Community Actions – Community-Owned Infrastructure

While individual actions are important, collective actions can have an impact at a greater scale. A third category of actions for the community to consider focus on leveraging a strength in numbers approach, and targeting “big picture” actions institutions to regain and maintain control of their data infrastructure. This category includes a broad range of possible structural solutions to foster an open, competitive landscape for data and data analytics that is aligned with the interests of academic institutions and the communities they serve. Open competition and transparency in this crucial space is essential if the academic community wants to ensure better terms and conditions from commercial vendors, more innovation from different sources, and more truly international/global solutions. By working at scale, these actions have the potential to have significant and long-term impacts, and they are intended to be pursued alongside the important campus-level actions that we recommend in the two previous sections.

Collectively Implement Strategic Practices

The most immediate step that the academic community can take is to strategically leverage its collective market power to change the behavior of commercial vendors. By working collectively in this area, it is possible to create a market where companies must not only compete on price and quality of services they provide, but also on how well they align with community values. Collective actions may happen through existing consortia or networks, state-level coordinating bodies, or new structures.

Common Contract Terms and Conditions

An important first step to consider is to have a critical mass of institutions to demand contract terms and conditions that support a more open and competitive market. In the Risk Mitigation section, we laid out multiple recommendations for institutions to adopt strong privacy policies, data policies, and engage in open procurement practices. All of these steps are important for institutions to take on their own, but the results will be even more powerful if institutions engage in these efforts together.

Broad adoption of common terms and conditions will have a market effect that favors products and services that are in the best interests of the academic community. This includes advantaging Open Source software over “black-box” algorithms and leveling the playing field for community-owned tools to compete with commercial options whenever available.

Buying Time

There are areas of infrastructure where the community can reassert or maintain control before it is lost. This is particularly true in some areas of teaching and student life, where digital tools and analytics have not yet been comprehensively deployed. However, it will take time for viable community-controlled infrastructure to come to market, and therefore an interim strategy may be to buy time. Strategies for buying time include avoiding new services with significant potential for vendor lock-in, putting a hold on new data or data analytics products until some of the actions we outline under Risk Mitigation and Strategic Choices are complete, and reconsidering steps that could accelerate vendor capture of new grey data – particularly “smart” devices and “inclusive access” digital textbook subscription programs.

Build or Acquire Academic Community-Controlled Infrastructure

The most direct path to ensure community control over data infrastructure is to build or acquire it. Currently, the vast amounts of data generated by academic institutions are largely under the control of commercial vendors. While the academic community can and should pursue strategies to ensure these vendors are more accountable, infrastructure that is truly owned and governed by the academic community is best positioned to align with its values and needs. As Bilder, Lin and Neylon pointed out in 2015,¹“everything we have gained by opening content and data will be under threat if we allow the enclosure of scholarly infrastructures…”

How can the community build or acquire infrastructure?

There are three main approaches, which may be pursued in combination or separately to build or acquire infrastructure. Each of these courses of action requires weighing their trade-offs, and each also requires academic institutions and funding bodies stepping up to invest the resources necessary to assemble viable alternatives to commercial solutions.

Build from Scratch. Building assets from scratch would ensure that their design aligns with the requirements of academic institutions and that their governance reflects the mission, values, and goals of the community. It would also likely be a cost effective approach. On the other hand, this approach requires academic institutions to build up very specific and potentially new capacities, and it may take longer to launch products that can compete in the marketplace with commercial vendors.
Funding Start-Ups. Funding start-ups or existing community-owned initiatives would have some distinct advantages over building from scratch, since they come with their own management teams and attendant competencies. In the case of start-ups, a mix of academic and private ownership is possible, allowing the community to seek additional funding across both academic institutions and venture capital funds. On the other hand, venture capital management can be risky and complex. Most important, the goals of academic institutions (launching and running a workable infrastructure) and those of management and funders (maximize financial returns) may come into conflict with each other over the long term.
Acquiring Existing Assets. Acquiring existing assets may be the most expensive option up-front, since sellers may want a premium that reflects their initial risk. On the other hand, existing assets could well provide the fastest way to come to market with the appropriate products and services, as well as bringing in an established user base. However, there may be some issues around interoperability.

