Introduction
As we’ve seen, reversion’s potential to redress creator-rightsholder imbalances has been severely undermined in the Commonwealth and US by a combination of rightsholder lobbying, poor design and ambiguous drafting. In this chapter we now ask – have European nations done a better job?
Reversion rights have been a longstanding feature of continental copyright laws, where they’ve been implemented with a much richer and more diverse array of triggers than the simple time-based mechanisms favoured by common law nations. This notably includes a wide variety of ‘use-it-or-lose-it’ rights, which premise retention by rightsholders on works’ continued availability. By freeing up unused rights for fresh investment, such mechanisms have obvious promise as a way of furthering both copyright’s access and rewards goals.
The range and scale of European reversion rights over time combines with space and language constraints to prevent us attempting any comprehensive history. Instead, we more modestly seek to determine their relative efficacy by examining the laws of key European nations immediately before and after implementation of the 2019 EU Digital Single Market Directive (‘DSM Directive’), together with the evidence we’ve been able to find out about their effects. Curiously, despite the broader range and availability of reversion rights on the continent, we find little evidence that they’ve had more practical benefit (for authors or the public) than the reversion rights of common law nations – though once again, we argue that this is attributable to deficiencies in their design and implementation rather than the concept itself.
4.1 Early Origins of Reversion on the Continent
Whilst a comprehensive history isn’t feasible, we can share two key eighteenth-century developments that help explain the early origins of the flourishing that was to follow, whilst also confirming that European legislatures have not been immune from the kind of undermining by rightsholders that has so impacted reversion rights in Britain and the United States.
4.1.1 The 1741 Danish Copyright Ordinance and Copyright’s Access Goal
Use-it-or-lose-it rights in continental Europe can be traced to its first copyright statute – the 1741 Danish Copyright Ordinance (the ‘ordinance’).Footnote 1
The ordinance prohibited anyone from publishing, reprinting, importing reprints or selling any book or writing that had been lawfully acquired by someone else.Footnote 2 However, those exclusive rights came with a twist: a corresponding responsibility to keep that material available to the public. The law gave publishers between three months and a year to reprint titles that went out of print. If they failed to do so, their right would lapse, and anyone would be free to print the work.Footnote 3
This law demonstrated a clear intent by lawmakers to promote not just the initial creation of works, but ongoing investments in their continued availability – in other words, a strong and explicit support of copyright’s access aims.
This mechanism put unused rights into the public domain, rather than returning them to authors, but it’s a clear ancestor of the use-based reversion rights that would begin to emerge in Europe soon after. The framers of the ordinance viewed failure to continue exploiting a work as abandonment of rights: ‘other persons must be allowed to publish the same writing, since it is to be assumed that the rightful owner has relinquished his right to it because of the long delay.’Footnote 4 If the rightsholder was not going to use the rights, that should not necessarily mean that society loses access. Others should have an opportunity to invest in them instead.
In Chapter 2 we invited readers to consider how the 1737 English bill proposing limiting transfers to 10 years would have radically changed the power balance between authors and publishers.Footnote 5 It’s equally worth considering how different access would look today had the copyright laws that followed the first European copyright statute followed suit in premising exclusivity privileges on continued availability – that is, if they had continued to tie ongoing rights to ongoing responsibilities.
4.1.2 The 1794 Prussian Law and Copyright’s Reward Goal
Our second historical vignette concerns a reversion mechanism that was included in the penultimate draft of what eventually became the 1794 Prussian law (Allgemeines Landrecht für die Preußischen Staaten, or ‘General National Law for the Prussian States’Footnote 6). That draft incorporated rigorous protections for authors, most relevantly by limiting the publisher’s rights to publication of a book’s first edition. If they wished to publish a second, they would need to obtain the consent of the author or their heirs to do so, and authors would themselves be entitled to publish a second edition after all copies of the first were sold.Footnote 7
Effectively then, the idea was to return the right to publish each subsequent edition to authors or their heirs, which they could then exploit at a time when all parties had a better idea of the work’s commercial value.Footnote 8 This was particularly important at that time because, as we discussed in Chapter 2, authors were often paid by way of lump sum rather than having any ongoing right to royalties.Footnote 9
At the 11th hour, however, intense opposition from prominent Berlin publisher Friedrich Nicolai led to sweeping changes that eliminated many of the draft’s author-protective features.Footnote 10 Under the version that was enacted into law, publishers’ rights were still limited to the work’s first edition (though the definition was broadened to include ‘any subsequent volumes and sequels’), and, more significantly, that limitation could be eliminated by contract.Footnote 11 That radically altered the proposed bargain by permitting publishers to negotiate up front for exclusive rights over all subsequent editions, and locking authors and their heirs into terms that had been agreed before anybody had a clear idea of the work’s value. This shows once again how concerned publishers are to prevent meaningful reversion rights from being enacted, and demonstrates that Europe was not immune from the same kind of rightsholder undermining that so influenced author-protective laws in Britain and the US.
4.2 Time and Exploitation-Based Triggers in Europe
Two main categories of reversion right emerged from these origins: those with time-based triggers (analogous to those we’ve already seen in common law nations) and ‘use-it-or-lose-it’ formulations (which, as introduced previously, allow authors to reclaim rights previously granted where rightsholders don’t exploit them, or exploit them inadequately). Below, we briefly explore key varieties of each type as they existed in EU Member States in the lead-up to the DSM Directive. In doing so, we draw substantially on the work of Yuvaraj and Furgal carried out as part of the Author’s Interest Project in 2019 and 2020, which we commend to readers interested in diving deeper.Footnote 12
4.2.1 Time-Based Triggers
Time-based mechanisms might return rights to authors after a certain period, or limit the amount of time an author can license rights. Either way, the effect is the same: rights boomerang back to the progenitors of the work, creating possibilities for fresh investments in making them available to the public and for creators to share in their value.Footnote 13
4.2.1.1 Spanish Origins
The earliest time-based European mechanism we have identified was part of Spain’s 1879 Act, and returned rights to authors’ estates 25 years after death.Footnote 14 At that time Spanish copyrights lasted for 80 years after the author’s death, much longer than surrounding nations,Footnote 15 and lawmakers were motivated by a desire to ‘reconcile the legitimate interests of the author during his life and the consideration owed to his family’.Footnote 16
The Spanish scheme was criticised for interfering with the law of succession and making negotiating copyright treaties with other countries more difficult,Footnote 17 but was nonetheless exported to parts of Africa and the Americas together as part of Spain’s colonial rule.Footnote 18 And, although there’s no documented link, its close resemblance to the Imperial law discussed in Chapter 2 (passed some 40 years later) suggests the Spanish law may have provided inspiration for that as well.Footnote 19
4.2.1.2 Time-Based Triggers as of 2020
In Europe, however, the Spanish formulation was an outlier. Time-based triggers were relatively rare in continental Europe, compared to the use-based triggers we discuss below.Footnote 20 Those countries that did adopt them tended to make them considerably shorter than the Spanish model, allowing authors to benefit in their lifetimes. Examples include:
1. Bulgaria, limiting copyright transfers to 10 years in length;Footnote 21
2. Italy and Spain, limiting the duration of publishing contracts (to 20 and 15 years respectively);Footnote 22
3. France, limiting rights granted by playwrights to five years;Footnote 23
4. Spain, limiting assignments of rights in ‘theatrical and musical performance’ rights to five years.Footnote 24
5. Portugal, mandating that grants of exclusive rights in works lapse if the work ‘has not been used’ after seven years.Footnote 25
In addition, there is an EU-wide time-based reversion right that originated in the 2011 Term Extension Directive (‘TED’), which we discuss in more detail next.
