Many factors raise barriers to open access. Some are specific to a particular GLAM, sector, collection, or jurisdiction. Barriers most commonly referenced among GLAMs and wider scholarship include:
A lack of copyright education and knowledge or access to expertise. For many GLAMs, the inability to determine whether works are in the (global) public domain versus in-copyright can override desires to prepare collections for open access release. This is a crucial area where the OpenGLAM initiative might be able to provide support and help build capacity among the GLAM ecosystem.
The expense of digitization. Digitization is expensive and can hinge on the costs of technology, labor, expertise, internal and external digital infrastructure, and its storage, preservation, and management. These costs are ongoing and require significant investment to keep up with advancements in new technologies and forms of engagement. With decreased funding, GLAMs are hesitant to give up any revenue (however small) that can support digitization programs, including licensing.
Private partnerships and their contracts. Because of these costs, GLAMs sometimes form exclusive partnerships with companies that provide digitization and commercialization services. While GLAMs typically receive copies as part of the agreement, the partnership contracts and third-party IPR claims typically prohibit GLAMs from releasing these copies under open access frameworks.
Desires to license “commercial use.” Some GLAMs view certain noncommercial uses as falling within educational missions, but license commercial use. How copyright law distinguishes between commercial and noncommercial use is fraught with jurisdictional and other highly-contextual considerations, which may not align with GLAMs’ views.
Desires to retain control. Fears around loss of control of a work and its educational context are often cited as why licensing models remain necessary. In these cases, IPR claims are also used to protect the work, the artist, its context, and the host institution.
Fears around the public domain and freeriding. Fears that releasing digital surrogates to the public domain will enable commercial photo libraries to freeride on GLAMs efforts are legitimate and already happening. This practice is legal due to the material’s public domain status.
Free entry onsite as justification for licensing fees. GLAMs without open access programs sometimes reference free onsite entry as a trade-off and justification for charging licensing fees. These GLAMs equivocate this practice with open GLAMs that charge high entry fees as evidence for why the open access program is possible. However, data suggests the majority of the museums and galleries that release collections under open access frameworks charge nominal fees or provide free entry onsite (the data excludes libraries and archives as they typically provide free entry).
Licensing as a (bad) business model. Citing licensing as a business model that drains resources, some GLAMs are exploring the new forms of commercialization, merchandising, and revenue opportunities made possible by open access and the public domain.
Practical barriers (beyond the scope of this resource). Other practical barriers will arise obstructing open access implementation. Such a major policy change requires staff buy-in across various departments (including those who worry open access may eliminate their jobs). More importantly, it requires a commitment by administrative staff at the highest levels.1 For many institutions, the lack of digital expertise or technology renders open GLAM participation impossible. Even GLAMs with digital support note difficulties keeping up with an area that evolves so rapidly. Open access can even exacerbate an institution’s technology gaps or weak spots, as it may lead to additional traffic and increased demand on technical infrastructure.2 For the most part, these practical barriers fall outside the scope of this paper. Resources are available that document individual experiences, how to approach or move past these barriers, and how to structure for digital success.3
Below, we’ll examine each the obstacles and evidence that suggest copyright claims in reproduction media pose more problems than solutions.
Let’s first address the financial elephant in the room. Digitization is costly. It requires long term planning, active maintenance, time and expertise, and sustained investment – all of which were scarce resources prior to COVID-19.4 Our collective governments have seriously underfunded the cultural sector for years. It is no wonder copyright’s economic benefits are sought to support digital operations. While not unwarranted, evidence shows this ambition is misplaced, even for the purposes of cost recovery.
As early as 2004, research by Simon Tanner revealed reproduction charging models were not profitable among a group of large American art museums.5 In his study, Tanner found little, inconclusive, or inconsistent data being monitored for factors like costs of production, administrative workflows, salary and overhead expenses, monitoring and enforcement, technological support, and delivery.6 Many museums reported funneling any income back into the service, which operated at a greater cost to the museum than it brought in.7 A few exceptions were able to profit off of a relatively tiny group of seminal or popular works.8 Even today, GLAMs cannot say with accuracy whether licensing models are profitable once costs are accounted for. One reason for this is because the granularity of financial modelling required to do so is not prioritized by non-profit organizations, nor should it be.9 Although Tanner’s 2004 study focused exclusively on museums, subsequent studies across the GLAM sector have reinforced his findings.10
Yet someone must pay for digitization and the costs it entails. Without a doubt, governments fail to provide GLAMs with crucial, adequate, and sustainable funding to enable them to fulfill their ever-expanding remits. Some public funding schemes obligate GLAMs to generate revenue via follow-on activities.11 Licensing thus provides an easy means to discharge this duty, pass costs on to consumers (i.e., their public), and feed revenue back into digitization programs.
