IPC on TwitterMy Tweets
The (unofficial) blog of the BCLA Information Policy Committee
By A. Christie Hurrell
Open access (OA) policies set guidelines – either voluntary or mandatory – for authors to make their published, peer-reviewed scholarly writings freely accessible online. Many universities around the world have developed OA policies for their institutions. Most of these policies involve archiving scholarly works in an institutional repository. This article outlines some of the factors prompting the development of OA policies, describes different types of OA policies, and outlines selected key stakeholders’ responses to them. The article is primarily designed to help librarians and other information professionals gain a basic understanding of OA policies in the university context, especially if they are to be involved with implementing or promoting an OA policy.
The impetus for OA policies
The development of OA policies has been prompted by numerous factors. A major issue for academic libraries has been the ‘serials pricing crisis’ of the past two decades, whereby average costs of journal subscriptions have increased exponentially, partially due to the consolidation of journal publishers.  Since library budgets have not increased at an equivalent rate to journal prices, the purchasing power of individual academic libraries has decreased, forcing them to cancel subscriptions, to reallocate budget items to maintain subscriptions, or to negotiate licensing agreements whereby access is granted to “bundles” of journals at a lower per journal price.  Open access represents another option for libraries to provide access to their institution’s scholarly output without the need for expensive journal subscriptions or potentially inflexible license agreements. Institutional administrators and policy makers may also see benefits to OA. By providing free access to their scholarly writing, they may be able to increase the visibility and impact of their institution. If the resulting OA collection is stored in a single institutional repository, an institution also gains a complete and easily-accessible record of its research output that can be used for promotion, analysis, and evaluation purposes.  Finally, OA publishing can increase the impact of research by making it available to a wider audience, especially to scholars and others in low- and middle-income countries where large serials budgets do not exist. 
Types of OA policies
There are two basic types of OA policies: voluntary and mandatory. The first type suggests or encourages researchers to make their work open access, while the second requires it. The Registry of Open Access Repositories Mandatory Archiving Policies (ROARMAP) is a searchable registry of policies developed by universities, research institutions, and funders. Currently, it lists 137 institutional mandates, 33 sub-institutional mandates, 1 multi-institutional mandate, and 14 voluntary policies – most of these developed by universities.  Since ROARMAP relies in individual institutions to self-register, it probably underrepresents the number of OA policies currently in existence. 
The first university-wide OA mandate was implemented in 2004 by the Queensland University of Technology in Australia.  Since then, there has been a clear growth in OA policies. They exist on every continent; a full list of countries with OA policies can be seen on ROARMAP’s website (http://roarmap.eprints.org).
Canadian funding agencies have also been active in developing OA policies. ROARMAP includes ten OA mandates developed by Canadian funding agencies, plus one proposed mandate.  Most funding agencies currently having OA mandates are involved in funding health and/or development research, including the Canadian Institutes of Health Research, Canada’s major health funding agency .
Library responses to OA policies
University library responses to OA policies have come primarily in the development, maintenance, and promotion of institutional repositories (IRs). IRs are web-based portals designed to collect, organize, and archive scholarly materials.  Due to their expertise in these areas, academic library staff are most often the ones leading the development of IRs, sometimes in conjunction with information technology specialists.  Libraries are also becoming involved with OA journal publishing: a 2007 survey of academic libraries by the Association of Research Libraries found that 44% of respondents were already offering publishing services, while another 21% were in the planning stages. Many of these services used open access models.  Canadian libraries have also been active in developing OA Authors’ funds, which provide subsidies for authors who wish to publish in OA journals that charge a publishing fee. A recent Canadian survey found that 12 out of 18 responding libraries maintained such a fund. 
Most major library associations in North America are involved in supporting and promoting OA initiatives. The Canadian Library Association’s Executive Council approved a position statement on OA in 2008. The statement encourages Canadian libraries to support and encourage OA policies, to raise awareness of OA among patrons, to support the development of OA both technically and financially, and to support and encourage authors to retain their copyright.  The Canadian Association of Research Libraries maintains a number of resources relating to OA, including tips on building an institutional repository, an OA bibliography, and links to recent OA-related news and events.  The Association of Research Libraries has developed SPARC, the Scholarly Publishing and Academic Resources Coalition, an international alliance of more than 800 institutions. SPARC focuses on education and advocacy around scholarly communications issues, including OA. It is also involved in incubating new business and publishing models that use OA principles.  The Association of College and Research Libraries has developed a Scholarly Communications Toolkit that is designed to help librarians learn about and effectively communicate numerous issues relating to OA, including authors’ rights, institutional repositories, journal economics, and publishing.  These resources have been developed, in part, because academic librarians have learned through experience that convincing faculty and administration to adopt an OA mandate and/or OA practices requires both evidence-based data on the advantages of OA, and strategic arguments that frame OA as a solution to a faculty/student problem, not just a library problem. 