4.2.1.3 Term Extension Directive
In 2011 the EU Parliament extended protection for recordings to 70 years from first lawful communication or publication, although rights in recordings that had not been lawfully communicated or published 50 years after fixation would still expire after that 50 years.Footnote 26 Our focus here is not on the debatable merits of that extension,Footnote 27 but on the accompanying time-based ‘use-it-or-lose-it’ reversion right that instrument gave to performers. In sum, it allowed performers to terminate contracts and regain rights in recordings 50 years after lawful communication/publication, so long as those works were not available in sufficient quantity or sufficiently available for public access.Footnote 28 If they wish to exercise the right, performers must notify the producer, who then has a year to fulfil both sale and public availability criteria, or else the rights revert.Footnote 29
The rationales for the TED’s extension of sound recording copyrights are firmly rooted in creators’ moral claims, with the Directive’s recitals referring to the ‘socially recognised importance of the creative contribution of performers’ and the undesirability of still-living creators being unable to benefit financially from their work, or prevent ‘objectionable’ uses.Footnote 30 The accompanying termination right is consistent with this recognition of an ongoing connection between creator and output: if their recordings are no longer being made available or sold, it’s the performer who has the greatest interest over what happens to them in their final 20 years of copyright.
Despite these laudable intentions, the evidence to date suggests that, in practice, the right has gone all but unused. We canvass that evidence and the probable reasons for the law’s failure below.Footnote 31
4.2.1.4 Time-Based Triggers: Author Benefits during Their Lifetimes
The above analysis shows that, while relatively rare, time-based triggers were present in EU Member States as of 2020, and that they often came into effect much more quickly than the US (35 years) or the UK/Commonwealth (25 years after the author’s death). As we explore further in Chapter 6, shorter time limits may encourage investors to actually use the rights they have been granted, while also allowing creators to renegotiate grants more frequently from a bargaining position informed by their work’s actual value.
4.2.2 Exploitation-Based Triggers
European countries have additionally developed a wide range of reversionary mechanisms linking control of rights to their ongoing exploitation. In this section we identify key examples of use-it-or-lose-it rights that existed as of 2020 through the lens of the biggest controversies that surround them, namely:
1. What kind of circumstances gave rise to a right to revert?
2. To what extent (if any) can rightsholders bypass these rights via contract?
4.2.2.1 What Circumstances Gave Rise to a Right to Revert as of 2020?
Use-it-or-lose-it rights may be general or specific to particular industries or situations.
4.2.2.1.1 General Rights.
The general model typically enables authors to reclaim rights over any kind of work in cases involving a lack of exploitation. The earliest historical example we could find was enacted in Austria in 1936, and entitled creators to reclaim their rights if they were not being used or if they were being used in a way that was prejudicial to the author’s interests.Footnote 32
Some countries, like Croatia, Romania, the Netherlands, Hungary, Germany, the Czech Republic, and Austria, had reversion thresholds that permit creators to reclaim rights that are being inadequately exploited.Footnote 33
Others, however, only appeared to allow reversion in cases of complete non-use. For example, Poland allowed creators to revert rights if the rightsholder had failed to begin disseminating the work within the agreed time (or two years if there was no agreed time for dissemination).Footnote 34 And the Danish provision only allowed termination for non-exploitation three years after the author had discharged their obligations.Footnote 35
We’ll discuss the competing merits of inadequate versus no exploitation thresholds below, in relation to Article 22 of the DSM Directive.
4.2.2.1.2 Industry or Situation Specific Rights.
Some European jurisdictions adopted approaches that tailor reversion rights to specific creative industries or situations.
For example, in countries including Portugal, Croatia, Finland and Sweden, contracts for rights in literary and musical works to be used in films could be terminated after certain delays in completing or distributing the resulting work.Footnote 36 Meanwhile, Spain provided that exclusive contracts for ‘theatrical and musical performance[s]’ could only last five years,Footnote 37 and authors could terminate them sooner if the work was not performed for a year after its initial performance.Footnote 38
Such industry-specific approaches show a sensitivity to the fact that copyright covers a wide range of cultural products, emerging from fields that each have their own (often very different) economic and industrial realities.
Some particularly interesting reversion rights have been enacted in relation to book publishing, which as we explained in Chapter 1 has a particularly strong reversionary tradition. For example, a 2014 French law entitles book authors to reclaim rights in a variety of circumstances, including for lack of exploitation, failure to provide accurate accounting statements, and where (at least four years after a book’s initial publication) no royalties have been paid for at least two years.Footnote 39 Unlike nearly every other European reversion right, this French law explicitly distinguishes between physical and digital exploitations: if publishers are appropriately exploiting a book in one form but not the other, authors can reclaim their neglected rights while leaving the other arrangements intact.Footnote 40 Meanwhile, Spanish and Lithuanian laws provide that if authors have licensed their publisher the right to publish their books in multiple languages, they can reclaim rights over any unused languages after five years.Footnote 41
The above examples are framed in ways that support creators wishing to take advantage of the new exploitation opportunities that flow from digital technologies and the internet. In other cases, however, rights have become outdated. For example, Furgal identifies Romania, Slovenia and Spain as among those that define ‘out of print’ by the number of copies remaining in stock – a formulation that made sense in the print context but simply does not translate to the digital context.Footnote 42 This demonstrates the importance of regularly reviewing statutory reversion rights to ensure they’re well suited to evolving social, technological and economic realities – a recommendation we develop more fully in Chapter 6.