Plus, for some GLAMs, licensing may generate not insignificant amounts. Any sum of revenue is difficult to abandon in the face of tightening belts. Indeed, austerity measures and decreased government funding over the years have caused GLAMs to over rely on copyright’s economic benefits and design strategies that support business-driven goals, sometimes out of survival. With budgets now at further risk due to COVID-19, making the case to walk away from licensing revenue and release digital materials to the public domain is only getting harder.
But what is the greater cost to the public and the economy when public domain collections remain locked up via (dubious) copyright claims? And what new forms of direct and indirect revenue might flow from open access business models that could surpass the amounts received via licensing revenue? There’s really only one way to find out.
While policy developments in the past decade have led to measurable progress, GLAMs continue to encounter significant financial barriers during the transition to open.12 This is unlikely to change anytime soon. But overcoming the first hurdle to open GLAM participation requires decision makers to change their mindsets around the reuse of collections, moving the digitization program from promoting business- to mission-driven activities.13
One (very) attractive option to reducing the financial burden of digitization is to bring someone else in to do it entirely.
Public-private partnerships are sought because they bring the technology and know-how to digitize collections and are often accompanied by forms of income that help augment operational costs. These partnerships can introduce third-party commercial interests into IPR and digital collections management. Questions will arise around contractual obligations to partners who claim IPR in the digital outputs, whether or not the IPR is valid.14 The cumulative effect is to entrench approaches that prioritize commercialization and risk aversion above open GLAM goals.
Contracts and IPR are both limited in scope and time, but those limits can conflict in practice. For example, contractual terms might create an exclusive partnership between a GLAM and third party for a limited period, after which the GLAM can form new agreements around digitization and collections management. But unless the limited period extends to the rights in the resulting media, GLAMs are effectively bound and limited with what they can do with the media for the term of the copyright. This means many GLAMs cannot release their copies of the digital collections via open frameworks even if they wanted to. Instead, new digitizations must be made for any purpose not covered by the copyright or contract.
Here, faults must be fairly attributed to the wider system that is defaulting on its obligations to adequately support cultural heritage institutions and their publics. Decisions to commercialize collections via private partnerships are never self-interested decisions made in a vacuum. Moreover, some of these partnerships can be incredibly exciting and open up new opportunities for GLAMs and the public. Digitization might only be the first step toward a more innovative initiative that produces new data, research, software, and technologies for something like artificial intelligence or machine learning. During negotiations for such initiatives, GLAMs should advocate for more nuanced approaches to how the non-original and more basic reproduction outputs can be released under open access frameworks.
Lastly, GLAMs are not without support from their public. New engagement via crowdsourcing and volunteer communities might be expanded to test crowdsourcing the digitization labor, data management, and expertise necessary to prepare collections for open access frameworks and platforms.15 This raises questions around capacity, privilege, and compensation for labor discussed further in Power Inequities.
Disparate and bespoke website terms on educational use among GLAMs add to the confusion: GLAMs might define “educational use” according to details around the image dimensions or resolution, by placing a cap on the number of works that can be used for free, or via specific contexts, such as “scholarly” use or use in an academic blog.17
Research on publication fees shows further disparities appear in pricing models that make blanket assumptions about the profitability and costs of publication models.18 By contrast, commercial publishers might place the burden of those fees on the author or provide limited budgets for reproduction fees, which authors must then supplement with personal funds.19 Charges for a single image can often outprice the cost of the actual publication.20 Others may seem reasonable in isolation, but the total costs quickly become unaffordable when multiple images and licenses are required.21 Faced with mounting fees and restrictions, some researchers reported reducing the amount of content quoted or included, or abandoning projects altogether.22
Similar to contracts, licenses are limited in scope and time. This means if a researcher licenses an image to study a work, that image cannot (legally) be used for any other purpose. Its inclusion in a lecture, blog post, or journal article will require a new license and fee payment to the rights holder. Licenses also expire. After a five-year license, a researcher will need to renew the license to (legally) continue using the image, subject to a new license and fee payment. And while researchers may pay to license a high-resolution image for publication (and study it, but don’t tell), most journals and periodicals print in black and white or at lower-resolution formats, undermining some value (and the price paid) for the image’s inclusion.23 One study noted that art history scholars expect to take extra steps to find a better quality image when the image paired with the article is sub-par or entirely absent.24 In other words, the public now turns to the internet, most likely Google Images.25
The difficulties distinguishing between commercial and scholarly use have motivated some GLAMs to reduce reuse barriers by adopting more open policies.26 Others have turned to licensing frameworks like Creative Commons. The Creative Commons NonCommercial (NC) license introduces some clarity due to the official interpretation of “educational use,” but the application of a NC license raises four concerns.27 First, Creative Commons does not support applying NC (and other) licenses to non-original reproduction media of public domain works. For this, CC0 is recommended.28 Second, the application of a noncommercial restriction to any material does not satisfy definitions of “open” for open access purposes. Both of these points are detailed further in Clarifying “Open”. The third issue is that “noncommercial” refers to the uses, not the users. This means NC licenses do not necessarily deter the types of actors that some GLAMs might think they deter. Finally, a number of GLAMs that apply a CC NC license to reproduction media also include a statement in the copyright policy of the specific activities considered to be commercial versus noncommercial. Many activities conflict with the official Creative Commons interpretation. In the event of a dispute, which terms should prevail?
