The internet has transformed the nature of scientific research. It has opened up new ways to collect, use and disseminate scientific information, leading to increased demand information. Open Access (OA) to scientific journal publications means making them freely available online, rather than charging readers to view them. OA to research data means making research data more widely available for re-use by others to support research, innovation and wider public use.
The growth of the World Wide Web in the 1990s has brought new opportunities for disseminating scientific publications. Scientific publishers have made substantial investments in online delivery and in digitising old content. Around 95% of journals are available online now. OA movement aims to make scientific publications freely available online, in a hope that there will greater transparency and more effective exploitation of public funds, and support to research and innovation.
Expanding access to scientific publications and data could eliver widespread social and economic benefits. In March 2011 the Minister for Universities and Science held a round table discussion to explore the issues. Commitments were made to supporting efforts to expand access to both research publications and data as part of its wider “Transparency Agenda”. OA is seen by many as a key option for expanding access. This POSTnote will examine separately the challenges facing OA movement and its benefits.
C. Open Access to Journal Publications
Scientific journals are a key channel of dissemination of scientific information. Journal publishers administer the peer review process before they produce, distribute and archive printed and digital editions of journals.
The “Subscription Model” of Publication
Traditionally publishers charge for their services through subscription fees to libraries and individual users. In the past decades, journals subscription fees have increased above inflation but library budgets have decreased, making access to this information more of a challenge. Following the recommendations of the 2004 House of Commons Science and Technology Committee, researchers, librarians, higher education institutes (HEI), funding agencies and publishers have been working together to improve access to publications. Although big progress has been made, there is still much debate as to whether the subscription model is the most effective way to disseminate research for maximum access and impact.
Some researchers, funding bodies and HEIs have argued that the cost of some journal subscription is compromising the UK’s capacity for research. They say that researchers in smaller UK universities do not have access to the publications they need, while the subscription model does not deliver adequate access to researchers in countries with lower income levels. However, many subscription publishers argue that the subscription model is highly competitive and efficient and that UK research libraries could make efficiency savings elsewhere, to accommodate subscription costs. They also say there are already several initiatives to expand access, such as provision of free or low cost access online access to developing countries.
In Sep. 2011 the government set up an independent working group to explore how expanded access might best be achieved and the result is expected to be published in 2012 spring. Many researchers, funding bodies and HEIs argue that OA could offer a key way to expand access and increase the impact of research.
Why Open Access to Publications?
OA proponents argue that the knowledge emerging from publicly funded research should be made freely available to society. It is also argued that free online access to publications would lead to social and economic benefits, resulting from increased communication and translation of knowledge. A recent study suggests that a shift to OA could increase the return on R&D by between £184 and £386m over a 20 year period. This reflects the increase in access provided to government, the NHS, corporations and small to medium enterprises, as well as universities.
OA has seen its growth in the UK and worldwide. Around 20% of all articles were made available through some form of OA in 2009. An increasing number of UK research funding agencies including the Wellcome Trust and Research Councils UK (RCUK) and many HEIs have now adopted OA mandates. Theses require researchers to make their published articles freely available online. Most subscription publishers have incorporated OA models in their business to allow authors to meet funder and HEI requirements. The main challenge of OA is seen to be how to meet the costs of publication, particularly the administration of peer review, on a sustainable basis.
Models of Open Access to Publications
There are two main models of OA and it is believed that there will be a mixture of the two alongside the subscription model for the foreseeable future.
• “Green” OA: the author deposits the “post-print” version of articles accepted for publication in an online repository which makes them freely available. Some subscription publishers specify embargo periods of 6-48 months, during which time the articles are available only from the journal. This is to protect their subscription income and secure a return on their investment.
• “Gold” OA: with an article processing charge (APC), published articles are made immediately and freely available by the journal publisher. Many subscription publishers offer “hybrid” models for some of their journals, under which authors can choose to make their articles OA by paying an APC.
OA can simply mean making articles free to read online, but it usually also means removing some of the copyright restrictions on the reuse of that information. The 2011 Hargreaves Review on Intellectual Property (IP) reported that some aspects of current IP law relating to publications were obstructing innovation and economic growth. One example is the placing of restrictions on “text-” and “data-mining”. With OA, such restrictions are removed.
Sustainability of “Green” and “Gold” Models
With green OA the costs of publication continue to be paid for through journal subscription fees, but anyone can access the publication via repository, regardless of whether they subscribe to the journal. The cost of HEIs could rise slightly because of the need to fund repositories alongside paying for journal subscriptions. However, proponents say that this is outweighed by benefits to the UK through increased returns on R&D. However, to maximise the benefits, there is pressure on publishers to make embargo periods as short as possible. Some subscription publishers argue that this could also lead to significant levels of subscription cancellations, which could reduce revenues to the point at which some publications become non-viable. Many journals are experiencing cancellations, although it is uncertain whether these cancellations are due to OA repositories. EU funded research is looking into this issue.