As we explained in Chapter 1, it makes sense for publishers to draft contracts that take sweeping rights so they’re covered in the eventuality they decide to use them, even though it’s vanishingly rare for all of those rights to actually be exploited. It’s not their role to consider the collateral damage that might result from such broad terms locking up rights to wide swathes of a nation’s literary heritage. In the examples set out above, lawmakers appear to have recognised both of those realities, and intervened in an attempt to reach a more appropriate balance. Of course, whether or not those attempts have been successful depends on how these laws actually work in practice. There’s little evidence available about that, but we canvass what we could find below.
4.2.2.1.3 Situation-Based Triggers: Going Out of Business.
In addition to industry-specific triggers, some nations also grant authors entitlements to reclaim their rights in specific situations, most notably where the rightsholder has gone out of business.Footnote 43 In Austria, this entitlement applied for grants of reproduction and distribution rights in all types of works.Footnote 44 Belgium, France, Italy, Luxembourg and Spain all gave authors who had signed publishing contracts the right to reclaim their rights in such circumstances,Footnote 45 although there were variations like the Spanish law preventing reversion where the publisher had gone into liquidation but reproduction had already commenced.Footnote 46
Meanwhile, the French Intellectual Property Code allowed creators of audiovisual works to ‘request the termination’ of an audiovisual production contract if the company was pronounced in liquidation or it had not been engaging in ‘business activities’ for over three months.Footnote 47 The Belgian Code of Economic Law had a similar trigger for audiovisual works: creators of audiovisual works could ‘request the termination of [the] … contract’ if the producer had not undertaken any activity for over a year, or when the audiovisual work had not been resold within a year of the producer’s liquidation.Footnote 48 In Austria, an author could ‘withdraw’ from a contract if an exploiter had taken exclusive reproduction and distribution rights but that exploiter had since become subject to insolvency proceedings.Footnote 49 However, they could only do so if ‘the reproduction of the work ha[d] … not yet begun’ at the commencement of the insolvency proceedings.Footnote 50
Such interventions appear primarily concerned with ensuring that the disappearance of cultural investors will not result in works being unnecessarily lost, to the detriment of both the original author and the broader public. As we will explain more fully in Chapter 5, it’s important to provide legal or contractual mechanisms for authors to regain rights if rightsholders go out of business – otherwise, administrators may be legally required to hold on to them as business assets, even if they are no longer in a position to meaningfully exploit the works.Footnote 51
4.2.2.2 Other Carveouts
As shown above, European reversion rights may be limited by industry or situation. However, both general and specific rights sometimes had other carveouts as well.
4.2.2.2.1 Exempting Reversion When Rightsholders are Not Responsible for Inadequate Exploitation/Non-exploitation.
Some provisions bar creators from enforcing the general use-it-or-lose-it provision in situations where the rightsholder was not responsible for the inadequate exploitation/non-exploitation,Footnote 52 for example where the author was instead responsible,Footnote 53 or where it was caused by a third party, ‘a fortuitous event or force majeure’.Footnote 54 Dutch law prevented reversions from being enforceable where ‘the other party [the rightsholder] has such an overriding interest in maintaining the agreement that the creator’s interest must deviate according to standards of reasonableness and fairness’.Footnote 55
4.2.2.2.2 Exempting Specific Works from Reversion.
Some reversion rights exclude certain types of work. In Finland and Sweden, for example, the ‘reversion’ of exclusive rights to perform a work in public to non-exclusive rights following three years of non-use did not apply to cinematographic works;Footnote 56 and the right to terminate publishing contracts for failure to begin or continue exploiting their works did not apply to newspaper and periodical contributions, or compilations.Footnote 57
Italian law excluded collective works from authors’ right to terminate publishing and performance contracts where the works had not been published within two years.Footnote 58 Its 20-year time limit on publishing contracts did not apply to ‘encyclopaedias, dictionaries … sketches, drawings, vignettes, illustrations, photographs and the like, for industrial use … works of cartography … [and] dramatic-musical and symphonic works’.Footnote 59 Bulgaria and Poland both had carveouts for architectural works.Footnote 60
Whether each carveout category of work listed above should indeed be exempted is beyond the scope of this book. However, the range and variety of these carveouts indicates the importance of carefully considering works for which reversion is not appropriate (a recommendation we return to in Chapter 6).
4.2.2.2.3 Contractual Carveouts for Works Made in the Course of Employment.
Work-for-hire carveouts (which, as we showed in Chapter 3, can be weaponised to disenfranchise creators of statutory reversion rights) were also present in some domestic European copyright laws.Footnote 61 Unlike the US termination system, rights in works made in the course of employment could, in some countries, revert to the employees, but employers could negate this by contract.
For example, under Lithuanian law, rights in works created by employees in the course of employment reverted after five years (except in the case of computer programs), unless their contracts specified otherwise.Footnote 62 Poland permitted employees who made works in the course of employment to regain rights if the employer had not distributed the work within two years, again unless contractually specified.Footnote 63 And in Slovenia, rights in works made by employees in the course of employment (except computer programs, collective works and databases) were presumed to have been assigned for 10 years, following which the rights would revert to the employee for reassignment to the employer.Footnote 64
In all these cases, the statutory provisions expressly allowed employers to derogate from reversion rights through their employment contracts. However, some European countries imposed stronger protections for employees in respect of works they created in the course of employment. In the Czech Republic and Slovakia, rights in works created in the course of employment would revert to authors if the employer died or went out of business and there was no legal successor.Footnote 65 The laws of corporate succession affecting publishers and other rightsholders in these jurisdictions is beyond the scope of our analysis. However, we highlight these provisions as examples where employees had at least limited reversion rights that it appears employers could not derogate from via contract. As we discuss in Chapter 6, the non-derogability of reversion mechanisms (except in limited circumstances) is an important protection for creators, even if, as we discuss in Chapter 6, the nature and features of employment relationships may justify statutory carveouts.
