Previous Next

What if we could reimagine copyright?


Reimagining copyright’s duration

Rebecca Giblin1


This collection’s foundational chapter revisited the vexed historical rationales for the grant of copyright. That uneasy juxtaposition of instrumentalist and naturalist motivations is perhaps most evident during debates about the duration of those rights. If we granted copyrights purely on instrumentalist grounds, we would grant the minimum we determined necessary to incentivise a socially optimal amount of creation. If we were driven exclusively by naturalist considerations, those rights would be perpetual.2 Instead, we are motivated by a combination of these desires. We want to incentivise creation of cultural and informational works in order to spread knowledge and culture, and we want to recognise the personality and labour authors pour into their works and reward them for those endeavours (or at least ensure that the fruits of their labours aren’t reaped by those who had nothing to do with the sowing). As this chapter will show however, current approaches to copyright terms – which international treaties lock us into indefinitely – are demonstrably counterproductive to both aims.

In 1906, author Samuel Clemens (known better as Mark Twain) testified to Congress in favour of extending copyright terms from the existing period of 42 years to a term of the author’s life plus 50. His claim was based not on the value of those books, but their lack of value: he argued that the commercial value of almost every book is extracted after its first few years, and doubted there were 20 Americans per century whose works were worth publishing beyond the existing copyright term. When copyrights expire, he argued, those few valuable books continue to be published, and the valueless continue not to be; the only difference is that the profits are diverted to publishers instead of authors or their heirs. In those circumstances Clemens argued there was no downside to giving lengthy copyrights to every work. That would enable authors to reap the benefits of those few that proved to be of lasting value, and the rest would be lost to obscurity regardless.3

This idea that there’s no downside in granting ever-longer terms may help explain the prodigious term extensions world legislatures have so casually locked us into over the last few generations. But while Clemens’ reasoning may have been sound a century ago, when high costs of production and dissemination made investment in obscure or unpopular works unfeasible, it no longer holds good. This paper demonstrates that current inordinate terms of protection are poorly justified by any of the dominant rationales trotted out in support of them, and result in real harms for authors and the broader public. It then shows how, if unconstrained by international treaty obligations and existing ways of doing things, we might retain incentives for creation, direct a bigger share to creators, and increase the social benefits that flow from access to cultural and informational works.

1. Justifications for current terms

Current approaches to duration

Since copyright’s inception, terms have ratcheted steadily upwards. In 1710, the Statute of Anne awarded rights for 14 years, reverting to authors for 14 more if they were still living after the initial term.4 Within a century, France had granted post-mortem rights (‘pma’ for post mortem auctoris) for up to 10 years after the death of the author.5 Those pma rights crept steadily upwards, with the Berlin Act of the Berne Convention recommending minimum terms of life plus 50 in 1908, and the Brussels Revision making them mandatory 40 years after that.6 Then in 1965 Germany unexpectedly (and with remarkably little debate) increased terms to 70 years pma.7 When the European Union harmonised copyright law in the 1993 Directive, all member states became obliged to match the German term, which was then the longest in Europe.8 In 1998, the US increased its durations to match the Europeans and, since then, has enthusiastically exported those extended terms still further through bi- and plurilateral trade agreements.9 The combined result of these various international instruments is that most countries now protect works for life plus 50 or 70 years; terms for other subject matter such as films and sound recordings have increased in parallel, and now commonly endure for 50, 70 or even 95 years.10 These terms are typically awarded as a ‘lump sum’ at the time of creation, and few jurisdictions seek to protect authors’ interests by limiting the alienability of those rights.11

We might expect that, before locking in such lengthy minimums via difficult-to-alter international instruments, policymakers would have given careful consideration to whether these constantly expanded terms would achieve their intended outcomes. That didn’t happen. As Ricketson has chronicled, the development of international norms governing duration has been ‘notable for an almost complete absence of debate of the policy and theoretical issues involved’.12

To fill the gap, this chapter evaluates the justifiability of existing copyright terms. As explored within the foundational chapter, the aims of copyright are many, varied and subject to fierce dispute. In the duration context however, we can identify four that are dominantly used to justify existing terms:

  1. To incentivise initial cultural production;
  2. To incentivise further cultural production by producing rewards that will subsidise investment in new works;
  3. To incentivise ongoing investment in existing works (i.e. to ensure their preservation and continued availability); and
  4. To recognise and reward authors for their creative contributions.

Do these rationales, in combination, justify existing mandated minimum terms?

Justification #1: That current approaches to duration are necessary to incentivise sufficient initial cultural production

Incentivising its initial creation is one obvious reason for granting monopoly rights over the fruits of creative endeavour. As the foundational chapter explained, this rationale motivates not only policymakers from instrumentalist traditions, but traditionally naturalist ones as well.13 The idea that monopoly rights are necessary to incentivise cultural production is at the core of the traditional economics approach to copyright.14 The expensive part of an informational ‘good’ is its initial creation; once created, it has characteristics of a public good, being non-rivalrous in consumption and minimally excludable.15 Economic incentive theory posits that, if a right holder has no right to impose monopoly rents, free-riding competitors driving the price to marginal cost would heighten their risk of being unable to recover their costs of initial creation.16 If competitors could simply come along and produce the same content without having to compensate authors, the original producer would be undercut. This suggests that some grant of monopoly rights is necessary to encourage the desired cultural production. (Of course, this is obviously not always the case, a point I’ll return to later.)

Significantly, existing minimum terms go far beyond incentive levels. To understand how far, it is necessary to consider two key concepts: the ‘time value of money’, and cultural depreciation. The time value of money recognises that, the further away in time a benefit will be received, the less it is currently worth. A dollar received today is worth a little bit more than a dollar received tomorrow, and a lot more than a dollar received a hundred years down the track. This means that, in determining the present value of future earnings, we need to discount them with reference to the applicable interest rate. In a brief to the US Supreme Court (‘the Eldred Brief’) a group of prominent economists, including five Nobel laureates, conducted some simple modelling to demonstrate how the time value of money affects incentives for creation. Assuming a real interest rate (net of inflation) of 7 per cent, they showed that a dollar today is worth $0.93 if received in a year, just $0.0045 in 80 years and a mere $0.0012 in 100.17

Works today are commonly covered by copyright for a hundred years or more. But given the time value of money, at the time of creation the value of those later years of protection is infinitesimal. Indeed, this model shows that the total present value of royalties received in the 80th to 100th years of protection is just 0.33 per cent more than the combined royalties for the first to 80th years.18 In other words, even if a work will still be generating a consistent stream of royalties in a century (an assumption I’ll return to shortly), the dollars that will be obtained in those later years are still worth virtually nothing at the time of creation.19 Of course, as the time of copyright expiry draws near, the present value of term extension becomes much higher. But in considering the scope of rights necessary to incentivise initial production and investment, it is the value of those rights at the time of a work’s creation that is relevant.

So the first important point is that the time value of money reduces incentives over time even if a work is still generating a consistent stream of royalties. The second point to understand is that very few works actually do so. ‘Cultural depreciation’ refers to works’ loss of market value over time. Cultural depreciation rates are important in determining the necessary incentives because, as the empirical evidence shows, the vast majority of a work’s commercial value is generally extracted shortly after creation.20 In most music genres, for example, revenues are typically a tiny fraction of their starting point within half-a-dozen years of release; in the case of books, the number of copies sold tends to drop sharply within a year.21 This isn’t the case for every work, of course, or even every sub-class. Classical music compositions are likely to have more enduring appeal than this week’s ‘Top 40’ hits, for example, and beloved novels such as Pride and Prejudice or To Kill a Mockingbird appreciate in the decades after their release rather than the other way around. As a rule however, most works have limited commercial lifespans. This can be illustrated by historical copyright renewal rates. Before 1976, US law granted a relatively short fixed term which could be renewed for a further period upon payment of a nominal fee.22 Prior to the introduction of automatic renewals in 1992, renewal rates ranged from 3 per cent in 1914 to 22 per cent in 1991, suggesting that even after a relatively short period, the vast majority of works had virtually no value to the owners of the rights.23 Those data also demonstrate that cultural depreciation rates differ across classes: far more films and music had their copyrights renewed, for example, than did technical documents, suggesting their value tends to be more enduring.24 But still, the commercial value of almost all works depreciates over time.