Ultimately, financial barriers to open access for GLAMs translate to downstream reuse barriers for the public, even for GLAMs that seek to enable noncommercial and educational use.
Fears around loss of control erect philosophical barriers in terms of who should be able to interpret or generate knowledge around a collection, and for what purposes. In Tanner’s 2004 study, museums reported that between fee collection and control, the more important factor for licensing was retaining control to ensure credit was maintained to the host museum, artists, and their works.29 Studies on archives, special collections, libraries, and the value of metadata report similar control-seeking outcomes.30 These fears may be reflected in fee models that price images out of reach for most audiences and safeguard an image’s scarcity.31
Licensing also ensures a wider system of “permissions” may be employed: even when fees can be paid, the desire to control how the image is used, interpreted, and discussed may lead the (alleged) rights holder to deny a user’s request.32 Refusals might be based on factors from the user’s intention to alter or crop the image, to perceived conflicts with an institution’s subjective values and ideas of what constitutes inappropriate use or even disrespect of the artist, artwork, or institution.33 A copyright claim provides a convenient protectionist measure with added economic benefits.
This paternalistic practice has no basis in law. It also communicates an inherent distrust in the public. As Maria Vlachou has argued, “objects in the public domain belong to everyone and no one has appointed museum professionals as arbiters of taste.”34 Open access thus requires GLAMs to be willing to educate users around best practices and to tolerate the risks associated with misuse.35
To date, there has been little evidence to suggest these fears will materialize.36 Even anecdotes of racist misuse cannot justify a practice that increasingly conflicts with copyright law.37 (Although other steps can and should be taken to reduce such risk.) For a discussion of pros and cons related to the good actor / bad actor licensing question, Michael Weinberg made a compelling argument in favor of less intimidation and risk aversion in licensing and reuse policies.38
Another fear relates to the possibility that once reproduction media is marked public domain, others will download and license it for profit. Unfortunately, this is unavoidable and inevitable.
Some GLAMs have already encountered instances of free and open access images offered for sale on commercial licensing websites. The images are identical in size and resolution to those uploaded to Wikimedia Commons or made available through the GLAM’s own website. However, they are obscured by watermarks and lack details about the image’s public domain status and host institution. Alamy and its contributors have been repeat offenders of this.39 While such freeriding may seem unfair and against the spirit of open access, it is entirely legal – and it has legal precedent to back it up.
In 2016, Getty Images was sued for doing this exact thing by photographer Carol Highsmith, who dedicated thousands of her original photographs to the public domain upon depositing them with the Library of Congress.40 Getty Images even sent Highsmith a cease-and-desist letter demanding she pay the company a licensing fee for using one of her own public domain images on her website. The $1 billion lawsuit for gross misuse and false attribution was dismissed on the grounds that the images were in the public domain.41 Accordingly, any reuse was permitted, including the commercial licensing of public domain works.42 It’s worth noting the parties settled out of court on New York state law claims related to deceptive business practices.43
Both Merete Sanderhoff of the National Gallery of Denmark (SMK) and Karin Glasemann of the Nationalmuseum Sweden have asked Alamy to remove their public domain images from the commercial licensing platform, with limited success. Any images removed were of gallery spaces and taken down for privacy reasons. Regarding the public domain artworks, Sanderhoff suggests playing offense: “all we can do is advertise that our images are free and keep pushing the OpenGLAM agenda to become more widely known, so the public comes to understand they don’t have to pay for the images.”