4.2.2.3 To What Extent Can Rightsholders Bypass These Rights via Contract?
As we’ve seen throughout this book, the practical impact of reversion rights can be much affected by whether rightsholders are able to eliminate them via contract. For example, in what Hugenholtz describes as a ‘weak point’, the author-protective provisions of Germany’s Verlagsgesetz (Publishing Act) of 1901 were overridable by contract until at least 2000.Footnote 66
Rightsholders have also sometimes attempted to contract around the operation of reversion rights, so that even if they operated as designed (e.g. rights did return to creators), the rightsholders would ultimately still get them back. For instance, Guibault and Salamanca report that publishers in Spain have tried to contract around the statutory 15-year limit on publishing contracts by including automatic renewal provisions in their contracts.Footnote 67 They provide limited detail, but ‘automatic renewal’ suggests initial contracts would contain clauses mandating that contracts would renew for an additional 15 years once the initial period expires, then the contract would renew for a new 15-year period.Footnote 68 If effective, such stratagems would prevent authors from renegotiating publishing agreements at or around the 15-year mark, when they would have a greater idea of the work’s value (and would be able to negotiate from a position of greater bargaining power).
Some European nations took strict stances against contracting out. The Netherlands and Slovenia, for instance, completely prevented creators from waiving their general use-it-or-lose-it rights,Footnote 69 while Romania and Slovakia prevented their general use-it-or-lose-it rights from being waived ‘in advance’.Footnote 70
Other nations barred or permitted contracting out in specific situations. For example, in Hungary creators could waive their general use-it-or-lose-it rights for up to five years after the contract was executed or the work was delivered to the rightsholder, whichever was later.Footnote 71 The Austrian copyright law similarly prevented creators from signing away their reversion rights except during the first three years of any transfer.Footnote 72 In Germany, the general use-it-or-lose-it right could be contracted away if this was done via a collective agreement.Footnote 73
And in Denmark, the general use-it-or-lose-it provision could not be waived, ‘unless it is [for] a mere change of the outlined time limits’.Footnote 74 As discussed above, the Danish use-it-or-lose-it provision allowed creators to terminate contracts for non-exploitation within three years.Footnote 75 Allowing these time limits to be varied could mean that rightsholders require creators to agree to longer notice periods than are reasonably required. Whether such contractual practices have taken place in relation to Danish works would require further investigation beyond the scope of this book. However, we note the allowance for such variation creates the potential for rightsholders to effectively neuter creators’ ability to regain their rights, by making them wait for lengthy periods before those rights return to them.
4.2.2.4 A Rich Range of Reversion Rights – but What Was Their Impact?
The above analysis shows that, as of 2020, EU members had already made a broad range of reversion mechanisms available to authors. These went well beyond the unsatisfactory time-based triggers adopted across Anglo-America – but have they been any more effective in practice? In the following section we canvass the evidence.
4.3 How Effective Were These Rights in Practice?
4.3.1 Little Evidence of Efficacy – or Harm
Despite the range of statutory reversion rights in force across the EU before the DSM Directive, including the kind of use-it-or-lose-it rights that we argue show real promise for promoting both of copyright’s core aims, there’s little evidence any of them have meaningfully assisted creators in practiceFootnote 76 – and indeed some evidence that they failed to do so.
One explanation for the lack of positive evidence is that there have been few attempts to measure the effects of continental reversion rights. We were able to identify just five studies to have done so, despite searching all databases to which we have access and seeking input from local experts.Footnote 77
It may be that access and language issues have prevented us from locating further evidence, despite our best efforts. It is also possible that some statutory reversion rights are having meaningful impacts for creators, but those benefits simply haven’t been measured or reported. However, the results of a 2013 EU consultation on copyright rules suggest that outcome is less likely. A report synthesising more than 9,000 responses to the consultation disclosed widespread discontent about the terms of exploitation contracts, with creators and their professional representatives accepting that rights often needed to be transferred to get works produced and distributed, but complaining about unfair terms arising from imbalances of bargaining power.Footnote 78 The issues they identified included widespread use of ‘buy-out’ contracts (which don’t permit creators any right to participate in the ongoing success of their works via royalties), lengthy licence terms with no possibility of renegotiation or termination, and inadequate transparency around how rights are actually being used.Footnote 79
If there were any examples of successful reversionary interventions in the laws of EU Member States, we would have expected to see them reflected in this evidence. Instead, the submissions called for EU-wide interventions to help creative workers secure better terms, including limiting the rights that could be transferred, giving creators the right to renegotiate contracts, imposing time limits on contracts, and making reversion rights available, especially in cases involving a lack of exploitation.Footnote 80 This all suggests that, while creators believed in the potential of reversion to help address these problems, existing legislative interventions were failing to do the trick. We now draw from the evidence that is available to us to identify the main reasons why those existing interventions may have been ineffective.
4.3.1.1 Ineffective Triggers
If reversion rights are to meaningfully benefit creators, they need triggers that enable them to do so. However, some existing rights permit reclamation only in circumstances where the copyrights are all but valueless. This appears to be the case for the TED’s hybrid time-and-exploitation based right, which doesn’t let performers get their rights back until 50 years after transfer – and then only in cases where they aren’t being made appropriately available, and where that hasn’t been rectified within a year of giving notice to the rightsholder.Footnote 81
You might wonder who would bother exercising such a tightly constrained right. The answer, according to the two studies to have evaluated the impacts of this law, is nobody.Footnote 82 As of 2018, it appeared that the right had not been invoked by a single performer. As Ramalho and Lopez-Tarruella note in their EU Parliament-commissioned evaluation, it had been available for a relatively short time at the time of these reviews, which might help contextualise this result.Footnote 83 Nonetheless, their finding of zero uptake remains remarkable and damning.
