This article is based on a talk I gave at Computers in Libraries in March 2019 (computersinlibraries.infotoday.com/2019/program.aspx#12685). It is not a comprehensive overview of the state of open access (OA)—this isn’t possible given that, almost weekly, OA agreements are signed, subscriptions are abandoned, and new models are attempted. Rather, it is a small-scale exploration, via examples from my home institution, Massachusetts Institute of Technology (MIT), and elsewhere, of two current approaches to moving the scholarly publishing system further toward open-publisher agreements (and disagreements) and academy-owned publishing.

My underlying assumption is that OA is good—that literature (typically journal articles), which is immediately, freely available on the public internet, improves the speed of research, is more equitable, and will help keep publishing costs from growing. That said, how best to make OA happen is unresolved. OA may have begun because the web allowed for easy sharing, but it has evolved into a complex movement with political, social, and economic dimensions.

SUBSCRIPTION-BASED MODELS

The scholarly publishing system is still largely subscription-based. It is:

• Expensive, as Ralf Schimmer, Kai Karin Geschuhn, and Andreas Vogler mention in their white paper, “Disrupting the Subscription Journals’ Business Model for the Necessary Large-Scale Transformation to Open Access” (pure.mpg.de/rest/items/item_2148961_7/component/file_2149096/content).
• Continuing to rise in cost, according to STM association reports (stm-assoc.org/2015_02_20_STM_Report_2015.pdf; stm-assoc.org/2018_10_04_STM_Report_2018.pdf).
• Dominated by three to five publishers responsible for about half of all papers published (including Elsevier, Springer, Wiley, and Taylor & Francis).
• Closed off to many people, because access is limited to those who can pay for it.
• Restrictive to authors, who give their articles freely to publishers and also turn over many or most rights, which often limits what they can do with their work.
• Hard to extricate from. Researchers want to publish in their favorite journals, where their work will be widely read and cited. They want to maintain familiar workflows and access content in existing journals. They also want to reuse and share their work. And they don’t necessarily like or trust change. So how can we make change anyway?

Variations in the type of OA that will help “fix” the system are at the heart of debates among researchers, funders, librarians, and publishers. The current fixes have their own problems. For example, although fully OA (“gold”) and partially OA (“hybrid”) journals proliferate—thus putting an increasing number of freely available works into the world—for-profit publishers in particular have benefitted from income brought in by the OA fees for these articles (often article processing charges, or APCs). A study in the U.K. found that APCs for articles in hybrid journals are higher than those in fully OA journals (“The ‘Total Cost of Publication’ in a Hybrid Open-Access Environment: Institutional Approaches to Funding Journal Article-Processing Charges in Combination With Subscriptions,” Stephen Pinfield, Jennifer Salter, and Peter A. Bath, Journal of the Association for Information Science and Technology, v. 67, issue 7; doi.org/10.1002/asi.23446). APCs are seen by many as further propping up the commercial publishing industry and putting low-income countries and scholars at a disadvantage.

A more common approach to OA in North America has been article self-archiving (“green” OA) via OA policies. (MIT faculty members have had an OA policy since 2009.) “Self”-archiving is a misnomer in that it is largely library staffs who do the work—and this is an issue: It takes a lot of time and effort to figure out what authors have published; search for these papers in the correct version—that is, the peer-reviewed manuscript; and deposit them into an institutional repository. Though authors may be philosophically in favor of OA, or even OA champions, they are busy people who, en masse, are difficult to motivate to participate in making their work openly available. Only about 15% of articles in the OA collection of MIT’s institutional repository, DSpace@MIT, come directly from authors.

According to a 2018 analysis on OA publishing in the journal PeerJ, between 28% and 45% of the world’s scholarly articles are OA via gold and green methods (“The State of OA: A Large-Scale Analysis of the Prevalence and Impact of Open Access Articles,” Heather Piwowar et al.; Peerj.com/articles/4375). (At MIT, about 45% of faculty papers published in the last decade have been OA.)

Clearly, there is some work to do.

Publisher agreements (AND disagreements)

Gold, or publisher-made, OA has been a priority in Europe for years. A 2012 report commissioned by the U.K. government, commonly known as “The Finch Report,” concluded that national funders should “prefer” gold OA over green (acu.ac.uk/research-information-network/finch-report-final). Funders currently offer grants to institutions to cover APCs. Also in 2012, the European Commission (EC) announced Horizon 2020, an € 80 billion funding program that requires grant recipients to make articles OA (ec.europa.eu/programmes/horizon2020/en). The EC is reimbursing APCs for funded authors. Because European countries have thrown the weight of government funding for research behind gold OA, there has been a lot of interest in making sure this money is spent wisely. Many have found this is not the case, and they’ve begun to push back on publishers.

In the last several years, consortia of libraries, universities, and research institutes in Germany, the Netherlands, Finland, Austria, France, Sweden, Switzerland, and the U.K., among others, have negotiated for license agreements that make their researchers’ papers more openly available for a reasonable cost. What each consortium wants is slightly different, and there has been varying levels of success. But, in general, consortia want a deal with a publisher in which articles published by their researchers are made OA in the final published version, often under a Creative Commons license. They want to pay APCs that aren’t obscenely high. They sometimes want to combine the cost for the subscription with the cost for APCs—a transformative agreement called “read and publish.” MIT signed North America’s first Read and Publish license last spring with the Royal Society of Chemistry (RSC; libraries.mit.edu/news/royal-society-chemistry-3/27769). The RSC wants “sustainable costs,” which can mean subscription fees that don’t get higher at all, or that increase less—or it doesn’t want to pay subscription fees and instead only pay for OA publishing.

The latter is what the German consortium Projekt DEAL wanted from Elsevier. The company balked, and in summer 2018, Projekt DEAL suspended talks with Elsevier after failing to reach an agreement (nature.com/articles/d41586-018-05754-1). Other stalled agreements from the past year include Sweden (nature.com/articles/d41586-018-05191-0) and Hungary (eisz.mtak.hu/index.php/en/283-hungarian-consortium-terminates-negotiations-with-elsevier.html) with Elsevier; the Netherlands (utwente.nl/en/news/!/2018/3/79107/no-new-agreement-with-royal-society-of-chemistry-and-a-new-agreement-with-springer) with the RSC; and France (the-scientist.com/daily-news/french-universities-cancel-subscriptions-to-springer-journals-29882) with Springer Nature. By not renewing contracts, researchers from these institutions no longer have access to journals from these publishers.

Some consortia have reached agreements: Projekt DEAL (projekt-deal.de/wiley-contract) with Wiley, and Finland (finelib.fi/finelib-and-elsevier-agreement-access-to-scholarly-journals-and-50-percent-discount-of-article-processing-charges) and Norway (insidehighered.com/news/2019/04/24/elsevier-agrees-first-read-and-publish-deal) with Elsevier—the latter is Elsevier’s first read and publish agreement. But there is now momentum behind institutions and consortia walking away from subscriptions if they don’t get at least some of what they want.