The OA Switchboard

On December 6^th 2018, a group of stakeholders representing research funding organizations, academic libraries, scholarly publishers, and open infrastructure providers met in London to discuss a proposal for addressing the growing set of challenges in the implementation of institutional and funder policies supporting open access publication. The result of this initial stakeholder meeting was broad support for this initiative, tentatively titled the OA Switchboard, and in the weeks since this initial meeting the support for this initiative has continued to grow. What follows is an overview from Paul Peters of the key challenges that the OA Switchboard aims to address, a description of the proposed solution, and a roadmap for the development and initial roll-out of this new system.

Background

As a growing number of academic institutions and researching funding organizations have established policies to centrally fund the costs associated with open access publications, there has been a significant increase in the proportion of open access publication fees being paid by institutions under a variety of models. While this has undoubtedly helped to encourage open access uptake by authors at these institutions, it has also led to a far more complex landscape with a multitude of overlapping business models, policies, and systems.

The increasingly complex network of agreements between publishers and institutions, along with a rise in the number of policies being developed by research funding organizations, has already started to pose serious implementation challenges. Universities are having to invest more of their internal resources to manage the administration of open access payments and to report on how their open access funds are being used. Research funders are struggling to monitor and enforce compliance with the open access policies they have put in place. Publishers are spending an increasing amount of time in administering open access agreements with individual universities and consortia, often under a wide range of business models that each require their own workflows and systems. Moreover, individual researchers are often caught in the middle, unsure as to whether a journal they have submitted to is compliant with the requirements of their institution and funder, or whether there will be funding available to pay for the open access publication costs of their work.

The problems that have begun to arise in the central funding of open access publications are likely to grow in scale and complexity in the coming years. If successful, initiatives like OA2020 and Plan S will likely result in a rise in the number of open access publications being centrally funded, either by universities or research funders. Not only will this result in higher administrative costs for institutions, funders, and publishers, but it may also lead to a more pronounced imbalance in the ability of small and large publishers to compete on equal footing. Already there are signs that a handful of large commercial publishers will be best positioned to negotiate open access agreements with individual institutions and consortia, often as part of existing “Big Deal” subscription agreements.

Many smaller publishers, including scholarly societies and fully open access publishers, have been unable to negotiate these kinds of central open access funding agreements. Not only do these smaller publishers lack the internal resources to make and implement agreements with a large number of institutions, but they often struggle to get a seat at the table in these sorts of discussions. The total open access output from any single institution may only amount to a few articles each year for many smaller publishers, making it difficult for these institutions to devote their scarce time and resources to setting up open access agreements with small and mid-sized open access publishers. Unless a solution to these problems can be found, negotiated deals with a handful of large publishers may be the only viable option for funders and universities to support the transition towards open access, which is likely to result in a publishing landscape that is even less competitive, transparent, and inclusive than the traditional subscription-based publishing market.

An Intermediary to the Rescue?

The challenges described above are in no way unique to the open access publishing landscape. In fact, they are relatively common in marketplaces that have an increasingly complex web of interactions between buyers and sellers. The introduction of a central intermediary is often the easiest way to reduce complexity for both buyers and sellers, and many successful online marketplaces (Amazon, ebay, Etsy, Uber, etc.) have established themselves as the most efficient means of connecting a large and diverse ecosystem of buyers and sellers. Even within the traditional scholarly communications market, library consortia and subscription agents have served as intermediaries to reduce the costs and complexity of individual libraries negotiating agreements with individual publishers.

One important benefit that a central payment intermediary can provide is that it allows each participant in the system to directly interact with only one single counterparty, reducing complexity for publishers, institutions, and funders. Additionally, payment intermediaries benefit from the existence of a relatively straightforward business model, in the form of modest transaction fees, to recoup their initial development costs and to fund the ongoing costs of maintaining and supporting their systems.

However, an important challenge for a solution based on a central payment intermediary is that some key stakeholders may be unwilling to participate in such a solution. Larger publishers who have already made substantial investments into developing their own workflows and systems may be reluctant to move to a system in which they would be required to pay a transaction fee on each payment they receive. Similarly, universities and research funders may be unwilling, or in some cases unable, to turn over the administration of their open access funds to an intermediary over which they have limited oversight or control. Unfortunately, the value that a central intermediary will be able to provide can be reduced if even a handful of publishers, institutions, and funders are unwilling to participate.