On the basis of the time value of money alone, Landes and Posner observe that the prospect of income that will be received 25 years in the future will have little impact on present decisions, and that the potential for royalties in a century’s time would do nothing to incentivise most authors.25 Pollock has developed a more comprehensive model factoring in not just the time value of money, but rates of cultural depreciation, the mathematical supply function for creative works, and the welfare associated with new works.26 Assuming a discount rate of 6 per cent and cultural depreciation of 5 per cent, his model estimates optimal copyright terms to be around 15 years.27 These economic models (and the dearth of empirical evidence to show that recent term extensions have resulted in an increase in cultural production)28 strongly suggest that existing minimum terms far exceed what can be justified as necessary to incentivise the initial production of works.29

Since current terms go beyond what is necessary to incentivise initial cultural production, they are justifiable only to the extent to which they satisfy other goals. What does the evidence have to say about that?

Justification #2: That current approaches to duration are necessary to incentivise new cultural production by subsidising investment in other works

Above-incentive terms impose additional costs on the creation of some new works by making it more expensive for subsequent authors to build upon them.30 But do they also result in greater investment in the creation and distribution of other cultural works? Longer terms are sometimes justified by claims that the windfall rewards resulting from above-inventive copyrights subsidise investment in new works, thus bringing about more new creation overall. The idea is that a right holder reaping profits from some successful works is more likely to reinvest them in unproven ones. If that’s so, it might offset the negative impacts current terms have on future creativity.

To establish the net benefit of such reinvestment, it’s necessary to distinguish between investments that would not have been made without those windfall profits, and those that would have gone ahead regardless. If the investment would have proceeded anyway, it cannot be attributed to the above-incentive grant of rights. In the Eldred Brief, those laureate economists also explained the circumstances in which rational producers might be expected to invest in new projects:

In general, a profit-maximising producer should fund the set of projects that have an expected return equal to or greater than their cost of capital. If a producer lacks the cash on hand to fund a profitable project, the producer can secure additional funding from financial institutions or investors. If the producer has resources remaining, after funding all the projects whose expected returns are higher than the cost of capital, this remainder should be invested elsewhere, not in sub-par projects that happen to be available to the firm.31

That is, rational producers will only fund projects which they expect to be profitable, and in those cases will do so as long as the expected return exceeds the cost of capital – regardless of whether that money comes from cash at hand or other sources. This suggests that a great deal of the investment paid for by above-incentive rewards would occur even in their absence. ‘If a producer pursues the same set of projects in any event, then its incentives will not be improved from the mere fact of a windfall from consumers.’32

The analysis changes, however, where a producer has no access to capital markets. Consider the individual artist or small publisher who is unable to obtain a loan or other source of funding. In such cases the profits generated from older projects may well be the only source of funding for new ones, and thus indeed lead to investment in creations which would not otherwise have occurred.33

Overall, this analysis suggests that the grant of additional incentives translates poorly into new productions that would not otherwise have been funded. In many cases the ‘reinvestment’ will result in works which would have been produced in any event. There is little justification for making blanket awards of above-incentive terms in the mere hope the resulting windfall will be reinvested in new creation.

Justification #3: That current approaches to duration are necessary to incentivise copyright owners to continue to invest in existing works

Another key rationale for terms that exceed what’s necessary to incentivise initial production is that they are necessary to incentivise copyright owners to continue to invest in existing works. This justification was relied upon heavily by supporters of the retroactive 20 year term extension enacted in the US in 1998.34 One of the most explicit references came from the then Register of Copyrights, Marybeth Peters, who testified to Congress that a ‘lack of copyright protection … restrains dissemination of the work, since publishers and other users cannot risk investing in the work unless assured of exclusive rights’.35 This reasoning suggests that, although the public has paid to incentivise the creation of the work, it must keep paying and paying again to persuade the right holder to continue to make it available.

In evaluating this justification it is necessary to consider two distinct kinds of investment: investments in continuing to make the original work available, and investment in works deriving from the original – such as audiobooks, films or plays from a novel.

Are existing terms necessary to incentivise copyright owners to continue to make original works available?

This argument assumes that nobody will invest in making existing works available if they have to compete with others providing the same content. The best empirical evidence in support of this claim is a study into the effects of US copyright law pre-1920, when it did not recognise copyright in foreign books. The study found evidence that publishers were hesitant to risk the heavy financial investments associated with typesetting and printing unless they had some guarantee of exclusivity.36

By contrast, a number of more recent and extensive empirical studies have found that works restricted by copyright are actually subject to less investment and narrower dissemination than their counterparts in the public domain. For example, research conducted in 2005 on behalf of the National Recording Preservation Board at the Library of Congress examined the exploitation of sound recordings created in the first 75 years of that medium’s availability. Under US law, sound recordings pre-dating 1972 can be protected by a patchwork of state and common law until 2067, with the consequence ‘that there are almost no pre-1972 US sound recordings in the public domain across the United States.’37 One of the key rationales for granting that lengthy term was to give owners an incentive to invest in making those older recordings available to the public.38 In a number of other jurisdictions however, those recordings have fallen into the public domain. This gives rise to a natural experiment: if it is true that works will be under-exploited in the absence of copyright, those older recordings should have been made more available by their owners, who are not only likely to have the best access to master copies, but greater incentives to produce them thanks to their ongoing monopoly rights. However, the study found that, on average, just 14 per cent of sound recordings published between 1890 and 1964 had been re-released by right holders on compact disc. Non-right holders re-released 22 per cent of those recordings without the benefit of any monopoly rights – over 50 per cent more than those that did.39

Similar findings were reached in Heald’s major study into the impact of copyright protection on the price and availability of books over time. Once again, US law provided ideal conditions for the experiment: books published between 1913 and 1932 had fallen progressively into the public domain from 1988 until 1997, but then a 20-year retroactive term extension ensured that books published from 1923 to 1932 would continue to be restricted by copyright until at least 2018. If the under-investment theory were correct, the data should show that the works in the public domain were being neglected by comparison to the works that were still subject to copyright. But once again the opposite proved true. Between 1988 and 2001, the public domain books were in print at the same rate as copyrighted books, and after 2001 they became available at a rate which was ‘significantly higher’.40 By 2006, 98 per cent of the public domain sample was in print, compared to just 74 per cent of the copyrighted works.41 Public domain books also averaged significantly more editions – 6.3 compared to 3.2 – and were available to the public at significantly lower cost.42 Only a small part of this discrepancy is explainable by the lower costs of digital distribution: there were still 5.2 editions of public domain titles available when ebook editions were excluded.43 Heald hypothesised that under-exploitation of works in the absence of copyright protection was likely only to be a problem where the costs of reproduction and distribution are high.44 There is less risk of ‘ruinous competition’ in the case of works that are easily and cheaply reproducible. In the digital world, this includes not only books, but also movies, music, photographs, software, music and more.

In a separate, subsequent study, Heald discovered that copyright seems to actually have a negative, rather than positive, impact on the availability of books. Using a random sample of 2,266 new editions of books available for sale on Amazon, and controlling for factors such as duplicates and multiple editions of the same title, the availability of copyrighted works was found to drop sharply shortly after release, before increasing dramatically upon entry to the public domain.45 When the figures are adjusted for the number of total books published in each decade, the trend became starker still.46 The lack of availability of print copies is not being filled by the ebook market either: though 94 per cent of public domain ‘bestsellers’ were available in electronic formats, the same was true of just 27 per cent of copyrighted books.47

The orphan works problem further illustrates how poor a job the blanket grant of long terms does of encouraging the continued availability of existing works. Countless millions of works are sterilised (in that their DNA cannot become part of future creation) or lost altogether because their owners cannot be traced.48 Very often, this is caused by works having commercial life spans shorter than their terms of protection.49 The range of orphaned works in cultural collections is enormous, ‘spanning published books, commercial photographs, journals, newspapers, television shows, films, sound recordings, plays and music compositions, as well as email messages, home videos, private letters, community pamphlets, postcards, government publications and other non-commercial ephemera.’50 Some 90 per cent of the photographs held by UK museums are orphans, as are about 13 per cent of in-copyright books;51 some EU archives claim orphaning rates around 40 per cent.52 Hargreaves decries this state of affairs as ‘cultural negligence’: ‘[a]s long as [it] … continues, archives in old formats (for instance celluloid film and audio tape) continue to decay, and further delay to digitisation means some will be lost for good’.53 The phenomenon is by no means limited to old works. A recent project to preserve New Zealand’s early video game heritage, for example, was unable to track down the owners of key pieces of software coded in the 1980s.54 There are people who wish to invest in preserving and disseminating these works; preventing them from so doing imposes all cost and no benefit on society. Even where adequate library exceptions exist, libraries have limited preservation resources. By the time such works finally enter the public domain, perhaps more than a century from now, many of them will be irretrievable from physically deteriorated and technologically obsolete containers.

