Summary and thoughts from a conference – attending the 7 th Conference on Open Access Scholarly Publishing

The Conference on Open Access Scholarly Publishing, COASP, is held annually with the aim of reaching professional publishing organizations, independent publishers and university presses, as well as librarians, university administrators and other stakeholders. Here, we outline some themes and highlights from this year’s conference.

The Conference on Open Access Scholarly Publishing, COASP, is held annually with the aim of reaching professional publishing organizations, independent publishers and university presses, as well as librarians, university administrators and other stakeholders.
The conference is arranged by the Open Access Scholarly Publishers Association. 1 The first conference was held in 2009 in Lund, Sweden.This year's conference took place in Amsterdam, the Netherlands.The conference's topics are, amongst others, open access usability and technical issues, policy issues, business models and research evaluation.All slides and recorded presentations are available from the conference website. 2 The conference themes have remained rather much the same over the years, but at the same time we have seen the idea of open access moving from theory to practice, standards being established and policies implemented.Here, we outline some themes and highlights from this year's conference.

TRANSITION TO OPEN ACCESS -BUSINESS AND POLICY ISSUES
SCOAP 3 Despite the fact that 97 % of the content of the 5 highest ranked journals in High-Energy Physics is freely available as preprints on arXiv, there is still a need to retain the traditional journals, not for content, but for the peer review process as well as for prestige and evaluation.The SCOAP3 project, Sponsoring Consortium for Open Access Publishing in Particle Physics, is a way of moving from the traditional model of publishing to open access.Salvatore Mele, from CERN and also in the executive committee for SCOAP 3 , presented the project that has been going on for 5 years.
The main purpose of the project is to convert key journals in the field of High-Energy Physics (HEP) to open access by re-using money spent by libraries for subscriptions, to pay for peer-review and publishing services.SCOAP 3 is paying publishers for the costs involved and publishers in turn reduce subscription fees for their packages.The funding has been collected from library subscriptions in universities or research institutes from 34 countries 3 .The approach has been successful in the sense that about 50 % of all formal HEP publications are now open content and the bidding with the publishers resulted in a lower article processing charge (APC) than the average, i.e. €1.042 in 2014.
Sharing of information via preprints has a long tradition in the HEP community.Already in 1953 it was stated that discoveries from CERN should be published or otherwise be generally made available.Publishing culture and routines in disciplines can thus be crucial for the success of a project such as SCOAP 3 .Salvatore Mele was of the opinion that libraries in general like SCOAP 3 .The project enables the possibility to harvest open access articles from the SCOAP 3 repository to include them in an institutional repository.Funding agencies also like SCOAP 3 , because it is cost efficient.At the same time the organization requires time-consuming and complex calculations to set the fee for each participating country and institution.Although SCOAP 3 has managed to transfer journals to open access, the solution is still built on the traditional publishing model.The journal prestige (impact factor) seems to be an important reason for maintaining the conventional model.It would have been interesting to see the project trying to create something completely new, like overlay journals built on top of the 97 % of HEP literature already freely available as preprints in arXiv.