Furthermore, there are important risks inherent in any solution that is based on a central payment intermediary, which tend to increase over time as the intermediary becomes more successful. Once any single intermediary has established relationships with a large number of universities, funders, and publishers, it can become difficult for alternative providers to compete. Over time, the absence of competition may result in a reduced quality of service, inflated costs, and a lack of innovation. In addition, relying on a single intermediary could further marginalize smaller publishers and institutions, particularly those with limited resources or specialized needs, since the requirements of larger publishers and institutions are likely to take priority. Finally, in the event that a well-established payment intermediary were to go bankrupt, as happened to two of the largest subscription agents (Swets and RoweCom) in recent years, it could have a disastrous impact on the open access publishing ecosystem.

The OA Switchboard

The OA Switchboard aims to leverage the benefits that a central payment intermediary can provide while avoiding the aforementioned challenges and risks that could be associated. The inspiration for this proposed solution has come from other examples of community-governed scholarly infrastructure, namely the Crossref DOI registry and ORCID, which have successfully brought together a large and diverse community of stakeholders to address complex challenges. An important distinction between the OA Switchboard and the sort of central payment intermediary described above is that the OA Switchboard is designed to enable publishers, academic institutions, and research funders to seamlessly communicate information about open access publications, without trying to serve as an intermediary for any payments that may be associated with these publications. In that sense, the OA Switchboard is simply another tool for passing metadata about scholarly publications between publishers and other stakeholders.

The following is a high-level overview of how the OA Switchboard will work, but there are many important details that need to be discussed as this initiative moves forward. The main function of the OA Switchboard will be to enable participating publishers, academic institutions and research funding organizations to send and receive a defined set of messages relating to the publication of open access research outputs.

The first type of message that a participating publisher would be able to send via the OA Switchboard would be an “Eligibility Inquiry,” which would be used to help submitting or potential authors understand whether a particular journal fulfills the requirements of their institution and/or research funder, and also whether central funding is available to pay for any open access publication charges that may be required. This Eligibility Inquiry would contain the essential metadata about the journal (e.g. journal and publisher identifiers, open access license terms, publication charge information, preservation/archiving information, etc.), as well as the standard identifiers relating to the authors (ORCID), their institutions (GRID and/or ISNI), and the funders who have been acknowledged in the publication (Crossref Funding Data). The specific list of required and optional metadata fields that should be included in an Eligibility Inquiry will be determined as part of the OA Switchboard’s development.

Once an Eligibility Inquiry has been sent to the OA Switchboard, the system would forward the message to any participating institutions that have been listed as an author affiliation, as well as to any funders that have been acknowledged as having funded the work. This routing of messages to institutions and funders would be done on the basis of institution and funder identifiers, which participating funders and institutions would select as part of their initial onboarding process. If none of the relevant institutions or funders for a given article are currently participating in the OA Switchboard, the system would send an appropriate message back to the publisher to let them know. In cases where the system is able to forward an Eligibility Inquiry to one or more funders or institutions, the system would send an initial response to the publisher to let them know that the inquiry has been submitted and a response is pending.

Funders and institutions would have a number of options for how to receive and respond to messages from the OA Switchboard. One option would be to use a simple online dashboard that would show the relevant information for each request that has been received, which they could respond to by clicking to approve or reject the request. Alternatively, if the funder or institution would like to process these requests via another system they are already using, they would be able to integrate with an API from the OA Switchboard to receive and respond to these messages. The third means by which an institution or funder could process requests would be to use a set of customizable criteria to automate the processing of these requests (e.g. requiring a CC-BY License, imposing a €2,000 cap on publication charges, requiring automated deposit into Europe PMC, etc), which would both reduce the administrative overhead on their side and enable inquiries to be replied to on a real-time basis.

