this allocation also supports the processing of Wellcome Trust APCs for which no support is provided by Wellcome Trust.

Headline numbers

  • In total Cambridge spent £1,288,090 of RCUK funds on APCs
  • 1786 articles identified as being RCUK funded were submitted to the Open Access Service, of which 890 required payment for RCUK*
  • 785 articles have been invoiced and paid
  • The average article cost was ~£2008

Caveats

The average article cost can be established by adding the RCUK fund expenditure to the COAF fund expenditure on co-funded articles (£288,162.28)  which gives a complete expenditure for these 785 articles of £1,576,252.42. The actual average cost is £2007.96.

* The Open Access Service also received many COAF only funded and unfunded papers during this period. The number of articles paid for does not include those made gold OA due to the Springer Compact as this would throw out the average APC value.

Observations

In our report on expenditure for 2014 the average article APC was £1891. This means there has been a 6% increase in Cambridge University’s average spend on an APC since then. It should be noted that of the journals for which we most frequently process APCs, Nature Communication is the second most popular. This journal has an APC of £3,780 including VAT.

Datasets on Cambridge APC spend 2009-2016

Cambridge released the information about its 2014 APC spend for RCUK and COAF in March last year and intended to do a similar report for the spend in 2015, however a recent FOI request has prompted us to simply upload all of our data on APC spend into our repository for complete transparency. The list of datasets now available is below.

1. Report presented to Research Councils UK for article processing charges managed by the University of Cambridge, 2014-2015

2. Report presented to the Charity Open Access Fund for article processing charges managed by the University of Cambridge, 2015-2016

3. Report presented to the Charity Open Access Fund for article processing charges managed by the University of Cambridge, 2014-2015

4. Report presented to Jisc for article processing charges managed by the University of Cambridge, 2014

5. Open access publication data for the management of the Higher Education Funding Council for England, Research Councils UK, Charities Open Access Fund and Wellcome Trust open access policies at the University of Cambridge, 2014-2016

Note: In October 2014 we started using a new system for recording submissions. This has allowed us to obtain more detailed information and allow multiple users to interact with the system. Until December 2015 our financial information was recorded in the spreadsheet below. There is overlap between reports 5. and 6. for the period 24 October and 31 December 2015.  As of January 2016, all data is being collected in the one place.

6. Open access publication data for the management of Research Councils UK, Charities Open Access Fund and Wellcome Trust article processing charges at the Office of Scholarly Communication, 2013-2015

Note: In 2013 the Open Access Service began and took responsibility for the new RCUK fund, and was transferred responsibility for the new Charities Open Access Fund (COAF). At this time the team were recording when an article was fully Wellcome Trust funded, even though the Wellcome Trust funding is a component of COAF.

7. Open access publication data for the management of Wellcome Trust article processing charges from the School of Biological Sciences, 2009-2014

Note: Management of the funds to support open access publishing has changed over the past seven years. Before the RCUK open access policy came into force in 2013, the Wellcome Trust funds were managed by the School of Biological Sciences.

Published 14 September 2016
Written by Dr Danny Kingsley & Dr Arthur Smith
Creative Commons License

Dutch boycott of Elsevier – a game changer?

A long running dispute between Dutch universities and Elsevier has taken an interesting turn. Yesterday Koen Becking, chairman of the Executive Board of Tilburg University who has been negotiating with scientific publishers about an open access policy on behalf of Dutch universities with his colleague Gerard Meijer, announced a plan to start boycotting Elsevier.

As a first step in boycotting the publisher, the Association of Universities in the Netherlands (VSNU) has asked all scientists that are editor in chief of a journal published by Elsevier to give up their post. If this way of putting pressure on the publishers does not work, the next step would be to ask reviewers to stop working for Elsevier. After that, scientists could be asked to stop publishing in Elsevier journals.

The Netherlands has a clear position on Open Access. Sander Dekker, the State Secretary  of Education has taken a strong position on Open Access, stating at the opening of the 2014 academic year in Leiden that ‘Science is not a goal in itself. Just as art is only art once it is seen, knowledge only becomes knowledge once it is shared.’

Dekker has set two Open Access targets: 40% of scientific publications should be made available through Open Access by 2016, and 100% by 2024. The preferred route is through gold Open Access – where the work is ‘born Open Access’. This means there is no cost for readers – and no subscriptions.

However Gerard Meijer, who handles the negotiations with Elsevier, says that the parties have not been able to come close to an agreement.

Why is this boycott different?

It is true that boycotts have had different levels of success. In 2001, the Public Library of Science started as a non-profit organization of scientists ‘committed to making the world’s scientific and medical literature freely accessible to both scientists and to the public’. In 2001 PLoS (as it was then) published an open letter asking signatories to pledge to boycott toll-access publishers unless they become open-access publishers. The links to that original pledge are no longer available. Over 30,000 people signed , but did not act on their pledge. In response, PLOS became an open access publisher themselves, launching PLOS Biology in October 2003.