Many other works are not technically orphans but are lost to society nonetheless because of what might be described as ‘parental neglect’. Copyrighted works are often so little valued that their owners don’t bother investing in making them available, despite having monopoly rights to do so. Unlike in Twain’s day, when high costs of production and dissemination meant that less popular works simply could not be kept available, many of these works could be rapidly and cheaply made available online for others to learn from and build on. Under current arrangements though, they are lost to society because their copyrights last longer than the owner’s interest in them.

The available evidence suggests that current terms do little to entice copyright owners to make their works available on an ongoing basis. In fact, the empirical evidence suggests they may have a negative impact on availability by making it more difficult for others to do so. A blanket grant of long terms in the mere hope they will lead to greater availability is poor policy. If we really are motivated to grant exclusive rights in exchange for continuing availability, those should go to those who do invest in those works, not those who could but don’t.

Are existing terms justified as necessary to incentivise investment in derivative works?

This one is easy, and can be answered by taking a quick look around our environment – derivatives of public domain works are everywhere. There have been countless reimaginings of Alice in Wonderland, for example, since it entered the public domain, including films, TV shows, illustrated and annotated editions, comic books, drawings, sculptures, and paintings. (Do an online image search for ‘Alice in Wonderland art’ to get an idea of the breadth of the inspiration drawn from that work.) Several of the film versions were made by Disney, a firm which famously built its empire on the retelling of stories from the public domain. Indeed, Walt Disney was so dedicated to using public domain works that he reputedly delayed making the first Alice film until he was completely sure the rights had lapsed.55 Disney has released dozens of major works based on public domain stories, including, notably, Snow White and the Seven Dwarfs (its first feature length animated film), Kipling’s The Jungle Book (released just a year after the story’s copyright expired), and recent hits Frozen and Maleficent (reimaginings of classic fairy tales by Hans Christian Andersen and the Brothers Grimm).56 Kipling himself drew extensively on other works in creating his story, and was drawn upon in turn by others (such as Edgar Rice Burroughs in writing Tarzan of the Apes).57 These anecdotal observations are confirmed by empirical work, finding, for example, that public domain works are exploited more than their copyrighted brethren in the case of audiobook recordings.58

The current explosion of creativity exploring the Sherlock Holmes universe further demonstrates enormous willingness to invest in creating derivative works from material in the public domain, even where that means competing with others who are doing the same. The grant of exclusive rights in these derivative works is more than ample to encourage their production. The grant of above-incentive rights in the original cannot be justified as necessary to achieve it.

Justification #4: That current approaches to duration are necessary to recognise and reward creators

This final justification is one of the most important (and perhaps the least understood).

As the opening chapter demonstrated, we are not only motivated by a desire to incentivise creation; we also want to reward creators with a bigger share of the social surplus of their creations in recognition of their contributions of personality and labour.59 Such motivations explain why even historically utilitarian copyright traditions often incorporate features such as moral rights, reversion, artists’ resale and performers’ rights.60 In Ricketson’s words, ‘There is a strong moral argument … that as authors confer benefits on society through their creative activity – the provision of learning, instruction and entertainment – this contribution should be duly rewarded.’61 Recognising and rewarding authors is fundamental to the function and legitimacy of any copyright system. Thus, even if creators would be willing to create works for relatively small incentives (or none at all), that does not mean that’s where the grant of rights should end.

The power and importance of rewards rationales can be seen in how persistently authors’ interests have been used to justify broader and longer terms of protection. For example, rhetoric accompanying the 1998 US term extension focused on the desirability of ‘[a]uthors [being] able to pass along to their children and grandchildren the financial benefits of their works’.62 The theme echoes Clemens’ entreaty to Congress of a century before. Such arguments are highly effective because they appeal to our inclinations to reward authors for their creative contributions. In practice however, relatively few of copyright’s rewards find their way to those creators. Indeed, such a huge proportion of the benefits of increased protection are captured by other cogs in the cultural production machine that authors are sometimes viewed as a mere ‘stalking horse’ masking the economic interests of others.63 In the case of the US term extension for example, the beneficiary of the unbargained-for windfall from the US term extension was the right holder at the time it was granted; very little of it accrued to the original author or their family if it had previously been transferred.64

Teasing out the reasons for creators’ poor economic outcomes requires revisiting that key tenet of economic incentive theory – that works will be created only if the fixed costs of production can be recovered, which is why we need to grant exclusive rights in the first place. This section canvasses the reasons why that rule does not always hold good in the context of creative labour, and how those realities might contribute to poor outcomes for creators.

As hinted at above, there are a great many intrinsic and extrinsic motivations for engaging in creative activities other than securing exclusive rights. Creative production might be triggered by external subsidies or prizes, assignments set by teachers, for the prestige associated with doing so, or because the creator is driven to do so by a problem she needs to solve, a story she needs to tell. Tushnet has written about the ways in which ‘the desire to create can be excessive, beyond rationality, and free from the need for economic incentive’.65 When authors explain why they write, they ‘invoke notions of compulsion, overflowing desire, and other excesses’ far more commonly than the urge to make money.66 In many cases, those reasons ‘are not the products of conscious choice or rational weighing of utilities’.67 This is not to say that a desire to make a living is not also a powerful motivator,68 but it does help explain why a great deal of creation occurs independently of economic motivations (or, at least, would occur even if those fixed costs of production could not be recovered).

Non-economic motivations may have translated poorly to outputs when high costs of production and dissemination meant the production of cultural works (inevitably) required significant financial investment. Today, however, many cultural artefacts can be made and disseminated for unprecedentedly little outlay, and there is evidence of widespread cultural production occurring independently of copyright’s incentives.69 Millions of individuals contribute product reviews, Wikipedia edits, and internet movie database entries without any expectation of financial reward. Others supply an extraordinary amount of original music and video to online platforms (Soundcloud has over 40 million registered contributors), produce free and open source software, dedicate their photographs to the public domain, and record audio versions of public domain novels to share online. As Boyle muses, that there seems to be some ‘innate human love of creation that continually drives us to create new things even when homo economicus would be at home in bed, mumbling about public goods problems’.70

There are other reasons too why creators may be inclined to create for smaller economic incentives than we would anticipate from rational economic actors. For one thing, creative labour can be more satisfying and enjoyable than other forms of work, and that can lead to individuals being willing to supply it at lower wages than they would otherwise work for.71 That’s why the Screen Actors’ Guild has to prohibit members from taking work that doesn’t comply with union minimums. As Johnson explains, ‘the only way actors can overcome the temptation to work for below scale is to enter into a group pledge to punish one another for doing so’.72

The desirability of artistic work is further highlighted by the fact that, though salaries in creative industries tend to be relatively low, there is nonetheless a considerable oversupply of artistic labour (defined as individuals wishful of engaging in arts work but who are un- or under-employed).73 Some creative production may also be explained as a consequence of over-optimism, with individuals overestimating their chances of emulating those who enjoy high-profile success.74

Creators’ keenness to practice their crafts can lead them to accept lower prices for their works than they would if motivated exclusively by their economic interests.75 As Towse explains, ‘[f]irms in the creative industries are able to “free-ride” on the willingness of artists to create and the structure of artists’ labour markets, characterised by short term working practices and oversupply, mak[ing] it hard for artists to appropriate rewards’.76 It also tends to put creators in relatively poor bargaining positions in their dealings with investors and intermediaries, which can lead them to transfer a great deal of the rewards of their creative labours to others.77 Rational firms don’t limit themselves to securing the minimum they need to justify their investment – they take everything they can get. For example, since at least 1919, the standard language in contracts between US music publishers and songwriters has commonly required the benefit of any and all future term extensions to vest in the intermediary.78 Even influential and famous composers have sometimes been forced by their publishers into sharing authorship credit and royalties.79 And Hollywood contracts often require artists to sign over their rights not just within the realm of planet earth, but throughout the universe at large.80 Thus, if a lucrative extraterrestrial market emerges, the bulk of its benefits will automatically be channelled away from creators. In the music industry, the ‘vast majority’ of musicians make little from their copyrights and performers rights, with the lion’s share going instead to intermediaries and a small minority of stars in more powerful bargaining positions.81