Publishers would be able to submit an Eligibility Inquiry in a number of ways, depending on the capability of the systems and workflows they are using. One option that a publisher could implement on their website, which could also be made available on a cross-publisher basis on the OA Switchboard website, would be an online widget that a potential author could use to select a participating journal and enter the key information about themselves and their co-authors, their affiliations, and any funders who have supported their work. Another means of generating and submitting an Eligibility Inquiry would be via their online peer review systems, which could collect the relevant information as part of the submission process. Alternatively, for publishers who are already performing a number of manual or semi-manual checks prior to peer review, they could collect the required information to submit an Eligibility Inquiry as part of these existing pre-review workflows.

In addition to the Eligibility Inquiry messages described above, participating publishers would also be able to send a Payment Request via the OA Switchboard, which would typically take place at the point of acceptance or publication. The information contained within a Payment Request would include the basic information included within an Eligibility Inquiry, in addition to a few other key pieces of information including submission and acceptance dates as well as the publication’s DOI. In cases where the requested payment would take the form of an open access publication charge, payment details and contact information would be provided either within the Payment Request message itself or using a link to an online invoice. Alternatively, in cases where an open access membership arrangement or an offsetting/hybrid/transformative agreement is in place, details about the relevant agreement could be included in the Payment Request message. This use of Payment Request messages would enable funders and institutions to automate the oversight, management, and reporting of the agreements that they have entered into, even if no individual payment is being made as a result of a specific publication.

This overview of the OA Switchboard’s functionality will clearly require significant refinement, which is expected to take place as part of ongoing discussions over the coming months with representatives from academic institutions, research funders, publishers, and other key stakeholders. The aim will be to solve, or at least simplify, many of the challenges that currently arise in the implementation of open access policies and funding workflows. An important secondary benefit that will hopefully come as a result of the OA Switchboard will be increased adoption of standard identifiers for authors, institutions, and funders, since linking the central payment of open access publication costs to the use of these identifiers will create a strong incentive for publishers and technology providers to integrate them into their systems and workflows.

Next Steps

Over the past two months, the OA Switchboard has started to develop from an initial seed of an idea into an active community initiative with a broad spectrum of stakeholders committed to making it a success. Over the coming months, several strands of development will need to happen in parallel in order for the OA Switchboard to become operational, particularly given the ambitious goal of launching a live version of the system in early 2020.

These key next steps include:

1)     Developing a detailed product and technology plan, both for the OA Switchboard itself and for the contents of the messages that will be passed between participating publishers, institutions, and funders. A key principle for the system is that it will adhere to the definition of open scholarly infrastructure as described in this blog post.

2)     Securing the necessary funding for the initial development of the OA Switchboard, including the system’s technical development, as well as the necessary community development and engagement. This initial funding is likely to require a combination of support from publishers, institutions, research funders, and possibly other philanthropic organizations.

3)     Defining the long-term governance and funding models for the OA Switchboard. If successful, the OA Switchboard will become a core piece of open scholarly infrastructure, and therefore careful consideration must be given to ensure that it will be able to meet the diverse and changing needs of the stakeholders it aims to serve.

4)     Engaging with the wide range of constituents and communities that will be required in order to make the OA Switchboard a success. The key challenge with regard to community engagement will be to strike an appropriate balance between inclusivity and expediency. If too few stakeholders are involved in the initial planning and development of the OA Switchboard, there is a risk that the system will fail to build the support it will require from these stakeholders. However, if too many stakeholders are involved in the initial planning and development process, there is a risk that progress towards releasing a functioning version of the system will be slow. Given the rapid development of the open access landscape at the moment, the risks of failing to move quickly may outweigh the risks of failing to involve all interested stakeholders in the initial discussions about the system.

One way to address the conflicting objectives of inclusive community engagement and a focused and efficient development process is to carefully limit the size of each of the working groups that will be formed to discuss the issues above, while making the outputs of these groups as transparent as possible to the wider community. In order to achieve this, a dedicated website will soon be launched at www.OASwitchboard.org where information about the initiative will be made available as it becomes available. Once this website has been created, it will provide detailed information about each of the working groups that have been formed and allow anyone who is interested to sign up for updates about the OA Switchboard’s development.