Copyright Evidence: 21 for 2021 (a year in review)

Posted on    by Bartolomeo Meletti

Copyright Evidence: 21 for 2021 (a year in review)

This blog post was written by Bartolomeo Meletti and Amy Thomas, and first published on the Kluwer Copyright Blog on 23 February 2022.


When we launched the Copyright Evidence Portal, our ambition was no less than to create a catalogue of all existing empirical studies about copyright. This ambition, and the resulting huge body of work (over 850 studies), inspired us to think about new ways to interpret this empirical literature, and to offer a state-of-the-art overview of the evidence on how copyright works in society.

CREATe, within the AHRC Creative Industries Policy and Evidence Centre (PEC), launched the 21 for 2021 project in May 2021. We invited experts to offer a synthesis of empirical evidence catalogued on the Evidence Portal in response to 21 topical copyright questions of importance for the 21st century.

In this post, we offer an overview of the project to date, stratified across CREATe’s core research themes: Creative Industries, the Public Domain, and Competition and Markets.

Creative Industries

The digital revolution has moved legal questions about copyright, information, and competition law to the regulatory centre of the creative industries. For investors, copyright has become a currency; users struggle with rights clearance (or ignore rights altogether); creators seek ever new ways to the market. We hear wildly conflicting claims about the value of intangible assets, about the benefits of open and closed models of innovation to firms and society. Evidence is contested and in short supply.

The Evidence Portal categorises studies across different creative industries, allowing for a comparative analysis of the existing evidence. Barr makes explicit a factor we had long suspected with the skew of evidence on the Portal: the music industry is disproportionately overrepresented in empirical copyright literature. This provokes the question of how we can transplant evidence gleaned from one industry to another (in Barr’s synthesis, to television). If copyright’s application has meaningful sectorial specific differences in practice, how might this impact the formation of, often broadly constructed, copyright policy?

The industry overview feature also enables us to see immediately which industries are underrepresented in the empirical literature. In the context of the news industry, Furgał reveals that there is almost no evidence on how journalists think about copyright, or how it impacts them in their professional lives. This is so despite the frequent and repeated collation of the interests of press publishers and journalists, often used to support increasing regulation of news media and digital platforms. Such a contribution makes transparent where regulatory interventions are being supported without a solid evidentiary base.

Surprisingly, one of the areas where we have the most empirical evidence relating to copyright is in the industries where IP does not apply, or is not enforced: the so-called ‘negative spaces’ of IP, or low-IP areas. Danna, Martinelli and Nuvolari demonstrate a rare reversal of the standard call for more empirical evidence; our knowledge of negative spaces has almost exclusively been informed from detailed, ethnographic, and qualitative research in particular communities. Now, there is instead a question of how to use this evidence to inform a grounded theory of copyright.

Other contributions have focussed on creators within these industries, and look to the interplay between law and practice, examining whether specific provisions in copyright are successful in their application. Towse investigates the utilisation of performers’ rights, concluding that this remains a relatively under-researched area of copyright law, compared to similar studies on performers’ authorial counterparts. Nonetheless, the limited evidence detects a few of the same patterns in findings: performers’ rights are underutilised, economic rewards skew towards a limited number of ‘superstars’, and earnings remain overall low.

Yuvaraj and Giblin also test the effectiveness of doctrinal provisions in practice: namely, reversion rights. In theory, this reversal mechanic has a huge potential to improve access to dormant works, and simultaneously the professional lives of authors, by giving them more control over their exploitation. The evidence suggests that, whilst rarely used, reversion rights do seem to prove this theory in practice. Now, further initiatives are needed to support the tailoring of a copyright regime, in contract and statute, to encourage the use of reversion provisions.

Craig emphasises how empirical studies exploring the connections between copyright and gender remain scarce in literature. While 46 studies available on the Copyright Evidence Portal explicitly take gender into consideration as a potentially relevant variable when answering other research questions, only 5 of these are specifically designed to assess the implications of gender. More empirical work is needed to explore the ethical and social values that underlie gendered differences in women’s role as creative users, authors, and copyright owners, not only as consumers.

Public Domain

The public domain in copyright may be understood as works or other subject matter that can be used without rightsholder permission. In a wider sense, what is public and what is private has important innovation and societal effects. A better understanding of what makes a functioning cultural and scientific public sphere is critical in the digital environment. The contributions in this category have clustered around the question of optimising reuse and public access to works, whether through the introduction of, or amendments to, exceptions or limitations.