If open GLAM becomes the norm, this could collectively knock the wind out some of these practices. At the very least, a diligent user might know to perform a reverse image search to locate the open access version before paying a licensing fee.
One final point to make is specific to galleries and museums, since it is uncommon for libraries and archives to charge entry fees.
Some decision makers have commented that the galleries and museums with open collections are able to recover lost licensing income through high ticket fees at the door. An example often relied on is the Rijksmuseum in Amsterdam, which charges €20 for entry. In 2015, the Rijksmuseum released 150,000 high-quality digital surrogates and collections data to the public domain (that number now exceeds 600,000). Afterward, the museum saw a marked increase in visitors. The Rijksmuseum example is often used and assumed to be representative of other galleries and museums with the availability to use ticket fees to offset revenue loss and support open access digitization programs.
The data shows otherwise. Of the (at least) 133 galleries and museums who release all eligible data open access, the majority (79) charge the equivalent of €5 or less at the door for entry.44 In fact, the 65 galleries and museums listed below provide both free access onsite and open access for all collections online. Many release rich, high-quality collections data.
Australia: Museums Victoria
France: Maison de Balzac, Maison de Victor Hugo, Musée Carnavalet, Musee Cognacq-Jay, Musée d'art et d'histoire de Saint-Brieuc, Musée d'art moderne de Paris, Musée de Bretagne, Musée de la Vie romantique, Petit Palais
Hungary: Semmelweis Orvostörténeti Múzeum
Portugal: Museu da Cidade de Aveiro
Slovakia: Slovenská národná galéria
Sweden: ArkDes, Armémuseum, Dalarna museum, Flygvapenmuseum, Hallwylska museet, Hälsinglands Museum, Karlsborgs Fästningsmuseum, Litografiska Museet, Livrustkammaren, Marinmuseum, Mölndals stadsmuseum, Nationalmuseum, Skoklosters slott, Sörmlands Museum, Statens maritima museer, Vänersborgs museum
United States: Albright-Knox, Arthur M. Sackler Gallery, Birmingham Museum of Art, Bowdoin College Museum of Art, Cleveland Museum of Art, Davison Art Center at Wesleyan University, Detroit Institute of Arts, Freer Gallery of Art, Hirshhorn Museum and Sculpture Garden, J. Paul Getty Trust, Minneapolis Institute of Art, National Air and Space Museum, National Gallery of Art, National Museum of African American History and Culture, National Museum of African Art, National Museum of Natural History, National Museum of the American Indian, National Portrait Gallery, National Postal Museum, Saint Louis Art Museum, Smithsonian American Art Museum, Walters Art Museum, Yale Center for British Art, Yale University Art Gallery
Of the remaining 54 galleries and museums:
21 charge fees between the equivalent of €5 and €9.99;
22 charge fees between the equivalent of €10 and €14.99;
7 charge fees between the equivalent of €15 and €19.99; and
5 charge fees of €20 or more.
Of course, this data cannot alleviate pressures to self-finance via ticketing or licensing fees (made worse by COVID-19 and the loss of foot traffic to support other commercial aspects, like the gift shop or café). It is also worth noting that in the year following the Metropolitan Museum of Art’s February 2017 announcement on the adoption of CC0 for all reproduction media, the museum announced it would resume the $25 admission fees for non-New York State residents.45 But the data does suggest adopting open GLAM without relying on ticket sales to replace licensing income is possible. Ultimately, governments must step in and adequately fund the cultural sector to facilitate digitization and open GLAM.
Ultimately, these barriers obstruct the potential of new innovation around public domain collections, as well as innovation around open access as a new business model. Instead of sustaining licensing systems that require overhead and drain resources, we might explore what new and exciting forms of commercialization, merchandising, and revenue opportunities flow from open access. Some examples of this are discussed in Benefits to Open Access, but more research and data on open access business models is necessary to inform more sustainable policy change.
In the meantime, mediated approaches to revenue generation can still be explored. Some examples include:
Transparent fees that pass costs to the user. Libraries and archives typically charge service fees when providing data, which are more transparent and affordable to users than copyright fee structures. (Digitization costs will obviously vary depending on the work, format, any data management and post production, and type of request.) For requests that require new digitizations, GLAMs can (and already do) pass costs on to the consumer. Fee models can build in and list hidden costs like the maintenance, storage, and delivery of digital heritage. GLAMs already operating such systems might publish their pricing schedules and modelling to aid the wider sector. The more transparent and reasonable these fees are, the more justified the ask of the service fee will seem. Why continue modelling costs on copyright licensing, which are opaque and more difficult to defend?