In 2012 a Cambridge academic Tim Gowers started the Cost of Knowledge boycott of Elsevier which now has over 15,000 signatures of researchers agreeing not to write for, review for, or edit for Elsevier. In 2014 Gowers used a series of Freedom of Information requests to find out how much Elsevier is charging different universities for licence subscriptions. Usually this information is a tightly held secret, as individual universities pay considerably different amounts for access to the same material.

The 2015 Dutch boycott is significant. Typically negotiations with publishers occur at an institutional level and with representatives from the university libraries. This makes sense as libraries have long standing relationships with publishers and understand the minutiae of the licencing processes . However the Dutch negotiations have been led by the Vice Chancellors of the universities.  It is a country-wide negotiation at the highest level. And Vice Chancellors have the ability to request behaviour change of their research communities.

This boycott has the potential to be a significant game changer in the relationship between the research community and the world’s largest academic publisher. The remainder of this blog looks at some of the facts and figures relating to expenditure on Open Access in the UK. It underlines the importance of the Dutch position.

UK Open Access policies mean MORE publisher profit

There have also been difficulties in the UK in relation to negotiations over payment for Open Access. Elsevier has consistently resisted efforts by Jisc to negotiate an offsetting deal  – where a publisher provides some sort of concession for the fact that universities in the UK are paying unprecedented amounts in Article Processing Charges on top of their subscriptions because of the RCUK open access policy.

Elsevier is the world’s largest academic publisher. According to their Annual Report the 2014 STM revenue was £2,048 million, with an operating profit of £762 million. This is a profit margin of 37%. That means if we pay an Article Processing Charge of $3000 then $1,170 of that (taxpayers’) money goes directly to the shareholders of Elsevier.

The numbers involved in this space are staggering. The Wellcome Trust stated in their report on 3 March 2015 The Reckoning: An Analysis of Wellcome Trust Open Access Spend 2013 – 14: ‘The two traditional, subscription-based publishers (Elsevier and Wiley) represent some 40% of our total APC spend’.

And the RCUK has had similar results, as described in a Times Higher Education article on 16 April 2015 Publishers share £10m in APC payments: “Publishers Elsevier and Wiley have each received about £2 million in article processing charges from 55 institutions as a result of RCUK’s open access policy”.

Hybrid open access – more expensive and often not compliant

Another factor is the considerably higher cost of  Article Processing Charges for making an individual article Open Access within an otherwise subscription journal (called ‘hybrid’ publishing) compared to the Article Processing Charges for articles in fully Open Access journals.

In The Reckoning: An Analysis of Wellcome Trust Open Access Spend 2013 – 14, the conclusion was that the average Article Processing Charge levied by hybrid journals is 64% higher than the average Article Processing Charge of a fully Open Access title. The March 2015 Review of the implementation of the RCUK Policy on Open Access concluded the Article Processing Charges for hybrid Open Access were ‘significantly more expensive’ than fully OA journals, ‘despite the fact that hybrid journals still enjoyed a revenue stream through subscriptions’.

Elsevier has stated that in 2013 they published 330,000 subscription articles and 6,000 author paid articles. There is no breakdown of how many of those 6,000 were in fully open access journals and how many were hybrid. However in 2014 Elsevier had 1600 journals offering their hybrid option, and 100 journals that were fully open access (6%). Note that the RCUK open access policy came into force in April 2013. It would be interesting to compare these figures with  the 2014 ones, however I have been unable to find them.

While the higher cost for hybrid Article Processing Charges is in itself is an issue, there is a further problem. Articles in hybrid journals for which an Article Processing Charge has been paid are not always made available at all, or are available but not under the correct licence as required by the fund paying the fee. Here at Cambridge, the five most problematic publishers with whom we have paid more than 10 Article Processing Charges have a non compliance rate from 11-25%. With this group of publishers we are having to chase up between three and 31 articles per publisher. This takes considerable time and significantly adds to the cost of compliance with the RCUK and COAF policies.

According to the March 2015 Review of the implementation of the RCUK Policy on Open Access, ‘Elsevier stated that around 40% of the articles from RCUK funding that they had published gold were not under the CC-BY licence and are therefore not compliant with the policy’ (p19).

We support our Dutch colleagues

In summary, the work happening in The Netherlands to break the stranglehold Elsevier have on the research community is important. We need to stand by and support our Dutch colleagues.

NOTE: This blog was subsequently reblogged on the London School of Economics Impact Blog and later listed as one of the Top Ten Posts for 2015: Open Access. It was also listed as one of the blogs that had an average minute per page measurement of over 6 minutes and 30 seconds.

Published 3 July 2015, added to on 22 January 2016
Written by Dr Danny Kingsley
Creative Commons License