Transition towards open access in the UK
SCOAP 3 is a quite unique example of a transition model within a specific discipline, another approach in the ongoing transition towards open access is by doing it on a national level using political and policy decisions, in this case in the UK.The UK open access landscape is somewhat different to the rest of Europe, because of the preference for gold over green open access (OA), which is a result of the decisions based on the Finch report from 2012. 4 Stephen Pinfield from the University of Sheffield based his presentation on the recently published report "Monitoring the transition to open access.A report for the Universities UK Open Access Co-ordination Group". 5his study was commissioned in response to a recommendation of the Finch Group that reliable indicators should be gathered on key features of the transition to open access in the UK.The results from the report show that OA publishing options are now widely available.Two-thirds of the world's journals offer an OA option of some kind and the most common model is the hybrid model.Nearly half of all journals across the world offer a hybrid option to authors.
Following the increased ability for authors to publish in an open access form, the UK Universities' expenditure of APCs has increased and now represents a significant proportion of their total expenditure on journals. 6entrally-managed APC expenditure has risen more than six-fold since 2012 and levels of APCs vary widely.APC payments varied from between 0 to £4.536 with a mean of £1.586, according to the report.Hybrid journal APCs were considerably more expensive than others.The majority of hybrid journals charge £1.000-£2.000.Only a small minority of fully OA journals charge more than £2.000.There was also a positive correlation between APC price and citation rates of journals.
Pinfield mentioned that the complexity of self-archiving policies for subscription-based journals makes compliance low.An estimated 36% of all postings are not in accordance with journal policies.Naturally, as posting policies are complex, self-archiving rules have to be manually checked.Policies are progressively more restrictive with posting in institutional and subject repositories, than on personal websites.Also, journal policies are often modified in relation to funder requirements, according to Pinfield.Frequencies of postings by subject area show health and life sciences to be highest.Most payments were made for articles in the health and life sciences and to large commercial publishers, probably as a result of large publishers managing successful journals within this subject area.
In conclusion, so far in the UK, there is a strong growth in open access options for authors and also a strong growth in take-up of open access.In terms of number of open access publications, UK authors are ahead of world averages.As the number of OA publications increases, there is also an increase in expenditures.

OTHER NATIONAL APPROACHES TO OPEN ACCESS
Open access policies are important tools for the transition to open access.The Netherlands has for long been actively promoting open access through institutional mandates and funder mandates.Also, the Dutch government wants to see a full transition to open access.By 2024, all scholarly publications by Dutch researchers should be freely available and the government has opted for the golden route as the most sustainable model in the long term.
At the conference, the national approach to open access was presented by Ron Dekker, Director Institutes at the Netherlands Organization for Scientific Research (NWO).He is also the project leader for Open Access in the Netherlands.The Dutch policy states that research results should be made open access without any embargo period.NWO has recently updated their funding conditions for open access publishing.The researcher is free to select a journal for publishing, which means that choosing hybrid journals is allowed, but this will not be funded by NWO funds.Dekker did not see hybrids as a sustainable choice.In his view, the development will either "go for gold" or we will fall back to the subscription based model.The current largest obstacle to policy implementation is misunderstanding, according to Dekker.In his presentation Dekker argued to go from policy to action through what he called "the polder model", which means building a well-working e-infrastructure with good leadership, also using quality assessment and policy monitoring.
In 2015 the Netherlands received special attention because of its national approach to open access and criticism against the publisher Elsevier's publishing policies.Dutch universities have been negotiating with Elsevier in the past year to reach a satisfying agreement on open access.In June this year, the negotiations reached a deadlock and were discontinued, which led to a boycott of Elsevier initiated by the Association of Universities in the Netherlands (VSNU). 7TURNING SUBSCRIPTIONS INTO GOLD -3 OFFSETTING MODELS Jisc Collections, Springer and IOPP representatives gave presentations on working practices with offsetting i.e. converting subscription costs to cover both access to read and open access publishing fees.The offsetting systems should support the transition to full open access and ensure that publishers do not charge the same institutions twice, through the payment of subscriptions and APCs (so-called double-dipping).