4.3.1.2 Uncertainty
Uncertainty can also play a key role in undermining reversion rights. As we saw in Chapter 3, lack of clarity about when rights accrue and what is required to exercise them can substantially depress creator willingness to actually do so.Footnote 84
Researchers evaluating the 2015 Dutch Copyright Act identified various uncertainties impacting its general use-it-or-lose-it right, which applies in cases where rightsholders fail to exploit works to a sufficient extent within a reasonable period.Footnote 85 In particular, they expressed concern about the lack of clarity around the interrelationship between this right and one entitling termination for breach under the general Dutch law of contract, as well as the meaning of ‘sufficient exploitation’.Footnote 86
Ramalho and Lopez-Tarruella criticised the TED right for uncertainty also. This stemmed largely from the Directive’s inclusion of ‘obscure and/or undefined terms’, particularly ‘sufficient quantity’ as the threshold for determining whether the performer can validly initiate the reversion process.Footnote 87
Only two countries defined this term in their domestic implementations, leaving considerable confusion elsewhere.Footnote 88 The resulting conflict in Portugal helpfully illustrates the problem. There, the performers’ collective management organisation took the view that ‘sufficient quantity’ should mean ‘sufficient quantity to be acquired in normal market conditions, that is with a balanced distribution in the territory at stake’, and that it should require producers to not just make recordings available on one online platform but on ‘a “significant part” of those that are active in the market’.Footnote 89 But the producers’ collective management organisation disagreed, arguing that ‘sufficient quantity’ should reflect ‘market demand’ and only cover physical copies.Footnote 90
These diverging views could lead to very different outcomes, especially given the fact that physical distribution of sound recordings has been largely replaced by streaming. Under the producers’ view, performers would have no redress so long as they were satisfying the (likely minimal) demand for physical copies. It is unclear which of these views (or some third alternative) will eventually prevail in Portugal, or how long it will take for the issue to be resolved throughout the remaining Member States. In the meantime, however, that avoidable uncertainty reduces the TED right’s potential value to creators even further.
4.3.1.3 Lack of Information
Reversion rights may also be undermined by creators being unable to access the data necessary to gauge whether the trigger is made out. This was the crux of yet another criticism Ramalho and Lopez-Tarruella made of the TED right: it doesn’t require producers to provide performers with access to the information they’d likely need to determine whether they were entitled to initiate the termination procedure. They recommended that Member States require producers to provide all necessary information to facilitate performers’ exercise of the rights, for free and within a reasonable period, but noted that none had yet done so.Footnote 91
4.3.1.4 Enforcement Challenges
Rights are only useful to the extent they can be enforced – and enforcement is particularly challenging in the context of reversion. This was made particularly clear in a 2016 investigation commissioned by the European Commission (‘the EC study’) to evaluate how author-protective interventions in 10 EU Member States (including some reversion rights) affected creator remuneration.Footnote 92 Its economic modelling found that the imposition of contractual obligations such as reversion rights could strengthen authors’ bargaining power, but that difficulties creators faced in actually enforcing those rights substantially weakened their efficacy.Footnote 93
Of particular relevance, the study found that most of the statutory termination measures in the EU Member States surveyed (as well as clauses requiring initial publication or permitting a revision in remuneration) needed to be enforced by judges.Footnote 94 As well as making exercise of the rights unfeasibly expensive, the study’s authors found this could deter creators in another way: ‘… requiring action to be taken by authors against the publishers … puts them in a position that might jeopardise their relationship with the latter; this could lead to authors being less willing to enforce such protective measures.’Footnote 95
Creators are understandably concerned about the risks of retaliation or other adverse consequences if they take action against their investment partners to enforce their rights.Footnote 96 Such risks are exacerbated by the fact that, as explained in Chapter 1, culture industries typically feature both a relatively small number of buyers for creative work and an oversupply of creative labour.
The aforementioned study evaluating the effects of author-protective reforms in Dutch copyright law found that such concerns are likely to play ‘a major role’ in explaining why its fair remuneration provisions had been so little used.Footnote 97 ‘[M]ost authors, out of fear of loss of contracts or blacklisting, do not dare to invoke or enforce their right to fair remuneration against exploiters.’Footnote 98
Such concerns are less relevant in the context of rights that don’t apply until the creator is near the end of their career (like the US termination right) or dead (like the Imperial right), but are nevertheless important and often overlooked when considering the real-world practicability of rights that rely on creators asserting them while their arts careers are still active.
4.3.1.5 Poor Design
Reversion rights can be undermined by poor design in other ways as well, including by failure to consider their interactions with other parts of copyright law (and the potentially unforeseen market impacts of such interactions).
For example, Ramalho and Lopez-Tarruella report that the TED use-it-or-lose-it right will likely never be used in the UK as a result of the way it was implemented.Footnote 99 The Copyright, Designs and Patents Act 1988 provides that when performers trigger that provision, the copyright in that sound recording expires.Footnote 100 This would mean that the producer would no longer be responsible to pay equitable remuneration for use of the recording, and the performer would then need to seek such remuneration from ‘[the] person who plays the sound recording in public or communicates the sound recording to the public’ (which they are entitled to even after the sound recording’s copyright expires).Footnote 101 As Ramalho and Lopez-Tarruella note, performers might be disincentivised from exercising the TED’s use-it-or-lose-it provision because it’s uncertain how they would get that equitable remuneration from the person who plays the recording in public or communicates it to the public.Footnote 102
The UK Intellectual Property Office’s (‘IPO’) study into the TED’s implementation in the UK also highlighted potential issues with the use-it-or-lose-it provision when there are multiple performers. The IPO noted that under the UK’s implementation of that provision, all the performers would lose out on access to continuing royalties if one performer invoked the use-it-or-lose-it right.Footnote 103 Ramalho and Lopez-Tarruella found other EU Member States had implemented the provision so as to require some type of agreement between the joint performers, or a majority of them, to invoke the right.Footnote 104 But the lack of such procedures in the UK highlights how domestic lawmakers can fail to design reversion laws to operate well given the realities of how works are created and distributed.
4.3.1.6 Rights Unsuited to the Digital Context
The lack of meaningful impact attributable to Europe’s reversion rights in the lead-up to the Directive may be a consequence of their being out of date.
In her 2020 analysis, Furgal identified France as the only country to have dealt explicitly with digital exploitation, explicitly entitling authors to end publishing agreements for lack of exploitation in digital or physical form.Footnote 105 While various other countries gave authors statutory ‘out of print’ rights, their usefulness was often limited by outdated formulations based on the realities of traditional print publishing.Footnote 106 Allowing authors to reclaim rights when their book is ‘out of print’ or has ‘less than 100 copies remaining in stock’ makes little sense in the context of publishing industries increasingly reliant on digital distribution and print on demand.
