Rocky road to open access

Share this on social media:

Topic tags: 

As open access policies come into force, UK higher education institutions are racing to meet the mandates. Rebecca Pool reports

It’s no secret that technology has thrown scholarly publishing into a state of flux. The digitisation of content, easy-to-use software, and the ever-growing opportunity to spread your words across the Internet has prompted academics and publishers to reassess how research is published.

For many in the UK, open access is the answer, as indicated by recent requirements from the nation’s two most significant providers of public funding, Research Councils UK (RCUK) and the Higher Education Funding Councils for England (HEFCE).

In short, each has recently introduced policies for UK research organisations to make published outputs openly accessible. However, the road to open access is anything but short.

‘We always said this would be a journey and not an overnight switch,’ highlights Mark Thorley, chair of the RCUK Research Outputs Network and head of Science Information for the Natural Environment Research Council.

‘We’ve seen issues over the cost of implementation, administrative processes and also policy compliance reporting,’ he adds. ‘But we’re also seeing an increasing volume of research available as open access, so the policy is making a difference.’

For Thorley, open access is a no-brainer. With a twitter account called ‘My life is Open Access and Open Data’, he points out how in today’s networked world, anyone can publish, literally anything. And, in his view, this makes instant access to peer-reviewed research more important than ever before.

‘Those in the research process have the responsibility to ensure that quality, peer-reviewed research is widely available to all who need it,’ he says. ‘Otherwise the void will be filled by make-believe and half-truths.’

So, for his part, Thorley is spearheading RCUK’s open access policy, which was introduced in 2013, a year before HEFCE announced its policy on the same. Each organisation’s policy aims to make research arising from its funds widely and freely accessible, although in practice, RCUK policy takes a firmer stance.

The Research Councils’ mandates aim to achieve immediate, unrestricted, online access to work funded by the UK research councils. Meanwhile, HEFCE’s approach requires that certain research outputs submitted to any research assessment exercise after 2014 be made as widely accessible as possible [see ‘RCUK and HEFCE Policy on Open Access’]. Exceptions exist, where, for example, publisher requirements mean research outputs cannot be deposited within policy obligations.

But firm or not, the two policies are deemed to be complementary and two years on, a lot has happened. For starters, the research councils are emerging from the first independent policy review, carried out by former vice chancellor of the University of Leicester, Bob Burgess.

Aiming to assess the impact the OA policy is having on education organisations and publishers, the review outlined key areas that needed more support or attention. These included administration and implementation costs, compliance monitoring as well as communication between the research councils and communities, and confusion over creative commons licence requirements.

At the same time, HEFCE’s policy – which comes into force April 2016 – has recently been amended to reflect research community concerns. For example, while the policy requires that research outputs be deposited in a repository upon acceptance for publication, information systems are not yet in place to do this, so this requirement has been shifted back to April 2017.

As Thorley highlights: ‘Many issues arise from trying to make this open access policy work at scale... we’re gearing up the whole sector to ensure the majority of papers are available in the open access corpus.

‘So the biggest issue has been how do you turn a practical open access policy into something that actually works, is do-able and doesn’t take up endless resources,’ he adds.

Costs count

Without a doubt, costs have been a key sticking point to the implementation of both the RCUK and HEFCE open access policies. Late last year, UK-based research funding and policy consultancy, Research Consulting, unveiled its report, Counting the cost of Open Access.

Commissioned by London Higher and SPARC Europe, the report outlined the total cost of implementing each policy, and the figures weren’t small.

The total cost to implement RCUK open access policy in the 2013/2014 academic year came in at £9.2 million. But factor in a £11 million expenditure on article processing charges (APCs) and this figure rose sharply to some £20 million.

Meanwhile the cost to the higher education sector of meeting HEFCE’s deposit requirements of the post-2014 Research Excellence Framework was estimated at £4m to £5m per annum. These figures excluded the costs of management, advocacy and infrastructure development, expected to be on par with RCUK figures for the same.

Interestingly, the report also shed light on the administrative costs of making articles open access through ‘gold’ and ‘green’ routes. Following gold was estimated to cost £81 per article and take two hours while green open access would cost £33 and take just over 45 minutes (see Green or Gold?, overleaf).

Jisc is tackling the issue of OA costs head on. The UK-based provider of research software has been heavily involved with open access policies from the beginning, collaborating with myriad research organisations to create tools and services to manage OA processes efficiently and cost-effectively.

Crucially, following the RCUK and HEFCE policies, Jisc also set up its ‘Open Access good practice initiative’ to help institutions comply with OA policies. Here, the company works with UK universities, across a series of so-called Pathfinder Projects, to discover what works, and what doesn’t (see ‘Meet the Pathfinder projects’).

As Sarah Fahmy, manager of the Jisc’s OA good practice initiative, highlights: ‘One of our main outputs is there really is no standard one size fits all for these institutions.

‘Sometimes institutions want a straight answer to the issues and we’re not quite at that point yet, but we have lots of examples of good practice,’ she says.