Jisc Collections
Jisc Collections negotiates with publishers at a national level in the UK to procure and license affordable digital content.Liam Earney, director of Jisc Collections, gave a presentation on how Jisc is merging their role as national negotiator of the electronic resources subscriptions in the UK, with the role of a national support organization of open access developments through project funding and competence development.A main driver in this merging of roles seems to be the open access mandates of RCUK, with a strong emphasis on gold open access, also allowing financial support to hybrid publishing, and supplying the major research universities with block grants to partly fund OA-related expenditures.
As mentioned previously, the costs of APCs have increased significantly in the UK.Liam Earney showed a graph on APC costs in the UK based on data collected by JISC, with a rise from £2.9 million in 2013 to £8.5 million in 2014.With publishers Wiley and Elsevier collecting a lion's share of these APCs, and their subscription prices (and profits) also rising it certainly looked like the accusation that the publishers were "double-dipping" had some substance to it. 8isc Collections has been trying to get offsetting agreements with the publishing houses since 2014 in the UK and a range of agreements are already in place, i.e. with Institute of Physics (IOP), Royal Society of Chemistry, Sage, Taylor and Francis, Wiley and Springer.The offsetting agreements where described as pilots and should be seen as tentative.Earney emphasized the need to monitor the changing environment and keep track of the effectiveness of the agreements in place.From their experiences, Jisc Collections has also formulated guiding principles for offsetting agreements.
Jisc Collections principles for offsetting systems should, according to Liam Earney: • support the transition to full open access • ensure that publishers do not charge the same institutions twice, through the payment of subscriptions and APCs • not be restricted to institutions that subscribe to a publishers "big deal" • apply at the level of subscribing institution • operate on a "cash basis" , rather than a voucher basis • be available at the institutional level rather than just the individual researcher Earney also pointed out, that possible schemes are reviewed with regard to the following: • Degree to which they support the transition to OA • Affordability • Ease of administration • Transparency If and when there are tensions between the different aspects, the degree to which a scheme supports the transition to OA is seen as most important.
Earney also emphasised that the success of offsetting schemes and the future of hybrid open access are closely linked, and one of the challenges is hybrid publishers refusing offsetting agreements.Springer has set an APC cost of €2.200 per published paper and the cost will be offset against the licensing fee for the institution.This model has some attractions for libraries and universities.

Springer Compact
• It would give access to Springers large portfolio of journals for both reading and OA publishing with a minimum of administration and invoice handling by the libraries and researchers.
• It offers a model for how it could be possible to shift funding from subscriptions to APCs gradually, without disrupting access to the subscription-based publications.Ritt/Springer believes that the OA publishing and the subscription model will run in parallel for at least the next decade.
Although there are advantages to this model, there are also potential problems.It extends the "Big deal" to OA publishing and will have the drawbacks of that model.A large portion of the institutional resources spent on scientific communication will continue to be locked up in agreements with a handful of major commercial publishers, if there is a general uptake of the model.This means that the resources available for scholarly publishers with few journals and new experimental ways of disseminating scientific results will continue to be scarce.The model would most probably keep the revenue streams as they are today, letting the major commercial publishers cling on to the driver's seat, making it possible for them to maintain their net-profit margins.Then they contacted the pilot institutions who were willing to pay the APC, who in turn contacted the authors to get permission to make the publication open access.Despite the supposedly strong incentive the RCUK mandate brings, the result was not that impressive.Out of 220 potential candidate papers published by IOP, only 22 where published open access in the pilot.Participating institutions were willing to publish 43 papers, but only 25 authors signed the required form to make the paper open access.

IOP Publishing
Steve Hall identified three major challenges in the first year of the pilot: • Identifying articles emanating from RCUK-funded research, which was something the universities could not do.• Providing information to relevant authors to encourage them to apply for funding from their institutions.• Varying open access policies and processes at the participating institutions.
Steve Hall and Juliane Ritt both stressed that these are experimental pilot projects and that the future direction is far from certain.

OA funds in Norway
Moving on from offsetting models to publication funds, Jan  (NRC) has set aside 8 million NOK (≈900 000 EUR) to refund up to 50 percent of institutions' APC costs.The rules for applying are the same as for the local ones.The HEI must have a local fund to be eligible for refunding.It is not tied to NRC-funded projects and sufficient funding will be set aside for the duration of the project.After 2019, APCs are expected to be part of general overhead costs at the institution, as subscriptions are today.

OA funds in Germany
Margo Bargheer, University Press Göttingen, also talked about a development similar to the Norwegian.Deutsche Forschungsgemeinschaft (DFG) is running a funding program that started in 2009 and will continue until 2019 with the aim to support German universities in setting up central funds for open access payments.The rules are similar to the Norwegian, with the addition that there is a price cap per APC of €2,000.More than 20 universities in Germany run open access publication funds with DFG funding, and approximately 20 universities run a fund without being part of the DFG funding program.Like the Norwegian publication funds, the German funds do not support publication fees in hybrid journals.
Margo Bargheer also spoke about experiences from the local fund at the University of Göttingen and the problems that arise when usage really takes off.The allocated budget for 2014 (€183.860)was overdrawn by €66.884.So, currently there is a challenge to model the fund for the future to handle the costs.Some of the ideas presented were: co-funding by authors and/or faculties, through individual research projects, and maybe most importantly, maintaining strict price caps per application.