In some cases, the transition to digital technologies has drastically changed the practical meaning of reversion rights from what lawmakers originally intended. Consider the general use-it-or-lose-it rights that are based on a complete lack of exploitation, for example. In the analogue era, making a work available to the public ongoing necessitated substantial investments in production and distribution. In those circumstances, any level of exploitation suggested that the investor was still demonstrating a meaningful commitment to the work.Footnote 107 In the digital era, however, most works can be kept available for remarkably little outlay, and so the mere fact a rightsholder is still technically making a work available to the public doesn’t necessarily indicate an equivalent level of commitment to what it did in the past.
The significance of this shift was explicitly recognised in the evaluation of the Dutch use-it-or-lose-it right. Despite having a broader threshold than some of the laws we canvassed above (based on lack of sufficient exploitation, rather than no exploitation), the report criticised it for failing to take into account the changed realities that come with digital distribution: ‘As a result of digitisation, a work can be made available online indefinitely, almost at no cost.’Footnote 108 As Professor JG Visser has noted elsewhere, ‘there will thus hardly ever be any question of [non-use]’ under this provision.Footnote 109
Concerned that authors might be prevented from reclaiming rights in circumstances where rightsholders are technically but not meaningfully exploiting them, the report’s authors recommended that both availability and promotion of works be taken into account when determining the sufficiency of any exploitation.Footnote 110
4.3.1.7 Contracting Out
Finally, the ability of rightsholders to contract out of some European reversion rights may further explain their apparent lack of benefit for creators.Footnote 111 None of the studies we surveyed directly addressed this issue, but it’s worth bearing in mind given the ongoing significance of this issue in the context of Commonwealth and US reversion rights, as well as the disproportionate bargaining power rightsholders often enjoy relative to creators. As we’ll see, contracting out and many of the other problems identified above also affected the design and implementation of Article 22 of the DSM Directive, undermining the great promise of that instrument to harmonise, for the first time, baseline protections for creators across the Union.
4.4 Article 22: A Missed Opportunity
4.4.1 The Lead-Up to the Directive
As we’ve seen, European creators expressed a strong desire for better reversion rights in the 2013 consultation even though the ones already on the books were largely letting them down – and they set their eyes on the EU to achieve reform.
There are two key reasons why EU intervention may have seemed a more attractive bet than domestic law reform.
First, creators from strongly laissez-faire traditions (such as the UK, which was then an EU member state, Ireland, Malta and Cyprus) may have believed that EU intervention was the only feasible way of achieving the kinds of author protections that had been largely absent in their own domestic copyright laws.Footnote 112
Second, creator groups may have been aware of Big Content’s track record of derailing creator-favouring laws at the national level. One example was a proposed 2010 Dutch law that would have banned copyright assignments (permitting licences only) and allowed authors to terminate most licences after five years.Footnote 113 After furious responses claiming it would be the end of the Dutch publishing industryFootnote 114 and would ‘lead to the destruction of knowledge and capital’,Footnote 115 those provisions were dropped.Footnote 116 Creators seeking rights within the EU would still need to contend with fierce rightsholder lobbying, but if they succeeded in obtaining a strong reversion right, it would have to be implemented throughout all Member States with little opportunity for powerful rightsholders to derail it. That may have made the EU a better focus for creator efforts than individual legislatures.
Rightsholder responses to the 2013 consultation asserted that further regulation was not needed, that author and performer remuneration was adequate and the most important thing when considering fair remuneration is the author’s freedom of contract.Footnote 117 However, the European Commission accepted the evidence of creators instead, and acknowledged the need for legislative intervention by way of what would become the 2019 Digital Single Market Directive.Footnote 118
Its initial proposal was for a package of three author-protective provisions: a transparency obligation for rightsholders (e.g. to report on royalties to authors and performers), a ‘bestseller’ clause allowing renegotiation of disproportionate remuneration arrangements, and a voluntary alternative dispute resolution system accessible to creators.Footnote 119
Creators responded by lobbying hard for reversion rights to be included as well,Footnote 120 further demonstrating their belief in reversion’s potential to ameliorate problems with their contracts. That eventually caused the European Parliament to propose a general ‘use-it-or-lose-it’ provision that would allow authors and performers across the Union to revoke copyright grants where there was an ‘absence of exploitation’. Under that initial draft, authors could revoke rights grants when there was ‘an absence of exploitation’ or where rightsholders had failed to comply with a separate transparency right, aimed at giving creators better visibility of how their works were being used, and their payments calculated.Footnote 121 Contracting out of the reversion right was excluded except ‘if concluded by means of an agreement which is based on a collective bargaining agreement’.Footnote 122
The DSM Directive was intended to ‘achieve a well-functioning and fair marketplace for copyright’ across the EU,Footnote 123 making this a once-in-a-generation opportunity to achieve a healthier balance between the interests of rightsholders (to have a reasonable opportunity to exploit rights and benefit from their investments), creators (to be able to reclaim and benefit from unused rights) and the public (to have better access to cultural heritage via rights being freed to facilitate new exploitations). Unfortunately, for the reasons we unpack below, the eventual Article 22 signally failed to provide the best practice solution creators were crying out for.
4.4.2 The Eventual Article 22
The enacted version differed markedly from Parliament’s first proposal. We briefly canvass what that version looked like, before returning to trace its journey to enactment and then domestic implementation throughout the Union.
Article 22, as enacted, required Member States to ensure authors or performers could wholly or partly revoke grants or exclusive licences of rights where there was a ‘lack of exploitation’.Footnote 124 This revocation could only happen: (a) a reasonable time after the transfer, or the licence had concluded; and (b) once the author or performer had given notice as to when that exploitation needed to take place by.Footnote 125 The provision excluded reversion where ‘the lack of exploitation [was] … predominantly due to circumstances that the author or the performer can reasonably be expected to remedy’.Footnote 126 Article 22 could also not be enforced by creators of computer programs.Footnote 127
Article 22 allowed Member States significant leeway to craft the reversion right according to their specific contexts. These included the relevant factors for different types of works and creative industries, the matter of how reversion would operate for coauthored works, whether authors should be prevented from exercising the right after a period justified by the industry or type of work, and whether authors were able to choose to only end the exclusivity of the grant/licence, rather than revoke it entirely.Footnote 128 While some of this discretion was appropriate, the breadth of it gave powerful rightsholders greater scope to intervene in domestic implementations than creators may have hoped for.