Given these realities, current terms are a poor tool for benefiting creators. The Gowers report observed that, ‘[i]f the purpose of extension is to increase revenue to artists, given the low number of recordings still making money 50 years after release, it seems that a more sensible starting point would be to review the contractual arrangements for the percentages artists receive’.82 Ricketson has also suggested that reform may better focus on formulating ‘appropriate safeguards for the licensing and assignment of their rights’.83 Some countries already have statutory limits to protect authors against too-broad transfers.84 In many jurisdictions however, creators often have little choice than to assign their economic interests in their works as a condition of investment or distribution.85 Given that current arrangements see so much of the benefit of above-incentive rewards diverted to intermediaries and investors, they are poorly justified as being for the benefit of creators.86

2. Reconceptualisation

The above analysis shows that the public is getting a poor return on its generous grant of above-incentive rights. Current terms are neither optimised to maximise continued investment in existing works nor to recognise and reward creators, and they cause knowledge and culture to languish underused, or even vanish altogether. If we were starting from scratch, unconstrained by international treaty obligations, existing business models and questions of political feasibility, how might we rethink our approact to duration – and do a better job?

Before getting to that question, it’s necessary to make explicit – a better job of what? What are my assumptions about what copyright policies should seek to achieve? There are tremendous possibilities for economic and social gains if we were to be more creative and explicit about copyright’s aims. However, this chapter’s rethinking is based on the assumption that we primarily wish to achieve the same aims that were identified above as justifying current duration policy. This is intended to enable a direct comparison of how much more effective copyright policy could be if it were freed from current proscriptions. Thus, I assume that we’re primarily motivated to:

  1. incentivise initial cultural production;
  2. incentivise further cultural production by producing rewards that will subsidise investment in new works;
  3. incentivise ongoing investment in existing works (i.e., to ensure their preservation and continued availability); and
  4. recognise and reward authors for their creative contributions.

So what might an alternative look like?

The foundational chapter set out a number of conceptual frameworks that might assist in determining where the public interest lies, including with reference to all individuals’ interests, the preponderance of those interests, or moral validity (assuming of course that moral validity is capable of determination). Here I adopt the framework of Ho’s ‘representative individual’, which usefully frees the mind from self-interest and vested interests, while desirably encouraging consideration of all aspects of the cultural production and access equation. The question then, is this: how would a person who had an equal chance of being anyone in society design duration policy to help achieve these aims?87

My starting point is to recognise that the twin desires of incentivising and rewarding creation need to be satisfied in different ways. Incentivising the creation of cultural and informational works may be achieved by granting just enough monopoly rights to give creators and investors a reasonable prospect of recouping their capital and some sufficient return on their investments. The grant of incentive-level rights is justified on the grounds that they are necessary for those works to be created. Rewarding creation requires securing to authors an additional share of the value of their creations above and beyond the incentive price. Here, the justification is that, even if they did not need it to create those works, they deserve it for having done so.

Having said that, even reward rationales don’t justify granting creators the whole value arising from their creations. Even in the current paradigm some parts of all works are always reserved to the public, including the ideas within them, de minimis or non-substantial takings, and tolerated or excepted uses. That’s because society is not motivated to incentivise or reward creation as an end in and of itself. Instead, we incentivise initial cultural production and ongoing investment in existing works to encourage the dissemination of knowledge and culture, and we reward authors (at least in part) out of gratitude for the benefits that flow from that access.88 Thus, some share of those benefits must be reserved to society as a whole. In considering where to divide the pie, the representative individual would be well advised to heed Lemley’s warning that the full internalisation of a work’s positive externalities may not only invite rent-seeking but could actually reduce them as well.89

The following sketch describes one possible way in which our representative individual might better achieve the four identified aims. It imagines a four-stage system of rights which would maintain incentives, increase rewards for creators, and simultaneously unlock a great deal of currently neglected value for the public. Since the very point of the exercise is unconstrained ‘blue sky’ thinking, it includes elements that would clearly be prohibited under Berne and other international instruments or simply politically unfeasible. The aim is to start the conversation about how we might do a better job if we were permitted to do so – and to raise awareness about what current treaty obligations force us to lose.

The incentive stage: A fixed initial term of exclusive rights

My reimagined approach begins with the grant of a fixed initial term of exclusive rights intended to incentivise initial production of the work (‘incentive rights’). These would be calculated at the approximate value needed to incentivise the creation of the work plus any initial investments necessary to get it to the market. While it would be impractical to determine this on a case-by-case basis, we could broadly classify different classes of works with little difficulty.90 Software, for example, would likely have a shorter term than films. Optimal figures would need to be approximated by economic modelling taking into account the complete contours of the law, particularly the scope of rights and exceptions. However, existing economic modelling suggests that, even for the most expensive works, it would be relatively short: given the time value of money and cultural depreciation, the potential to earn money from a work after the first 25 years would do little or nothing to persuade a rational investor to make the work in the first place.91 For the sake of illustration, the following explanation proceeds on the assumption of a 25-year initial term for all types of works.

As under the current system, the initial grant of incentive rights would not be subject to any formalities. Consistent with the approach favoured by Gangjee in the succeeding chapter, this gives artists an opportunity to assess which works are most valuable and reduce their overall costs of registration. This is important given the incredible number of works that can be quickly generated using existing technologies: a photographer who takes tens of thousands of pictures every year will need some time to establish the ones in which she has an ongoing personal or economic interest. It is also desirable to ensure individual artists are not disadvantaged relative to commercial producers. As Elkin-Koren has explained, authors may initially create without any profit motive, but still seek to share in the rewards in the event that their creations achieve popularity.92 An upfront registration requirement may prevent them from doing so; it may also enable corporate players to co-opt the fruits of their labours.93 As demonstrated by the use of the GPL (GNU General Public License) for open source software, copyright can be a useful tool for preventing corporations from free riding on unpaid creative labour.94

The reward stage: Rolling creator-rights

After the initial fixed period, copyright in the work would expire. Then would commence a second stage of rights. Here the reward rationale comes into play, and the overall regime starts departing sharply from the status quo.

After the expiry of the initial incentive-value rights, creators would be entitled to assert their continuing interests over works by registering for a ‘creator-right’. Creator-rights would give their owners slightly less comprehensive rights than the initial grant of rights. As noted above, even in the current paradigm some parts of all works are always in the public domain. When a work moves from the incentive stage to the reward stage, some additional uses could join that list. They might include enhanced rights to engage in non-commercial transformative uses and a broader range of unremunerated educational uses.95 Key national cultural institutions might gain greater rights to engage in preservation activities; given the cheapness and ubiquity of storage and transmission, systems might even require deposit of copies to guarantee future preservation and access.

An increased range of uses might also be permitted-but-paid:96 subject to compulsory licences that require payment of equitable remuneration. This could facilitate the creation of social welfare-enhancing initiatives such as digital public libraries. This could enable us to remunerate authors on a per-loan basis similar to some existing public lending right arrangements while simultaneously increasing access for the public. The use of such mechanisms could also help facilitate the emergence of new distribution platforms and make it easier for artists to find audiences. Compulsory-type licences could also help facilitate uses of works which had a great many creative contributors, as is often the case with films.

For uses falling outside the more expansive exceptions and compulsory licences, creators would be entitled to make fresh bargains with intermediaries and investors, and have veto rights over undesired uses. Thus in many cases the author would retain the right to exclude. This is not inconsistent with the broader ‘access to culture’ aims outlined above; it is simply part of the evolving balance between improving access to culture while recognising the creator’s continuing interest in her works. Alongside this system of rights there would need to be an appropriately tailored moral rights regime existing independently of registration in order to protect artists’ non-economic rights.