Under gold OA publishing, the costs of publication are shifted from the reader to the author, in the form of an PAC, which is usually paid by the author’s funding agency or institution. This is seen by some publishers and by RCUK as providing a more financially sustainable business model compared with green OA. However, the one-off cost to the UK of a transition to gold OA is estimated to be in the region of £7m, because of the need for HEIs to create central funds to meet APC charges, and for funders, HEIs and publishers to develop administrative systems to deal with APCs.
Nevertheless, great uncertainties exist – how gold oa will evolve and whether there would be a net decrease or increase in publishing costs in the UK under wholly gold OA system. HEIs would be expected to reduce journal subscriptions and divert the savings into paying APCs. However, savings are hard to calculate as there is significant variation in APCs between journals. The average APC is around £1,500 but they can be as high as £3,000. A recent report suggests that if the average APC is under £2,000, costs to HEIs would stay the same; at the lower level there could be significant savings to HEIs, but at the higher level, the cost to HEIs would rise significantly above the current subscription model.
Subscription publishers argue that APCs for high-status journals must reflect the costs associated with handling large numbers of rejected papers. However, some funding agencies, libraries and OA publishers said that subscription publishers calculate APCs on the basis of current profit margins rather than costs incurred. Some worry that “hybrid” models can lead to “double dipping”, whereby institutions are paying twice, for APCs and for subscriptions, although subscription publishers said that subscription costs will reduce as the proportion of gold OA articles grows.
D. Open Access to Research Data
Sharing research data more widely comes with many benefits. In addition to enabling errors or flawed research to be detected, it also deters selective reporting of results. The recent legislation Freedom of Information (FoI) has transformed the degree of openness. Under this law, HEIs and public research institutes are required to share certain information if requested by the public. Sharing of some private sector research could also be in the public interests, such as data from clinical trials and safety analyses which are used to inform policy decisions (e.g. 2011 oil spill in the Gulf of Mexico).
Making Data Open
Timing of Data Sharing and Intellectual Property
Many funding agencies and HEIs argue that there is a balance to be struck between making data openly available for wider re-use and allow universities to exploit their IP. Because this balance varies across disciplines, funding agencies’ requirements on when data should be shared vary greatly.
HEIs have expressed concern that FoI legislation could enforce data sharing at too early a stage, which could hinder research and commercial collaboration. They have also raised concerns over the inclusion of research information in the FoI Act, suggesting that there needs to be an exemption to protect ongoing research, as there is under the Scottish FoI Act. Whether there would be an amendment to the legislation will not be known until later 2012.
There are tensions between sharing data openly and protecting privacy, especially in the biomedical and social sciences. The Data Protection Act 1998 (DPA) regulates the processing of “personal data” through restrictions on how the data can be recorded, stored, altered, used or disclosed. The DPA implements the EU Data Protection Directive 95/46/EC, which is being revised to take account of new technologies and the changing ways that personal data are being used.
Under the DPA, personal data means data related to a living individual who can be identified, either directly or indirectly, from the data, or from other information held by the same organisation. The only legitimate way of using personal information for a purpose different from that for which it was originally collected is by obtaining consent. However, if it has been anonymised (personal identifiers or codes removed) data protection legislation does not apply.
Data Management and Arching
To enable others to find and to re-use research data effectively, data need to be archived and made publicly accessible, with sufficient “metadata”. The Environment Research Council and the ESRC have longstanding systems in place for sharing and archiving data. However, overall the systematic retention and archiving of data and “metadata” are patchy. To address this, funding agencies are introducing data management plans as an integral part of funding applications.
E. Cross-Cutting Issues
The Role of Publishers
In all types of OA, subscription and OA publishers both have a role to play in ensuring data are made publicly and permanently available. However, there is agreement that it is not their role to manage large datasets, and that the wider research community should take the lead in data management.
The Role of Funding Agencies and HEIs
Although most funding agencies and HEIs now have policies regarding OA to publications and data, compliance is acknowledged to be patchy. Some suggested that processes need to be simplified for researchers, communicating the benefits to researchers should be more effective and there should be clearer indication on budgetary and resource provision. Also, monitoring of compliance of the policies is necessary. There was also suggestion that data management and sharing should be recognised as a professional achievement, to encourage researchers to invest their time in making their datasets publicly available for re-use.
The Role of Researchers
OA movement impacts some disciplines more than other, so it requires a substantial shift in community attitudes and behaviour in certain groups. Although funder and HEI mandates could increase OA, there is wide agreement that a persuasive rather than punitive approach is preferable and the reward structures of academia could be adapted to incentivise researchers to adopt OA practices. The Higher Education funding bodies are considering how OA could feature within the new system for assessing the quality of research in the UK, although it will not become a formal part of the Research Excellence Framework until 2020.
• Open Access (OA) to scientific publications could provide more effective dissemination of research and thus increase its impact;
• It is necessary to consider the costs and benefits of different models of providing OA to publications if a comprehensive shift to OA is to be financially sustainable;
• OA to research data could enable others to validate findings and re-use data to advance knowledge and promote innovation;
• OA movement needs to be built on effective data management and archiving. It also presents challenges relating to protecting intellectual property and privacy;
• OA movement requires collaborative relationship between researchers, librarians, higher education institutions, funding agencies and publishers.