Article 22 was accompanied by several other author-protective provisions: a requirement that Member States impose obligations on rightsholders to provide creators with rights to ‘appropriate and proportionate remuneration’ (Article 18); regular information about how their works were being used and how their share of income was calculated (Article 19); additional remuneration if the success of their works makes their initial remuneration arrangements disproportionate (Article 20); and a voluntary alternative dispute resolution (‘ADR’) mechanism to resolve disputes related to the transparency and contract adjustment provisions (Article 21). Member States needed to ‘ensure that any contractual provision that prevents compliance with Articles 19, 20 and 21 shall be unenforceable in relation to authors and performers’ (Article 23). As we demonstrate below, this meant Member States could technically allow investors to contract out of the application of Article 22 and its intended benefits for creators, even though those states could (but were not required to) limit contracting out provisions to those ‘based on a collective bargaining agreement’.Footnote 129
The DSM Directive became law on 17 May 2019,Footnote 130 requiring Member States to implement the Directive’s requirements into their domestic laws by June 2021.Footnote 131 Poland was the last Member State to implement the Directive in September 2024, despite previously challenging it.Footnote 132
4.4.3 How the Sausage Was Made
Before getting into the specific deficiencies of this provision, it’s worth providing a little more context about Article 22’s passage into law, and the changes that were made along the way.
As noted above, the Parliament included reversion in its amended draft legislation after the Commission had failed to do so in its initial proposal.Footnote 133 The first draft had two triggers: authors could revoke rights grants when there was ‘an absence of exploitation’ or where rightsholders had failed to comply with the new transparency requirement.Footnote 134 Additionally, it excluded contracting out except where collectively bargained.Footnote 135
Parliament passed that draft in September 2018,Footnote 136 but the reversion provision met substantial opposition during the subsequent trilogue negotiations between the Parliament, Commission and Council.Footnote 137 There were six rounds of negotiation, and as late as the fifth, still no agreement on whether to even include the reversion right in the Directive.Footnote 138 After the fifth round, however, with most of the Directive’s other content broadly agreed, a compromise text was finally agreed.Footnote 139
That final version weakened the Parliament’s original proposal in two critical ways.
First, it deleted the second trigger for reversion – the one that would have allowed creators to reclaim their rights in circumstances where rightsholders had failed to comply with their obligations under the accompanying transparency obligation. The impetus for this appears to have originated with the EU Presidency, which was adamant that it would not accept that trigger (albeit for reasons not apparent in the primary documents available to us).Footnote 140
We’ve previously flagged that author-protective provisions are only helpful to the extent they can be enforced. Had creators been given the right to terminate their contracts where rightsholders failed to provide them with the data to which they were entitled by law, it would likely have made it much easier to get them to actually do so. Unfortunately, the deletion of the second trigger meant this potential was never realised.
Second, the compromise text flipped the contracting out restriction. Whereas previous contracting out was banned except when the result of collective bargains (reflecting the German contracting-out model for its own use-it-or-lose-it provision, discussed aboveFootnote 141), Member States were now permitted (but not required to) impose such restrictions.Footnote 142 Effectively, that permitted rightsholders to contract out of domestic implementations of Article 22 unless Member States individually restricted them from doing so.Footnote 143 Once again, this left individual Member States vulnerable to lobbying from powerful corporations during their domestic implementations.
In addition, the law as passed changed the (now sole) reversion trigger from ‘absence of exploitation’ to ‘lack of exploitation’.Footnote 144 However, it’s not clear that this change made any substantive difference.
Below, we critically evaluate the ‘lack of exploitation’ standard and rules around contracting out, including the further problems that emerged during Article 22’s domestic implementations.Footnote 145
4.4.4 ‘Lack of Exploitation’
One of the most significant problems with Article 22 arises from its ‘lack of exploitation’ trigger. As explained above, some countries with general use-it-or-lose-it rights before the DSM Directive had based them on no exploitation, while others had more creator-protective standards that allowed rights to be reclaimed where exploitation was inadequate.Footnote 146
‘Lack of exploitation’ could have either meaning on its face, and neither view is unanimously adopted in the scholarship or by industry stakeholders.Footnote 147 Acknowledging the text’s ambiguity, we proceed with a plain reading of the phrase in the context of the Directive’s recitals, which suggest Article 22 is intended to be limited to situations where rights are ‘not exploited at all’.Footnote 148 The equivalent phrases in the Directive’s other published languages are consistent with ‘no exploitation’ as well.Footnote 149
Problematically, this high threshold doesn’t go as far as it might to further copyright’s access and reward goals. As we’ve noted at various times throughout this book, digital technologies make it very easy and often virtually costless to make most works available online.Footnote 150 In this paradigm, mere availability of a work does not say the same thing about the rightsholder’s commitment to it as it did in the analogue age. Under a ‘no exploitation’ standard, reversion can be avoided where rightsholders have made it available for sale online, even if they’re doing nothing to actively invest in making the work available in new formats, or promoting it to new audiences.Footnote 151
While a standard based on no exploitation is better than no revocation right at all, we query whether it’s desirable to allow rights to remain locked up on the basis of such minimal commitment – especially when others may be willing and able to do a better job.
A trigger based on exploitation being ‘inadequate’ in the circumstances would have been more consistent with promoting copyright’s access and reward goals. We agree with the recommendation of von Gompel et al in their evaluation of the Dutch use-it-or-lose-it right that, in determining whether the exploitation threshold has been reached, consideration ought to be given not only to whether the work is available, but the extent to which it is being appropriately and actively promoted.Footnote 152 In determining whether exploitation is in fact appropriate in all the circumstances, relevant circumstances might include the author or an alternative investor being keen to re-launch the work to the public, or make it available in new formats.
As we will argue in Chapter 6, any such standard would need to be supported by regularly updated industry-specific guidelines to help creators identify whether their work would qualify – and give rightsholders certainty about what they needed to do to retain them.Footnote 153 Done well, such a standard could do a much better job of promoting public access to knowledge and culture, and helping creators achieve recognition and financial rewards for their work.
The ‘no exploitation’ standard also raises a controversy over whether an author can reclaim some of their rights in cases where some but not all are being exploited. For example, imagine a situation where a publisher is exploiting its print rights over a book, but not its audio or ebook rights. Can an author revert just the unused rights?