Careful thought would need to be given to drawing the boundaries of the scheme, and particularly to the kinds of creators who would be eligible for these rewards. Existing legal systems already draw that line in different places; such differences could no doubt be accommodated in this new system as well. Regardless of where the boundary falls however, creator-rights would vest only in those who made creative contributions. They would not benefit mere investors, and, vitally, they would not be transferable until shortly before the expiry of the previous term of rights. This arrangement would impose little in the way of fresh burdens – an artist receiving a creator-right could assign it in its entirety each term if she wished to do so – but it would position creators to reap more of the fruits of their labours. It would also open up new opportunities for investment by individuals or firms who believe they could do a better job of exploiting them than those who previously held the rights.

Unlike incentive rights, creator-rights would arise only upon registration. (In the current paradigm, of course, this would be prohibited under Berne.) That threshold requirement would make it simple for prospective users to determine whether works have entered the public domain and reduce the tracing costs involved in getting in touch with the creator to negotiate licences for those that have not. By requiring contact details to be kept up to date, orphan works could be virtually eliminated. Gangjee more deeply explores the potential of formalities to fix the deficiencies of current approaches, including those caused by too-long terms of protection, in the following chapter. By imposing a proactive obligation on authors to register their continuing interests in works, the system would dramatically reduce the losses of knowledge and culture that occur by granting lengthy blanket terms irrespective of whether the author or owner has any continuing interest in their work. While it may sometimes be difficult to identify who made what creative contribution, and apportion the rewards, those are the same difficulties we already confront in determining ownership. In any event, they could be significantly ameliorated by the development of policies, guidelines and precedents as the system became established.

This is a system of rolling rights: after each period of rights draws to a close, a further registration would cause a new creator-right to vest afresh in the author. Granting creator-rights for a period equal to each full incentive-based term ensures there will be enough protection to encourage even the most lavish new investments. That leads to the question: how many renewals might there be? As explained above, though the rationale of rewarding and recognising creators dictates that they should receive a greater share of the rewards of their work, it does not justify infinite protection. Sooner or later the rights must expire for good – but when? It is easy to make the case the creator should be entitled to continuing rights over her lifetime, but less clear that it should be transferable to heirs after death. While the idea that copyright terms should allow authors to provide for their dependants has ‘almost assumed the status of an article of faith’, significant changes in social conditions since the 19th century reduce the power of the rationale,97 as do systems which reserve for others the benefits ostensibly intended for descendants.98 However, as outlined above, enabling authors to provide for their descendants has a powerful moral appeal. As long as protection is granted on an opt-in basis rather than as default, post-mortem protection would ensure fairness in outlying cases while causing little harm elsewhere.99 Empirical evidence could be gathered to help determine when creator-rights should ultimately expire. It might be however that in the case of works involving a limited number of authors, creator-rights could be renewed for life plus a generation; larger ensemble works might last until the demise of the last registered creator. This would be unlikely to comply with the Berne minimum of author’s life plus 50 years (and almost certainly be shorter than the life-plus-70 terms that trade agreements have locked in to an increasing number of domestic laws).

The transition stage: Between expiry and registration

Artists could be encouraged to preregister their works so they can be alerted when the initial expiry of rights draws near. However, even with strong awareness about how the system works, some creators would inevitably fail to register works: perhaps because they do not realise that they still hold value, or simply because they forget to do so. To avoid that, there could be a transition stage after the previous expiry of rights during which time creators would be permitted to retroactively register their continuing interests.

Use of works in the transitional stage should be encouraged by preserving those who engage in them from obligations to pay for past use. That’s because they would have an important signalling function informing artists about continuing interest and value in their works. New industries would no doubt spring up to identify works which were proving to have continuing value in the period after the initial lapse of rights, and to assist creators in registering their continuing interests in works.

The transition stage would end upon registration (with the works then being protected by creators-rights) or lapse of time.

The public domain

The final stage is entry into the public domain. This could occur upon donation by the creator, after failure to register by the closure of the transition window, upon lapse of creator-rights (i.e. non-renewal) or at the point of ultimate expiry, where no further renewals are permitted. As noted above, appropriately tailored moral rights would be a necessary complement to protect authors’ non-economic interests. Though thinking about their shape is outside the scope of this chapter, the need for them would be particularly important given that authors who chose not to register continuing interests might well outlive their economic rights.


The following case studies illustrate how the above-described system might work in practice:

  1. A global digital public library obtains a blanket licence over all books of 25 years and older. Where their authors have registered for creator-rights, the library pays them equitable remuneration each time one of them is borrowed by a user until such time as they enter the public domain.
  2. An author writes a book, and gains a 25-year copyright. Assume a film studio exclusively licenses the film adaptation rights, and makes the book into a film five years later. The film is protected for its own 25-year term. When the book reaches the end of the copyright period (20 years after the film is made), the author registers her interests and receives a creator-right in her book. That enables her to negotiate with others to make new derivative works including further film adaptations (or perhaps even reach a deal with the original studio not to agree to grant a licence elsewhere for the term of the creator-right). When the copyright on the film expires, the author of the book may register a creator-right over that as well, in recognition of her creative contribution to the work.100 She would share in the proceeds of the film’s exploitation with its other creative contributors.
  3. A number of 25-year-old computer games have fallen out of copyright and not been registered. Shortly after, a revival of interest in that era results in people beginning to convert those games to currently playable formats. An investor tracks down the original programmers and asks them to register their creator-rights and grant a licence to enable the investor to commercially exploit the works. Those who previously converted the games have no liability for doing so, but can no longer commercially exploit them. Some right holders cannot be traced; when they don’t act within the registration window, their games officially enter the public domain.
  4. A band records a song, and assigns the copyrights in music and lyrics to their record label. After the initial term of protection lapses, the band members register creator-rights. Anyone can continue to use the sound recording, subject to equitable remuneration being paid to the creators. A producer wants to put it in a movie soundtrack, negotiates a licence to do so, and the proceeds are shared among the creators.
  5. A band records several albums but none of the creative contributors register for creator-rights. After the works enter the public domain, another musician samples some of their tracks and includes that material in her own release without requiring any licence to do so.
  6. An artist produces various culturally important photographs, posters and political cartoons, and, when the copyrights expire, registers her creator-right over each of them. Non-profit cultural institutions are freely able to make copies for preservation purposes, but commercial users are obliged to negotiate a licence.

Benefits of this reimagined approach to duration

This reimagining of duration would have a number of benefits over the existing system.

First, it would secure to creators a larger slice of the pie, regardless of their relative lack of bargaining power or legal nous. This might include the session musicians who contribute instrumentals or vocals at live or recorded performances, typically in exchange for a flat-fee rather than a right to a royalty.101 It could also benefit those who make creative contributions as part of large ensembles, giving some ongoing financial interest in their films (in contrast to the current system, which typically sees investors ending up with the bulk of the rights).102 By better securing the rewards of copyright to creators themselves, this reimagined system would also enjoy greater legitimacy than existing approaches. For reasons explained by Weatherall later in this collection, that could in and of itself do a great deal to promote effective enforcement of rights.103

Another crucial advantage of the proposed approach is that it would facilitate the emergence of new distribution platforms. This is important, because distribution, visibility and access to their outputs are vital to creators. As author Heidi Bond explains:

Distribution is an issue – it doesn’t matter how many people have potential access to your book; if it’s not in a distribution channel that is, in fact, accessed by people, it doesn’t exist. Access to distribution channels – and the decreasing number of those channels – is a serious issue. Visibility is an issue.104

Thus, the attractiveness of new distribution mechanisms depends on the size of the audience they reach. The ‘network effect’ is the value that one user of a network has to others. The more users there are, the more valuable the network becomes. The catch is that those users will come only if the service is offering the content they want. Accordingly, new distribution platforms often live or die depending on whether they can secure the rights to existing popular artists and material. In many fields, rights are highly concentrated among a handful of intermediaries. That gives those right holders considerable power to shape emerging new platforms in ways that align with their interests. Consider the example of online music streaming. It’s extremely difficult to attract listeners to such a service without securing the rights to popular music, which makes it necessary for new providers to do deals with existing major labels. Not surprisingly, such deals end up being negotiated to suit the record companies, and not the artists (who typically neither control the rights nor have a spot at the negotiating table). Thus the majors took a stake in Spotify as a condition of access to their catalogues, and pay only a small fraction of the royalties they receive from the service to artists.105 A recent industry study found that the after-tax payments of French music subscription services were split 73 per cent to major labels, 16 per cent to writers and publishers, and just 11 per cent to performing artists.106

Since artists are often poorly placed to negotiate a better deal (for the reasons already discussed), the current system perpetuates the status quo, with creators obliged to sign over their rights in order to get the visibility and access that are so often the keys to success. By putting works back up for grabs after 25 years, the proposed system would reduce the concentration of rights and facilitate the emergence of new platforms that could offer creators a better deal.