According to Fahmy, a pathfinder project headed up by Bath University set out to unravel the costs associated with managing a gold route for RCUK-open access funded research. The team created a functional cost analysis of the workflows and related labour costs for this route’s article processing charges (APCs), identifying areas of inefficiency.

‘Other institutions can now look at this and see whether or not they can replicate, or streamline, their processes around this,’ says Fahmy.

Likewise, a second project led by Northumbria University has developed a shareable cost modelling tool that institutions can use to model difference OA scenarios to make more informed decisions on funding policies. ‘Each pathfinder project is trying steadily to put together a piece of this puzzle,’ adds Fahmy. ‘Levels of open access implementation vary from institution to institution... and this programme is really trying to find something for everyone.’

Indeed, as other projects reveal, costs are not the only issue. So far, the Pathfinder projects have thrown up several key issues including difficulties in identifying accepted articles, a key step in the OA process flow now notoriously known as ‘point-of-acceptance’. Other issues include monitoring compliance, tracking deposits in subject repositories, uncertainties over audit requirements including what exactly HEFCE means by an exception.

‘A few institutions raised the issue that they didn’t know what number of exceptions were acceptable,’ highlights Fahmy. ‘HEFCE has since updated policy to say this isn’t an issue, but we have a pathfinder project, led by University College London, that’s looking at how institutions can evidence a good approach to logging an exception.’

No doubt as the open access journey continues, pathfinder projects will continue to chip away at the issues, and for its part Jisc is also developing a dizzying array of packages to support OA repositories and more (see Jisc open access projects and services).

For example, so-called Sherpa services provide help at the point of submission. Sherpa RoMEO provides an analysis of publishers policies for authors’ rights when using OA repositories, while Sherpa RED will advise on compliance.

Further along the open access process, Jisc Publication Router will direct content to the appropriate repository while Jisc Monitor will help institutions to track OA publication processes, especially with reference to compliance and costs.

And come pubication, RIOXX Guidelines and Application Profile provides a mechanism for institutional repositories to comply with the RCUK policy on open access, supporting the consistent tracking of OA publications across scholarly systems. CORE also provides the text and data-mining interface for all open access literature from repositories and some journals.

As Fahmy’s colleague, Neil Jacobs, head of scholarly communications support at Jisc, puts it: ‘Take RIOXX, I think it’s fair to say that before we developed this, most repositories encoded information in a fairly idiosyncratic way and there were interoperability issues.’

‘We hope [the tool] will standardise this, and also help with other policies, such as the European Commission Horizon 2020 programme,’ he adds. ‘From an institutional point of view, there is a lot to open access and it really matters to institutions that they get it right.’

For his part, Jacobs is confident that the tools and services are coming together so institutions can begin to comply with OA policies, more seamlessly: ‘We have compliance, interoperability and issues such as these, so there’s work to do, but it’s not an amazing amount of work,’ he says.

Case-in-point, a recent independent study from several funding bodies including RCUK and the Wellcome Trust, on Jisc’s Sherpa/Fact  tool, revealed positive results. This funders and author’s compliance tool was found to be more than 95 per cent accurate when checking publisher policies against funder mandates for open access.

Such results indicate that step by step, progress on fulfilling the latest OA policies is being made, and as the RCUK’s Mark Thorley asserts: ‘This is all part of what research means in an Internet age.’

‘We have to make the outputs of our research more easily accessible,’ he adds. ‘Come 2020,  we’ll be a lot further down the road and my hope is innovative new business models to disseminate research will have started to come through.’   

RCUK OA policy: http://www.rcuk.ac.uk/research/openaccess/

HEFCE OA policy: http://www.hefce.ac.uk/rsrch/oa/Policy/

JISC services for OA: https://www.jisc.ac.uk/open-access

Green or gold?

From word go, the 2013 RCUK policy has leaned towards the gold open access route, rather than the green route. Indeed, if a researcher follows the gold route, he or she publishes the final version of work in an open access journal, which is freely available via the publishers’ website immediately.

In contrast, the green route, or self-archiving, sees the researcher publishing his or her work in a journal and then depositing the final peer-reviewed manuscript in an institutional repository, such as ePrints, or central repository, such as arXiv or PubMed Central. What’s more, access may be subject to an embargo, depending on publisher policy.

Clearly the two are quite different, and as RCUK Research Outputs Network chair, Mark Thorley puts it: ‘Anything with an embargo period is not open access, its delayed access.’

‘In the world we live in we should be looking for research that is open and immediate access, and re-usable as of the date of online publication, now six, twelve or twenty four months down the line.’

And while many favour green open access, experience indicates this route demands a sturdy infrastructure between publishers and repositories to process peer-reviewed manuscripts. Still, the gold route demands article processing charges (APCs)  in the region of £1,000 to £2,000, that do not apply to green open access.

However, the RCUK can provide funds, and as Thorley emphasises: ‘A green repository model that still relies on subscriptions is just a transition mechanism. We have to find a more robust and open way of dealing with access.’