The choice between these options is largely a function of the level of funding provided and the availability of potential assets to acquire. As an example, the cost to outright acquire the full set of infrastructure for the scholarly communications process would likely be in the low hundred millions of dollars. To build this infrastructure from the ground up would likely cost less up-front, but would take substantially longer. However, it is worth noting that there are Open Software-based solutions across the entire workflow of scholarly communications that could help the “build” process go faster. The recent “Mind the Gap” work by John Maxwell and team contains a detailed map² that provides an excellent blueprint of this particular space for the community to consider.

It is important to point out that the financial commitment necessary to build or acquire assets is not a one-time investment.

Any movement towards true community control of infrastructure will require institutions to be willing to invest in digital infrastructure with the same commitment as they currently invest in physical infrastructure.

Like buildings and roads, this will require regular additional investments to ensure that their underlying technology remains up-to-date, stable and sustainable. Emerging initiatives like the Invest in Open Infrastructure (IOI)³ and the Sustainability Coalition for Open Science Services (SCOSS)⁴ could play important roles in ensuring that there professionally managed investment options for the community to consider.

How would the community pay for it?

The first step before any investment would be to conduct a detailed analysis of business plans and potential acquisition valuations. This would allow the community to conclude which options are most attractive in financial terms. Performing this assessment is timely, as many institutions are at a crossroads in reconsidering their current financial relationships with commercial content providers, and consequently, are strategically rethinking their scholarly communications, course material, and infrastructure spending.

More and more frequently, these explorations are taking place within the context of the institutions larger strategic infrastructure investment discussions. For example, one major research institution’s President recently tasked the library to analyze how it could cut its spending on scholarly communications by 50%, with the goal of reinvesting its spending into initiatives aimed at radically changing how scholarly communications and infrastructure are managed. If such an approach became a socialized and coordinated effort among leading research institutions in North America or - ideally - globally, it could yield resources approaching the level needed to fund a collective approach to community-owned infrastructure. Likewise, a number of institutions have announced six-figure funding for local open educational resources programs, viewing it as an investment that will pay dividends in reducing the overall cost of education for students. As more institutions begin to approach open educational resources as infrastructure for teaching and learning, there is an opportunity to coordinate shared investments toward meeting common needs.

Redirecting current institutional content spending towards infrastructure is an important potential source of funding, but is not the only one. Multiple sources will be required in order to ensure scale and sustainability. Establishing new partnerships between private and public funding bodies and higher education institutions to support community-controlled infrastructure is also critical. There are already examples of funding bodies independently taking an active role in supporting new dissemination platforms (e.g., eLife⁵and Wellcome Open Research⁶). However, it remains to be seen whether joint initiatives between academic institutions and funders can be established and gain traction.

Establish Inclusive Governance Structures

Another approach is to identify or construct governance structures that would allow new kinds of relationships between academic institutions and vendors – commercial or otherwise. While members of the academic community often participate in the “governance” of commercial vendors in an advisory capacity, they rarely have the opportunity to do so from a position where they can exert real operational influence – such as a position on a fiduciary board. If a critical mass of academic institutions were to demand such a role, interesting new opportunities for community-aligned governance could be explored.

It is vital for the governing bodies of infrastructure services to include representation from the communities they serve in order to ensure that management stays accountable to the community’s evolving needs. Iterations of this approach have long been a part of the governance of Open Software initiatives, some of which might serve as useful foundations for governance models in other types of infrastructure. Governance bodies should be deliberate about considering which voices are important to include, and strive for diverse representation across a wide cross-section of factors, including institution type, geographic location, career stage, gender identity, racial or ethnic identity, disciplinary background, as well as other relevant factors.

Leverage Policy to Support Community Control

Updating Federal and State Policy

Another avenue to expand community’s control over data infrastructure is to advocate for favorable federal and state policies. Multiple studies have found that the current legal framework is insufficient to prevent the commercial exploitation of student data.⁷ In the U.S., the most immediate opportunity centers on data privacy protections, and the primary federal law that governs student data privacy in higher education, the Family Educational Rights and Privacy Act (FERPA). This law was established before the internet, and is limited in both scope of coverage and protection it offers students. Updating FERPA is a clear opportunity to advocate for stronger provisions that guarantee students full control over their education data, protect it from exploitation, and set stronger security standards for vendors.