Commentary of the European Copyright Society published to assist the implementation process suggests that Member States would need to explicitly make it clear that reversion is available in relation to each right in order to achieve such an end, suggesting that simple enactment of Article 22’s text would not be sufficient.Footnote 154 And the French implementation, for example, seems to rule this out explicitly. Its implementation provides that, ‘when author has transferred, on an exclusive basis, all or part of her rights, she may, in the absence of all exploitation of their work, automatically revoke all or part of this transfer’.Footnote 155 By contrast, as we saw above, France’s existing use-it-or-lose-it provisions for books not only use broader ‘inadequate exploitation’ triggers, but allow authors to reclaim part of their rights (e.g. digital or print) if only one of them meets the threshold.Footnote 156
Given the breadth of rights cultural investors tend to take in their dealings with creators, the fact that in winner-takes-all markets few works are likely to be fully exploited, and the ease with which minimal exploitation (such as making it available online) can be achieved, it’s difficult to discern a normative case for so strictly limiting the reversion right’s scope.
As the Directive only prescribes minimum standards,Footnote 157 EU Member States could have provided protections more in line with copyright’s access and reward goals by implementing versions that included a broader trigger – for example by enabling authors to reclaim rights in case of inadequate exploitation, as well as no exploitation at all.Footnote 158 However, only a minority of countries took the opportunity to do so. Eight countries already had a pre-DSM Directive general use-it-or-lose-it right with a trigger that included inadequate exploitation: Slovenia, Romania, Germany, the Czech Republic, Croatia, Austria, the Netherlands and Slovakia.Footnote 159 Six of these countries maintained that standard through implementation, ensuring authors had a broader right than the minimum mandated via Article 22.Footnote 160 But the remaining two states, the Czech Republic and Romania, actually wound that protection back, reducing their standard to allow reversion for non-exploitation only.Footnote 161
Most of the other Member States implemented the minimum ‘no exploitation’ standard, however (n = 18). Malta alone appeared to implement Article 22 by directly adopting its (rather ambiguous) ‘lack of exploitation’ language.Footnote 162
It’s unclear the extent to which countries decided to pass the minimum standard was influenced by rightsholder lobbying, as there was not always transparency around the implementation process. In Ireland, for example, which adopted an explicit ‘no exploitation’ standard, the government’s consultation did not ask for comment on the standard to be adopted, and did not make responding submissions available to the public.Footnote 163
4.4.5 Contracting Out
4.4.5.1 Contracting Out of Article 22 Is Now Allowed
We’ve seen throughout this book that when reversion rights can be avoided by contractual terms to the contrary, rightsholders routinely respond by using their superior bargaining power to require creators to sign away their reversion rights at the time they enter into the initial contract.Footnote 164 If we’re concerned with ensuring that copyright law promotes ongoing access to works and recognition and rewards for creators, it’s undesirable to allow rightsholders to so cheaply and easily eliminate protections designed to achieve this.Footnote 165
EU lawmakers recognised and provided for the risk of contracting out by requiring Member States to ensure contracts can’t bypass or neuter various other of the DSM’s author-protective provisions, but not Article 22.Footnote 166 By doing so, they have suggested that it is indeed possible to do so.Footnote 167 Member States can limit contracting out of Article 22 to situations where the relevant provisions were ‘based on a collective bargaining agreement’.Footnote 168 But again, this is permitted rather than mandatory, and thus further suggests that if they don’t so regulate, exploiters will still be able to contract out of revocation unhindered.Footnote 169
The enacted text effectively flipped the European Parliament’s original reversion proposal, which prohibited rightsholders from contracting out except as part of a collective bargain.Footnote 170 The closed-door nature of (and lack of transparency around) the trilogue negotiationsFootnote 171 makes it difficult to prove any causal connection between rightsholder lobbyingFootnote 172 and the shift away from the right being mandatory, but there is no doubt that the change benefited big businesses over creators.
Failure to make Article 22 expressly mandatory, when most of the DSM’s other author-protective provisions are mandatory, has been criticised for either ‘lack[ing] … coherence’ or indicating a desire to prioritise rightsholder interests over those of creators,Footnote 173 neither of which are consistent with the Directive’s creator-protective rationales.Footnote 174 It also made individual Member States vulnerable to lobbying from Big Content during the course of their domestic implementations: if they chose to actively depart from the DSM’s text to make reversion rights inviolable, they would not have any cover from the EU in doing so.
4.4.5.2 Some Courageous Implementations
Despite this, most (n = 19) of the EU Member States have implemented stricter approaches to contracting out than mandated by Article 22. Of these 19 states, only five (Croatia, the Netherlands, Slovenia, Spain and Slovakia), went further than required by the Directive, banning contracting out of the Article 22 equivalent in their domestic law completely.Footnote 175 The remaining 14 states permitted contracting out only in certain circumstances:
1. Eleven limited contracting out to collective bargaining situations (as permitted by the Directive);Footnote 176
2. Two (Austria and Hungary) retained their pre-DSM Directive restrictions on contracting out of their use-it-or-lose-it rights, preventing creators from waiving their statutory reversion rights beyond three and five years respectively;Footnote 177
3. One (Estonia) permitted contracting out if it was to the author’s benefit, but not their detriment.Footnote 178
However, the eight remaining Member States enacted rights which were silent as to whether it was possible to contract out,Footnote 179 which (for the reasons explained above, and the fact that other author-protective provisions in the copyright laws of these countries are mandatoryFootnote 180) strongly suggests that it is. Given the ease with which such terms can be inserted into contracts and the disproportionate bargaining power culture investors typically have over creators, we expect that this will translate to this right being of even less benefit to creators in those nations.
4.5 A Wasted Opportunity
Continental Europe has a tradition of protecting authors’ rights much more actively than common law nations, and that’s reflected in the panoply of reversion rights that existed in the lead-up to the DSM Directive, as well as the author-protective provisions that instrument included.
However, few of the reversion rights we were able to identify have been implemented in a way likely to meaningfully change outcomes for creators, and Article 22 is perhaps the biggest disappointment of them all. The DSM Directive was a once-in-a-generation opportunity to remedy this by achieving meaningful, technologically appropriate rights, but it signally failed to do so. While the opacity of the process makes it difficult to understand exactly what determined the final framing, the inappropriately high exploitation threshold, elimination of reversion as a consequence for rightsholder failures to comply with their transparency obligations and approach to contracting out resulted in a right that we suspect will do little to secure a fairer share of rewards to creators, or promote better public access to valuable knowledge and culture. Article 22’s use-it-or-lose-it right is better than nothing, but it’s far from what could have been.
Open access