Researcher responses to OA policies
The research literature on OA policies shows that the baseline level of self-archiving among university faculty is approximately 15%, and that voluntary OA policies do not increase this level of activity, even in an institutional environment that is very supportive of OA.  However, a 2006 study of seven Australian universities showed that mandatory policies, coupled with an effective author support system (defined primarily as supportive interactions with library/IR staff) result in the majority of an institution’s scholarly output being archived in an OA format.  An international survey of scholarly authors carried out in 2005 found that the majority of respondents (81%) would willingly comply with an OA mandate, while only 13% would comply reluctantly, and 5% not at all.  Other studies show that OA mandates result in a higher citation impact, which is generally a huge motivator for academic authors.  These data seem to suggest that an OA mandate, coupled with effective author support systems, will result in a complete institutional collection, and mostly satisfied faculty.
However, not all of the research literature supports this conclusion. Some authors have raised compelling questions regarding the effectiveness of OA mandates, arguing that they may undermine the peer review process or interfere with tenure and promotion practices.  These criticisms reflect deeply-held value systems among various academic communities, and as such are not likely to be erased by simply enacting an OA mandate (especially if that mandate is not the product of a democratic process). In addition, more recent research on the results of OA mandates at institutions have found no solid evidence for an increase of faculty awareness of OA or an increase in self-archiving as a result of an OA mandate.  Increases in the number of items deposited in an IR may have more to do with the activities of library staff than of researchers themselves, these studies argue.  Certainly, depositing materials in an IR takes time and effort, especially if authors must negotiate with traditional publishers and/or co-authors over permission to deposit. The time and effort required to deposit materials is a barrier for authors, and for IR managers, who are often struggling with a lack of staff and resources.  As such, these studies, which take into account the knowledge, attitudes, and practices of academic authors, as well as the realities of today’s academic libraries, emphasize that the construction of university OA policies must be tailored to fit the specific needs of each institution, and must be marketed and enforced effectively. In addition, librarians are urged to promote structural changes so that OA publishing is recognized in the academic reward system (e.g. tenure and promotion). 
Publisher responses to OA policies
Traditional commercial publishers have had varying responses to the introduction of OA policies at universities. Institutions implementing OA policies must be aware of publisher policies, in order to assist authors who are required/requested to self-archive, but still wish to comply with the policies of their journal of choice. The SHERPA-RoMEO project, based at Nottingham University, provides a searchable database that allows authors or support staff to discover the policies of various publishers.  SHERPA-RoMEO uses a colour-coded system to categorize archiving policies of various journals. Over half of the journals included in the database allow authors to self-archive some version of their article, for more information on different types of publisher policies and embargo periods, see the SHERPA-RoMEO FAQ page. 
Commercial publishers are understandably concerned about the impact that OA policies may have on their bottom line. They have formed advocacy groups, such as the PRISM Coalition and the Association of American Publishers, to lobby for the traditional model of publishing.  However, these groups have focused their energies primarily on OA mandates enacted by funding agencies such as the US National Institutes for Health, arguing that such mandates represent government intrusion into the publishing marketplace. Their lobbying efforts in the US have resulted in the introduction of the Research Works Act, which would repeal government funders’ OA mandates, and have ignited much controversy among academics and OA advocates.  Major publishers, including Elsevier, eventually withdrew their support from the Research Works Act, and it was declared “dead” in the US House of Representatives in February 2012.  Publishers have been less concerned with repealing the OA policies of individual institutions.
The movement towards OA policies and practices has been steadily growing in Canada and internationally for a number of years, and there are now many policy statements, resources, and research evidence to assist librarians who are charged with developing and implementing OA-related initiatives in their institutions (see the Further Reading section below for some helpful links). However, for any OA initiative to succeed, librarians and others must carefully consider the specific culture, needs, and attitudes of their institution (or of various communities within their institution). Librarians are urged to research these issues in advance of launching any new initiative, so as to build support and buy-in for OA policies and practices from the ground up.
32. Howard, Jennifer. 2012. “Legislation to Bar Public-Access Requirement on Federal Research Is Dead.” The Chronicle of Higher Education. http://chronicle.com/article/Legislation-to-Bar/130949/ (Accessed March 29, 2012).