Another promising option derives from federal-level discussions around potential consumer data privacy legislation, akin to Canada’s Personal Information Protection and Electronic Documents Act (PIPEDA) and Europe’s General Data Protection Regulation (GDPR). Ensuring that any proposed legislation also applies in the education environment (and carries any additional protections the community deems necessary), would also provide leverage to institutions for maintaining control over critical data.

In both the U.S. and Canada, higher education is closely tied to the state or provincial governments, so legislation at this level could also offer advantages. For example, state law typically governs procurement at public institutions, and open procurement practices could be advanced through a ban on NDAs or other unwanted provisions in relevant contracts. State and provincial governments might also consider introducing regulations that guarantee that students are informed about the terms of use for digital course materials before registering for a course, or regulating “inclusive access” programs to prevent course materials from becoming an unaccountable student fee.

Antitrust Actions

There are also possible actions that can be considered under existing antitrust law. Both the scholarly communications and courseware publishing markets have become increasingly concentrated in recent years, despite concerns raised within the academic community. The rise of data analytics adds another layer of concern: if the economic model of digital publishing favors oligopolies, that of data analytics favors the rise of monopolies. As our experience of social networks and search engines demonstrates, the future of research and education could end up being defined by singular firms with excessive market power.

The recent decision of the Department of Justice to initiate an investigation into the practices of leading tech companies suggest that the view of regulators over these issues may evolve, particularly in regard to the collection and usage of data. While it is early days, the academic community should monitor how regulators (and the courts) decide on concentration of data and data analytics services and their ties to the provision of other services and – if the situation demands it – initiate or support antitrust actions at the appropriate point in time. The community can also work collectively to take proactive antitrust action by filing comments with antitrust enforcement agencies or working actively to oppose mergers.

Realign Stakeholder Relationships

These community-based actions portend several possible realignments within the academic community and its stakeholder groups that should also be considered as efforts move forward.

Academic and Research Libraries

Within academic institutions, there is need for realignment between libraries and the rest of the institution. Library professionals live and breathe data and information flows every day, and have a unique opportunity to contribute their expertise. In addition, there is a clear need for the senior administration to identify the leaders who will organize these actions, and librarians could well lead some of them. In order to do so, however, libraries will have to upgrade their project management competencies, and become comfortable mobilizing resources from outside their traditional core activities (for example, from legal, ethics, economics and business experts).

Funding Bodies

The second realignment is within the broader research community. Historically, funding bodies and academic researchers have worked at arm’s length, and some of that separation will have to continue (particularly in the grant approval and review processes). However, at a broader level, the issues posed by data analytics portend a much closer and aligned relationship between funding bodies and academic institutions, as they both share some of the objective of keeping research data infrastructure open to competition.

Scholarly Societies

As new partnerships and financial arrangements that ensure greater alignment with community values are considered, the role of scholarly societies might also be reexamined. Historically, the relationship between academic institutions and scholarly societies has been complex and sometimes disconnected, as many academics consider their society to be their primary affiliation before their institution. Additionally, scholarly societies have been perceived, for right or wrong, to be among the least enthusiastic supporters of open scholarly practices, given their concerns over the potential revenue loss in a transition to open access.

In the development of community-owned infrastructure, the relationship between academic institutions and societies might be productively reexamined and realigned to support mutual interests. The development and management of community owned infrastructure requires many functions, including some that are largely dependent on disciplinary expertise that societies alone possess. New kinds of direct fee-for- service arrangements may offer an alternative source of revenue to societies, while supporting direct community control of the communication of research outputs. Similarly, as more institutions invest in open educational resources, scholarly societies are poised to play a potential role as a service provider for vetting or publishing educational materials.

“For example, a number of learned societies have established the Society Publishers Coalition.⁸ This is, effectively, a “coalition of the willing” with the aim to establish closer working relationships with academic institutions and funding bodies and may help taking the initial steps to establish more collaborative relationships between the academic community and societies.”