TRANSITION TO OPEN ACCESS -STANDARDS
Although we have come far in implementing open access for publications, there are still ways of making access to research simpler, according to the keynote speaker Kaitlin Thaney from Mozilla Science Lab.We still face challenges in evaluating research and we can still develop the ways we reuse research and data.We should look beyond open access and think about usability, argued Thaney.
Considering the use and reuse of data, the initiative JATS4R (JATS for Reuse) presented by Kaveh Bazargan, seems important as it aims to improve XML to enhance reusability.JATS4R stands for Journal and Archiving Tag Set for Reuse and was formed to provide guidelines and tools to standardize the tagging of XML in publishing workflows.Many open access publishers provide article information in XML, but to become reusable, the content must be correctly tagged.This will increase accessibility, but also possibilities for text mining and future proofing.The JATS4R validator can help you check your XML content, and is an initiative supported by a number of publishers. 10Kaveh Bazargan showed an example on how simple pagination (start-page/end-page), something you might think is fairly unambiguous, might cause problems.The following example shows how the same start-and end-page can be tagged in different services: Tagging 1.

EVALUATION
The problem of 'predatory' publishers is constantly under discussion.To support researchers when choosing a journal for publication, the 'Think.Check.Submit' initiative could help researchers to identify trusted journals.The initiative, presented by Lars Bjørnshauge, is led by several organizations involved in publishing as well as individual publishers. 11The tool is a simple checklist of things you should consider when submitting to a journal and an attempt to help researchers to determine whether a journal or publisher could be trusted.This is a suitable approach considering that new journals are launched every week and hopefully the guide will become a useful tool to researchers.
Moving on to research and researcher evaluation, Stephen Curry discussed the problems with measuring scientific output.Curry is a professor of structural biology at Imperial College London and also a member of the steering group, assigned by the UK Minister for Universities and Science to assess the role of metrics in research assessment and management, reported in July 2015 "The Metric Tide". 12he main findings from the report could be summed up as using responsible metrics: • Metrics do inform, but do not replace judgement.Metrics should be used as indicators.
• Institutions need to be transparent about use, with clear statement of principles for assessment and in dialogue with staff.• Data used need to be transparent, which is a challenge to providers.
The responsible metrics should build on "The Leiden manifesto for research metrics", which summarizes ten principles to guide research evaluation 13 .Stephen Curry discussed in his presentation the importance of rethinking the definition of "good research" and how we measure good research.He described good research as transformative, complex and multi-dimensional.He also questioned how well we measure research when we use impact factors, citation counts or h-index.The present culture of metrics sustains traditional publishing.The advantage is that the high bar to publish in "top" journals spurs competition, but according to Curry there is a list of disadvantages in today's model that outweighs the advantage.
Stephen Curry listed following downsides with the existing model for metrics: • But how do we get the incentives for a change right?Funders could reward speed, openness and content, not "the journal wrapper", i.e. the journal impact factor.Furthermore, we could evaluate researchers based on other academic activities and not only on journal impact factors.It is important that we de-throne the impact factor and instead try to publish citation distributions.