My reimagining has some commonalities with the idea of the ‘paid public domain’ (known in France as a domaine public payant, and Germany as Urhebernachfolgevergütung). As first mooted by Victor Hugo in 1878, the concept would have seen works become the property of the nation immediately upon the author’s death, subject to a fee being paid for their exploitation in perpetuity. Revenue would go to direct heirs during their lifetimes, then to a general fund to support emerging authors.107 Variations on the concept have since been adopted in a number of countries: some require payments to be made in perpetuity while others are limited in time; some attract remuneration for commercial uses only; fees can be flat percentages of revenue, lump sums, or determined by an independent body; some apply only to certain classes of works. Rights may belong to the state, authors’ societies or unions. Some require prior permission, while others allow any use upon payment of a fee.108 The author-right scheme envisaged in this work achieves many of the social and cultural advantages sought by paid public domain schemes while minimising their detriments. By operating on an opt-in basis instead of capturing every work by default, it would enable the public to obtain the full benefit of works that are worth less to authors than the nominal cost of registration. Combined with ongoing renewal requirements, it would enliven rather than stultify the public domain. Further, by allocating the rewards to specific authors rather than a general pool for reallocation, it avoids criticisms that have been made of some paid public domains: that they give the state too much control over culture109 or that the revenues are ‘just another tax’.110


This reimagined system doesn’t just envisage giving creators a bigger slice of the cultural pie – it seeks to make the pie itself bigger. The online registry would actively facilitate markets between creators and exploiters. By increasing the opportunities for both commercial and non-commercial exploitation, reducing tracing and other transaction costs, and unlocking the value of works in which the creator has no further interest, the system would likely increase both revenue and consumer surplus. At the same time, the social welfare costs of granting rights in excess of what is necessary to incentivise production would be reduced by carving out certain socially valuable uses from the scope of the creator-right. That, combined with greater use of compulsory licenses, would facilitate the development of a broader range of initiatives for preserving and sharing our cultural heritage, including via digital public libraries, music repositories, film archives and more. Overall, the proposed system would maintain incentives while substantially improving access and securing a greater share of revenues to creators.

Despite its advantages however, this reimagined system would be impossible to implement. That is not because of technical limitations: the widespread ability to cheaply compile, store, communicate and access information, and the low transaction costs now involved in tracking usage and transferring funds, would make such a system perfectly feasible right now (at least in countries with developed communications infrastructure, which are also the biggest producers and users of copyrighted works).

It’s impossible because we are welded to inexorable copyright terms via international treaties that make it ‘virtually impossible to deal with term on a logical basis’.111 Various elements of this proposal would fall foul of existing international obligations: by being shorter than mandated minimums, by requiring formalities, and by introducing carve-outs that could violate the three-step test.

It would technically be possible to adapt elements of this proposal into an authors’ reversion scheme which would still do a better job than our existing approach (by at least securing more of copyright’s above-incentive rewards to creators) while being compliant with international law. However, such a compromise would do little to reduce the extraordinary societal loss that arises from the automatic locking away of our cultural heritage regardless of authors’ continued interest in their works. And even this far-from-optimal solution would be politically infeasible in practice. Despite growing evidence that existing terms are actively counterproductive to what we want to achieve, there is persistent political pressure to extend them still further.112 The powerful intermediaries who benefit most from existing frameworks and engage in constant rent-seeking to expand the rights of owners would fight ferociously against any proposal to allocate more rights to authors.

In recognition of this intractable practical problem, Landes and Posner have proposed a system of ‘indefinitely renewable’ copyright.113 It would permit those right holders who indicate continued interest in their works to renew them ad infinitum. They argue that this would reduce the rent-seeking which occurs when still-valuable works near their expiry of protection, and at least cause those unwanted or less valued works to enter the public domain.114 Such a system would do little to incentivise the production of additional works (for the reasons discussed at the beginning of this paper). And it would do little to recognise or reward the author’s creative contributions, since the benefits would go to those who owned the works at the time of renewal. In fact, it would perpetuate the problem of copyright’s rewards being co-opted by entrepreneurs or investors rather than creators or their heirs. But the proposal is nonetheless attractive, despite being unjustifiable on any recognised rationale for protection, because it would at least unlock a great deal of the culture that is cumulatively valuable to society despite being of no further interest to its owners. This kind of deal is perhaps the best we could hope for given current legal and political realities.

It is not the point of this project to propose such compromises. This chapter unabashedly proposes an ideal. Copyright could do a much better job if freed it from existing constraints. We could design terms to simultaneously incentivise ongoing creation, increase social welfare, give creators a fair go, improve the preservation and dissemination of our cultural heritage and eliminate orphan works. Currently neglected value could be unlocked. We could make the pie bigger. If we better understood our motivations for granting copyright protection, and acted on them by disaggregating incentives and rewards, authors and the public alike could strike a far better deal. But although the ideas proposed for achieving these aims are not revolutionary, current frameworks simply won’t admit them.

1 Faculty of Law, Monash University. A Research Accelerator Grant from Monash University substantially supported the development of this work (and the project) over 2013–2014. I am also grateful to the Centre d’Études Internationales de la Propriété Intellectuelle at the Université de Strasbourg which hosted our project workshop in 2014, and to each of the project participants for the valuable suggestions and feedback they provided during that process. Some of this work was completed while I was in residence as a Senior Visiting Scholar at Berkeley Law School. Thanks also to Professor Jane C Ginsburg, who rigorously critiqued an earlier version of this paper.

2 Indeed, perpetual copyright was ‘nearly adopted as the ideal standard’ in the preliminary negotiations to the original iteration of the Berne Convention. Sam Ricketson, ‘The Copyright Term’ (1992) 23 International Review of Intellectual Property & Competition Law 753, 755, citing Sam Ricketson, The Berne Convention for the Protection of Literary and Artistic works: 1886–1986 (Kluwer, 1987) at 41–42. Ricketson argues that recognition of the interests of the public is a key reason why natural rights theories ‘have … never triumphed in their pristine form’. Sam Ricketson, ‘The Copyright Term’ (1992) 23 International Review of Intellectual Property & Competition Law 753, at 755. Even countries which once offered perpetual protection, such as Mexico, Guatemala and Portugal, have now reined it back. Sam Ricketson, ‘The Copyright Term’ (1992) 23 International Review of Intellectual Property & Competition Law 753, at 755 n 7.

3 James Boyle and Jennifer Jenkins, Intellectual Property: Law & the Information Society (CreateSpace Independent Publishing Platform, 1st ed, 2014) 284–288.

4 Act for the Encouragement of Learning (Statute of Anne), 8 Ann, c 19 (1709/1710), section c11 (Great Britain).

5 Gillian Davies, Copyright and the Public Interest (Sweet & Maxwell, 2nd ed, 2002), 136–137.

6 From the time of the Berlin revision in 1908, a term of 50 years pma had been recommended, but it did not become a mandatory minimum until the Brussels Act of 1948. See Ricketson, ‘The Copyright Term’, above n 2, 777–778.

7 Gillian Davies, above n 5, 201.

8 See Council Directive 93/98/EEC of 29 October 1993 harmonizing the term of protection of copyright and certain related rights [1993] OJ L 290/9, recital 11 (emphasising the importance of harmonisation as justification for the increase to 70 years pma protection).

9 See e.g. Australia–United States Free Trade Agreement, signed 18 May 2004 [2005] (entered into force 1 January 2005), Art 17.4. A leaked draft of the Trans-Pacific Partnership dated 16 May 2014 showed that a range of proposed minimum terms for works ranging from 50 to 100 years pma. An earlier draft dated 3 August 2013 showed that countries including the US and Australia were proposing a term of 70 years pma protection. See <>.