Donation or pay-what-you-can systems. To illustrate, one unnamed GLAM website previously allowed users to make contributions immediately prior to image download. Donation amount options were aligned with the costs of conservation, purchasing equipment, or running an educational program (e.g., €5 buys one paintbrush for conservators; €50 buys art supplies for a drawing class of 25 students). Donors could also include personal messages, which then were advertised alongside their donation with notes of gratitude by the GLAM. This system could easily be adapted to support and promote the costs of digitization and digital collections management. These costs remain so transparent users are often unaware they exist. Donation amount options could align with costs of digitization programs (€29 enables us to digitize one new painting; €100 pays for 100G of cloud storage services for one year). Moreover, there are many emerging examples of pay-what-you-can business models in other sectors. Some GLAMs even have pay-what-you-can entry fees or suggested donations. Why not also open GLAM?
More nuanced data releases. Data releases don’t need to be all-or-nothing, or tied up in big announcements around a massive website design and digital program overhaul. GLAMs can begin by releasing smaller datasets and basic digital outputs as open access. More than 600 GLAMs have taken this approach so far, rather than adopting an institutional wide policy. This approach can even be with more sophisticated projects (with private partners) that produce a range of IPR. The non-original or basic media appropriate for open access can be released while more sophisticated IPR can be retained for commercialization. The sector might collectively explore and share how to make better business decisions around digital media management that do not cling to low-hanging IPR and rely on scarcity in a (shrinking) market.
More equitable contracts with private partners. This one is tricky and will depend on a GLAM’s bargaining power and familiarity with contract negotiations. How can we encourage and establish more equitable contracts with various partners? This is not to say such partnerships are by default imbalanced. But there are many boilerplate contractual provisions or standard terms that could result in a GLAM assigning or limiting rights that, for example, affect access, reuse, or royalties generated by patents or other sophisticated IPR generated during a partnership. Sharing examples and tips and creative solutions could positively impact future rights management and revenue generation (including how to get around confidentiality clauses, if included – not everything is commercially sensitive information).
Merchandising! This one is obvious, and already happening.
And, importantly, any combination of these can be used.
Note to the reader: what else should be included in this list?
It may seem difficult to design more sustainable systems that are not based on copyright, but the evidence consistently suggests that literally anything might be more sustainable than clinging to the licensing status quo.
Data on image licensing after open access implementation reveals varying impacts. Some early adopters of middle-of-the-road policies reported a marked decrease in commercial licensing that was already on a downward trajectory;46 some were not concerned with licensing revenue loss to begin with;47 others noted little to no decrease in licensing revenue following open access adoption.48 There are even a few examples of temporary increases in image sales.49 One reason is because commercial publishers and licensors typically prefer to pay (and have budgets) for that precious piece of paper ensuring use is legal.
Finally, relevance is a significant risk with closed collections. As more GLAMs go open, online traffic and consumption will increasingly follow the openly-licensed materials. Accordingly, the risks shift from those associated with adopting open access to the downsides of maintaining restrictive policies.50 Copyright income is already limited and will continue to decrease as reuse patterns redirect to open collections. Other downsides could include decreased opportunities around brand licensing revenue, limited public engagement with collections on popular websites, or reduced general awareness of collections.51 Maintaining relevance in an attention economy is important, but it is particularly so when leveraging the potential of open collections to connect with users globally through new technologies to enable new knowledge and markets to develop around reproduction media.52 Through open access, many GLAMs have seen gains from new business opportunities and industry relationships, new methods of engagement, knowledge generation, and collections management, and greater fulfillment of public missions, in general.
There is a certain irony when GLAMs fail to level the public domain playing field by locking their public(s) out of its reuse. But governments must also support the cultural sector with adequate financial, practical, and legal infrastructure to digitize collections and actually implement open access programs. As business models around cultural content diversify and digitization incorporates new technologies, we risk allowing public domain collections and their relevance to become exclusive markets for commodification by only those trusted with their care or permitted to access them. Instead, GLAM decision makers and their governments must heed legal scholars’ warnings against the privatization of cultural heritage and the distortion of future knowledge.53 We’ll discuss how this unfolds in Power Inequities.
Continue to Benefits to Open Access