CONCLUSIONS AND SUMMARY
From a library perspective, the COASP conference is both relevant and of great interest.Libraries are getting increasingly more involved in the process of scholarly publishing, with competences in open access-publishing as well as in infrastructure and workflows.Attending the conference is a good way of getting updated on current issues related to the open access publishing environment.
When we look back at previous conferences, we can clearly see some trends.First, the conference target group has been slightly shifting from small publishing societies, special interest groups, university departments and libraries to an increasing presence of the professional traditional publishing organizations.This includes both participants and presenters.Second, the proportion of participating librarians seems to decrease for each conference.Possibly one of the reasons for this is the increasing involvement of the large commercial publishers, both in the conference program and number of attendees.The conference might be perceived more as a publisher's trade meeting than an open access event from a library point of view.Libraries are still largely divided in different units or departments when it comes to working with open access issues.On one hand there are librarians with competence in how to publish open access and working with institutional repositories and on the other hand we have the librarians with a long experience in dealing with commercial publishers, making agreements on subscriptions for e-resources and "big deals".In general, we think that COASP nowadays would be just as interesting to the e-resources librarian as to the librarians working with open access-issues.Some interactions between the two are on the rise, at least in Sweden, but they are mostly still in the very beginning of something that might lead further.It would be interesting to use NOPOS to explore what is happening in the Nordic countries in this area.We invite the National consortia leaders to write a piece each on what's happening and what they would want to happen!As we have outlined in this summary, several interesting initiatives aiming to make the transition to open access as sustainable as possible (and as economically sustainable), are taking place.Both academic publishers, research funders and other stakeholders are trying to move to a more open and transparent way of publishing.Finally, a few words from some of the presenters to summarize from this years' conference.Dekker and Frantsvåg both pointed out that information and manpower support are important High bar to entry slows publication (rejection in one journal forces the author to submit to a new journal, which is time consuming) • Reduces researcher productivity • Conservative peer review • Eye-catching research trumps quality?• Impact factor-based rewards foster cheating • Cheating undermines public trust • Restricts access • Poor fit to public policy in the digital age When re-imagining scientific communication in an open era, Stephen Curry sees a universal pre-print culture, i.e.Open-Access Mega journals "with PLOS-style (which means sound science is enough) peer-review" and open access to underlying data.Openness and public scrutiny would work as an added quality control and with the underlying data also open it should discourage scientific fraud.
Juliane Ritt presented Springer's offsetting model Springer Compact, with the goal to support the transition from a subscription model to a gold open access model.The Compact model combines open access publishing in Springer's hybrid open choice journals and gives full access and archivalrights to all subscription-based and licensed journals on SpringerLink, Springer's online delivery platform.The Springer Compact is a traditional "Big Deal" combined with a flat-fee that give authors from the institution the possibility to make their papers freely available as gold (or hybrid) open access without any additional APC-payments.The total institutional cost, besides a reduced "big deal" cost, will be based on the number of papers published earlier in Springer journals by authors from the institution.
Steve Hall from IOP Publishing, part of the Institute of Physics, presented their take on a transition from subscriptions to open access publishing.The IOP model is just like Springer Compact based on using the hybrid journals as the way to make the transition, although not based on a big-deal subscription model.Instead, APC payments for hybrid open access papers are offset against both local and global subscription prices.At low level of hybrid, bulk of offset is local.At higher levels the global subscription prices are affected more and the local less.IOP is running a pilot with UK universities in 2014-2016.Steve Hall reported the results so far from 20I4.IOP began the pilot with identifying RCUK-funded research publications where open access was mandatory.
Erik Frantsvåg from UiT the Arctic University of Norway, presented the state of open access in Norway.In Norway there is a growing number of higher education institutions (HEIs) with local open access policies.There are 17 central publication funds for APCs run by HEIs, which is way ahead compared to Sweden. 9In 2013 the Norwegian Government formulated a white paper requiring open access to publicly funded research.The white paper referred to the UiT fund (established 2010) as a good example and added a recommendation that other institutions should follow.It seems that the recommendation had an effect, since the rules for applying for funding are more or less similar between funds.To receive financing for publication fees the corresponding author should be affiliated to the HEI.The journal, that the author plans to publish in, should be registered in DOAJ and accredited in the Norwegian HEI financing system.Publications in hybrid journals are not financed by the publication funds.A problem with this, which also recurred in other presentations, is that the accreditation system for researchers favors publishing in established, subscription-based journals.So, as in many other countries, the researchers are caught between two conflicting messages from the same sender: on the one hand demands to publish open access, and on the other hand demands to publish in traditional non-open access high-impact journals to get research funding.The Norwegian research council has an open access requirement, first adopted in 2009 and later revised in 2014.The revised policy emphasizes gold open access and the Research Council has as a consequence established a new funding scheme to boost publishing in open access journals.This national scheme for funding support was implemented in 2014 and will run until 2019.It aims to support the transition to gold open access, but bypassing the hybrid model.The Norwegian Research Council