10 In Europe, film copyrights potentially last even longer still, since they expire 70 years after the death of the last survivor of the director, screenwriter, or composer. See Directive 2006/116/EC of the European Parliament and the Council of 12 December 2006 on the term of protection of copyright and certain related rights [2006] OJ L 372/12, Art 2.

11 For discussion of some jurisdictions that do limit creators’ abilities to trade their rights, see Hugenholtz et al, ‘The recasting of copyright and related rights for the knowledge economy’ (Research Paper No 2012-38, Institute for Information Law, 2012) 165.

12 Ricketson, ‘The Copyright Term’, above n 2, 777.

13 See R Giblin and K Weatherall, ‘If we redesigned copyright from scratch, what might it look like?, this volume.

14 See e.g. Neil Weinstock Netanel, ‘Copyright and a Democratic Civil Society’ (1996) 106 Yale Law Journal 283, 307; Stephen G Breyer, ‘The Uneasy Case for Copyright: A study of copyright in books, photocopies, and computer programs’ (1970) 84 Harvard Law Review 281; William M Landes and Richard A Posner, ‘An Economic Analysis of Copyright Law’ (1989) 18(2) Journal of Legal Studies 325.

15 Anne Barron, ‘Copyright infringement, “free-riding” and the lifeworld’, in Lionel Bently, Jennifer Davis and Jane C Ginsburg (eds), Copyright and Piracy: An Interdisciplinary Critique (Cambridge University Press, 2010) 93, 94.

16 Landes and Posner, above n 14, 327.

17 Economists Brief in Eldred v Ashcroft, 537 US 186, 4–7 (2003).

18 This modelling was conducted by a team of economists (including five Nobel laureates) and is based on the conservative assumption that revenues would remain stable over time. This is unlikely, given that most copyrighted works tend to lose value over time, so in fact the present value of later years is likely to be even lower than the model reflects. See Economists Brief in Eldred v Ashcroft, 537 US 186, 4–5 (2003).

19 See Landes and Posner, above n 14, 326, 363; Economists Brief in Eldred v Ashcroft, above n 17, 7–8. But cf Stan J Liebowitz and Stephen E Margolis, ‘Seventeen Famous Economists Weigh in on Copyright: The Role of Theory, Empirics and Network Effects’ 18(2) (2005) Harvard Journal of Law and Technology 435 (questioning the assumptions in the brief).

20 See, for example, detailed empirical analysis in William M Landes and Richard A Posner, ‘Indefinitely Renewable Copyright’ (2003) 70 University of Chicago Law Review 471, 501–507.

21 See e.g. HM Treasury, Gowers Review of Intellectual Property (December 2006), <>, 52–53; Hugenholtz et al, above n 11, 113. It has been estimated that more than 95 per cent of copyrighted back catalogue recordings are not currently released, further demonstrating their lack of current commercial value. See Hugenholtz et al, at 114.

22 The fees over time are set out in Landes and Posner, above n 20, n 10.

23 See Barbara A Ringer, ‘Renewal of Copyright’, in Copyright Society of the USA (ed) 1 Studies on Copyright: Arthur Fisher Memorial Edition 503, 616–620; Landes and Posner, above n 20, 499–500.

24 William F Patry, How to fix copyright (Oxford University Press, 2011) 106.

25 Landes and Posner, above n 14, 361–362.

26 Rufus Pollock, ‘Forever Minus A Day? Calculating Optimal Copyright Term’, (2009) 6(1) Review of Economic Research on Copyright Issues 35; available at <>, 3.

27 Pollock, above n 26, 21. Boldrin and Levine have also developed an economic model based on the idea that, as the elasticity of total monopoly revenue increases, the scope of copyright should decline with the size of the market, and reached the conclusion that optimal terms are likely seven years or less. See Michele Boldrin and David K Levine, ‘Growth and Intellectual Property’, (Working Paper 12769, National Bureau of Economic Research, December 2006) <>.

28 One study linked recent 20-year term extensions in a number of OECD countries to increases in the number of movies produced. However, the study was never peer-reviewed or published. See IPL Png and Qiu-hong Wang, ‘Copyright Duration and the Supply of Creative Work’ (2006), available at <>. A subsequent non-peer-reviewed paper by the same authors incorporated that research with other work they had conducted, and ‘found no statistically robust evidence that copyright term extension was associated with higher movie production’. See IPL Png and Qiu-hong Wang, ‘Copyright Law and the Supply of Creative Work: Evidence from the Movies’ (2009) <> (unavailable from original source but accessible via; copy on file with editors).

29 See e.g. William Cornish, ‘Intellectual Property’ in (1994) 13 Yearbook of European Law 485, 489–490, explaining that these terms cannot be justified on utilitarian grounds: ‘it cannot be that an extension of the right from fifty to seventy years post mortem auctoris is required as an economic incentive to those who create and those who exploit works. They make their decisions by reference to much shorter time scales than these. It is only considerations of moral entitlement which can possibly justify even the present minimum term in the Berne Convention …’.

30 See e.g. Mark A Lemley, ‘Property, Intellectual Property, and Free Riding’ (2005) 83 Texas Law Review 1031, 1058 (recognising that ‘intellectual property rights interfere with the ability of other creators to work, and therefore create dynamic inefficiencies’).

31 Economists Brief in Eldred v Ashcroft, 537 US 186, 9 (2003).

32 Economists Brief in Eldred v Ashcroft, 537 US 186, 9 (internal note omitted) (2003).

33 Economists Brief in Eldred v Ashcroft, 537 US 186, 9 (2003).

34 Mark A Lemley, ‘Ex ante versus ex post justifications for intellectual property’ (2004) 71 University of Chicago Law Review 129, 133–134, citing Orrin G Hatch and Thomas R Lee, ‘“To Promote the Progress of Science”: The Copyright Clause and Congress’s Power to Extend Copyrights’ (2002) 16 Harvard Journal of Law & Technology 1, 3; Orrin G Hatch, ‘Toward a Principled Approach to Copyright Legislation at the Turn of the Millennium’ (1998) 59 University of Pittsburgh Law Review 719, 736–737; Brief of Amicus Curiae American Intellectual Property Law Association in Support of Respondent, Eldred v Ashcroft, No 01-618, 16–17 (S Ct filed 5 August 2002) (available on Westlaw at 2002 WL 1822117); Brief of Amici Curiae of the Nashville Songwriters Association International (NSAI) in Support of Respondent, Eldred v Ashcroft, No 01-618, 14 (S Ct filed 2 August 2002) (available on Westlaw at 2002 WL 1808587); Robert A Gorman and Jane C Ginsburg, Copyright: Cases and Materials (Foundation, 6th ed, 2002) 347.

35 Copyright Term Extension Act: Hearing on HR 989 Before the Subcommittee On Courts and Intellectual Property of the House Committee on the Judiciary, 104th Congress (Statement of Marybeth Peters, Register of Copyrights and Associate Librarian for Copyright Services, Library of Congress).

36 See Paul J Heald, ‘Property Rights and the Efficient Exploitation of Copyrighted Works: An Empirical Analysis of Public Domain and Copyrighted Fiction Bestsellers’ (2008) 92 Minnesota Law Review 1031, 1035, citing B Zorina Khan, ‘Does Copyright Piracy Pay? The Effects of US International Copyright Laws on the Market for Books, 1790–1920’ (Working Paper No 10271, National Bureau of Econ Research, 2004) 21, 21–24.

37 See Eva E Subotnik and June M Besek, ‘Constitutional Obstacles? Reconsidering Copyright Protection for Pre-1972 Sound Recordings’ (2014) 37 Columbia Journal of Law & the Arts 327, 330–331.

38 Tim Brooks, Survey of Reissues of U.S. Recordings (August 2005) <> v: ‘One consideration by Congress in extending copyright protection to owners for such a long period was to give those owners an incentive to reissue, and thereby preserve, older recordings.’

39 Tim Brooks, Survey of Reissues of U.S. Recordings (August 2005) <> 7–8. Note that some reissuers may have claimed copyright in the ‘remastered’ recordings; other producers claim copyright in the accompanying artwork.