Heald offers an important synthesis of empirical evidence relevant to policy discussions on the extension of copyright duration and an optimal copyright term. Drawing on a well-established body of quantitative studies, Heald concludes that extensions to copyright duration have largely negative impacts on the availability of works, development of derivative works, and prices. Situated amidst the international trend to increasing the length of copyright protection, this review poses a challenge to such extensions on the grounds of economic efficacy, suggesting various possible directions for future research.

Wallace examines empirical evidence around the role of cultural institutions in shaping access to and reuse of both in-copyright and public domain digital heritage. Despite being only one piece of the puzzle, copyright has dominated the debate in this area. New empirical research is needed to support policy making and legal reforms that embed considerations of the public domain into rights management and new content creation, rather than focusing solely or primarily on copyright. The implementation of the CDSM Directive offers a fresh opportunity to investigate policy options in this area for non-EU jurisdictions.

Meletti offers a comprehensive overview of the vast number of studies on copyright exceptions, broadly construed, totalling 137 studies published between 1982 and 2021. Meletti’s synthesis (also available as a CREATe working paper) creates a navigable taxonomy of these studies in order to identify essential research gaps and common patterns in findings. Most importantly, the review takes a forward-looking perspective to reiterate the purpose of such studies: ‘to help courts interpret exceptions consistently; policy and law makers to draft them; and users to understand rely upon them’. Despite the vast number of studies here, there is apparently still room for gathering evidence on whether open-ended or closed-ended exceptions promote innovation, serve the public interest, or respond to market failures more effectively.

Other contributions focus on specific case studies of optimising particular uses. Margoni explores the justification for text and data mining (TDM) exceptions, largely inspired by transposition of Arts 3 and 4 of the CDSM Directive. Despite the very limited (and new) research in this area, it appears that the introduction of TDM copyright exceptions and limitations help promote the adoption of innovative, computational uses-based research, especially in jurisdictions where such exceptions are more robust and permissive.

Porangaba from the perspective of a legal scholar, offers an interdisciplinary overview of how empirical evidence can help solve the doctrinally complicated question of user-generated content. Throughout the Porangaba’s review, there is an omnipresent question of the viability of the (oft-discussed) user-generated content exception or, perhaps more provocatively, to what extent copyright is even the best, or most efficient, regulatory tool for encouraging or constraining this type of creativity.

Competition & Markets

The digitisation of the economy has drastically changed the way markets operate, and may affect the balance of powers between producers, creators and users or consumers. This makes it even more crucial to understand the avenues for innovation and creativity the law unleashes, while identifying the limits it may place on commercial activities. Contributions under this category have clustered around online platforms as an increasingly important avenue for innovation and creativity. Platform liability has been at the forefront of discussions surrounding the modernisation of copyright in recent years, particularly in relation to the introduction of Art. 17 of the CDSM Directive.

The purported change from an ex post to ex ante system of platform liability prompted some of the contributions in the 21 for 2021 project. Erickson and Kretschmer retrospectively examine the ‘before’: did the notice and takedown system, as enacted in the US Digital Millennium Copyright Act (1998) and (in a related form) in the EU E-Commerce Directive (2000), work efficiently based on empirical evidence? Drawing on an established body of quantitative studies, they show a rapid acceleration of automated content recognition systems since 2012 and evidence the dangers of false positives but also that the system largely appeared to be working.

In the wake of this automation, Peukert and Windisch look more to the ‘after’ of platform liability, in a future of ContentID, algorithmic licensing, and blockchain technologies. The evidence here suggests that a central and imminent question is how to optimise the new platform market to make automated detection efficient: at the moment, it seems we are well aware of the problems of automation, but not necessarily of the solutions. In particular, the relationship between intellectual property rights, supply behaviour and cultural participation needs further investigation.

Finally, Thomas queries whether copyright or contract is the more apt regulator of online platforms. Through an analysis of the empirical evidence on platforms’ terms of service, the synthesis makes explicit the imbalance of power between parties in the digital environment. The take-it-or-leave-it nature of the agreement, combined with sweeping and broad terms disproportionately in favour of the platform, typifies many of the problems at the heart of big tech: unrestrained scale and missing accountability.


A common thread throughout the project is that empirical evidence is crucially important for the formation of both policy and theory: but it is often sorely lacking. It has long been a core ambition of CREATe to improve this situation through initiatives such as the Copyright Evidence Portal and 21 for 2021.

Directions for future research are only likely to grow. We anticipate exciting new contributions on topical issues such as litigation data, trade, and diversity in the coming months.

The 21 for 2021 project will continue in 2022 and conclude with a closing conference, so please do subscribe to the CREATe newsletter to stay updated with the latest contributions!