40 Paul J Heald, ‘Property Rights and the Efficient Exploitation of Copyrighted Works: An Empirical Analysis of Public Domain and Copyrighted Fiction Bestsellers’ (2007) 92 Minnesota Law Review 1031, 1040.

41 Ibid.

42 Ibid 1048–1050.

43 Ibid 1043.

44 Ibid 1050.

45 Paul J Heald, ‘How Copyright Keeps Works Disappeared’ (2014) 11(4) Journal of Empirical Legal Studies 840–841.

46 Ibid 842–843.

47 Ibid 852–853.

48 Ian Hargreaves, Digital Opportunity: A Review of Intellectual Property and Growth (Department for Business, Innovation & Skills, 18 May 2011), 38.

49 See e.g. Ellen Franziska Schultze, ‘Orphan works and other orphan material under national, regional and international law: analysis, proposals and solutions’ (2012) European Intellectual Property Review 313, 213. Orphaning is also partly caused by the lack of formalities in copyright: see Dev Gangjee’s fuller discussion of this in ’Copyright formalities: A return to registration?’, this volume.

50 Australian Digital Alliance and Australian Libraries Copyright Committee, Submission No 586 to Australian Law Reform Commission, Copyright & the Digital Economy, November 2012, 52.

51 See Anna Vuopala, ‘Assessment of the Orphan works issue and Costs for Rights Clearance’ (May 2010) European Commission, 5.

52 Hargreaves, above n 48, 38.

53 Ibid.

54 See Susan Corbett, The law – is it an ass? (1 April 2014), Play It Again Project <>.

55 Amicus Brief of Peter Decherney, Golan v Holder, 609 F 3d 1076, 17–18 (US 2012).

56 For a list of 50 Disney films created from stories in the public domain, see Derek Khanna, 50 Disney Movies Based On The Public Domain (2 March 2014) Forbes <>.

57 Kembrew McLeod and Rudolf Kuenzli (eds), Cutting Across Media: Appropriation Art, Interventionist Collage, and Copyright Law (Duke University Press, 2011) 256.

58 This study found that more than twice as many public domain books were available as recordings than protected books (33 per cent compared to 16 per cent). See Christopher Buccafusco and Paul J Heald, ‘Do bad things happen when works enter the public domain? Empirical tests of copyright term extension’ (2013) 28 Berkeley Technology Law Journal 1, 22. Even when only commercial recordings are taken into account, public domain works were still more available to the public than their protected counterparts. See page 29.

59 Canvassing the possible reasons for this intuition, see Stephen Breyer, ‘The Uneasy Case for Copyright: A Study of Copyright in Books, Photocopies, and Computer Programs’ (1970) 84 Harvard Law Review 281, 284–291.

60 Even the US Supreme Court, against a backdrop of constitutionally mandated utilitarianism, has sometimes showed echoes of such thinking in the jurisprudence, once observing that ‘sacrificial days devoted to … creative activities deserve rewards commensurate with the services rendered’. See Mazer v Stein, 347 US 201, at 219 (though note that the Court’s ultimate decision relied heavily on utilitarian justifications).

61 Ricketson, ‘The Copyright Term’, above n 2, 757.

62 Committee Reports, 105th Congress, 2nd Session, House Report 105–452, 105 H Rpt 452; Copyright Term Extension Act <>.

63 See e.g. Peter Jaszi, ‘Toward a Theory of Copyright: The Metamorphoses of “Authorship”’ (1991) 40(2) Duke Law Journal 455, 471.

64 See Copyright Extension Act of 1995, S 483, 104th Congress, 1st Session (1995). The only exception was that the Extension Act did amend the Copyright Act to give authors and their heirs a termination right at the end of the original 75 years from first publication, but only if they hadn’t exercised their prior termination right at the end of 56 years from first publication. See 17 USC § 304, (c) and (d).

65 Rebecca Tushnet, ‘Economies of Desire: Fair use and Marketplace Assumptions’ (2009) 51 William & Mary Law Review 513, 515.

66 Ibid 523.

67 Ibid 524.

68 See e.g. TJ Stiles, ‘Among the digital Luddites’ 38 Columbia Journal of Law & the Arts 293.

69 See e.g. Mark A Lemley, ‘IP in a world without scarcity’ 90 NYU Law Review 460, 492. Regarding the phenomenon of commons-based production, see Yochai Benkler, ‘Coase’s Penguin, or, Linux and the Nature of the Firm’ (2002) 112(3) Yale Law Journal 369; see also Michael J Madison, Brett M Frischmann and Katherine J Strandburg, ‘Constructing Commons in the Cultural Environment’ (2010) 95 Cornell Law Review 657.

70 James Boyle, ‘The Second Enclosure Movement’ (2003) 66-SPG Law & Contemporary Problems 33, 45–46.

71 See e.g. discussion in Ruth Towse, Creativity, Incentive and Reward (Edward Elgar, 2001) at 53–58; see also Throsby and Thompson, 1994 (cited in Towse at 57) (results of an extensive survey study of Australian artists demonstrating that artists often worked in non-arts jobs, but only to the extent necessary to support their artistic occupation).

72 Eric E Johnson, ‘Intellectual Property and the Incentive Fallacy’ (2012) 39 Florida State University Law Review 623, 668–669.

73 See e.g. Towse, above n 71, 58.

74 See e.g. Towse, above n 71, 58; Christopher Buccafusco and Christopher Jon Sprigman, ‘The Creativity Effect’ (2011) 78 University of Chicago Law Review 31.

75 Tushnet, above n 65, 545.

76 Ruth Towse, ‘Copyright and Cultural Policy for the Creative Industries’, in Ove Granstrand (ed), Economics, Law and Intellectual Property (Kluwer, 2003), 427.

77 Towse, above n 76, particularly 429; Tushnet, above n 65, 545.

78 William F Patry, ‘The Copyright Term Extension Act of 1995: Or How Publishers Managed to Steal the Bread from Authors’ (1996) 14 Cardozo Arts & Entertainment Law Journal 661, 675–676.

79 Ibid 665.

80 See e.g. Dionne Searcey and James R Hagerty, ‘Lawyerese Goes Galactic as Contracts Try to Master the Universe’ The Wall Street Journal (online) 29 October 2014, <>; Done Deal Professional, Work for Hire Agreement #1 (31 January 2001) <>.

81 Towse, above n 71, 126.

82 HM Treasury, above n 21, 51.

83 Ricketson, ‘The Copyright Term’, above n 2, 784.

84 See e.g. discussion in Hugenholtz et al, above n 11, 165.

85 Ricketson, ‘The Copyright Term’, above n 2, 757.

86 See e.g. Silke von Lewinski, ‘EC Proposal for a Council Directive Harmonizing the Term of Protection of Copyright and Certain Related Rights’ (1992) 23 International Review of Intellectual Property and Competition Law 785, 788–789.

87 See Giblin and Weatherall, above n 13.

88 Brett M Frischmann and Mark A Lemley, ‘Spillovers’ (2007) 107 Columbia Law Review 257, 258–261.

89 See generally Mark A Lemley, ‘Property, Intellectual Property, and Free Riding’ (2005) 83 Texas Law Review 1031.

90 Ricketson, ‘The Copyright Term’, above n 2, 770–771.

91 See e.g. Landes and Posner, above n 20, 476 (‘[T]he incremental incentive to create new works as a function of a longer term is likely to be very small (given discounting and depreciation) beyond a term of twenty-five years or so’).

92 Niva Elkin-Koren, ‘Can formalities save the public domain? Reconsidering formalities for the 2010s’ (2013) 28(3) Berkeley Technology Law Journal 1537, 1553.

93 Ibid 1555–1556.

94 Ibid 1556.

95 In order to exercise their rights, authors, preservation societies and members of the public would have to have a right to bypass any technological protection measures that had been applied when the works were protected by copyright.

96 Jane Ginsburg authored this catchy descriptor: see Jane C Ginsburg, ‘Fair Use for Free, or Permitted-but-Paid?’ (Working Paper No 481, The Center for Law and Economic Studies Columbia University School of Law, 2 June 2014).

97 Ricketson, ‘The Copyright Term’, above n 2, 761.

98 von Lewinski, above n 86, 789.

99 Ricketson, ‘The Copyright Term’, above n 2, 762–763.

100 This is similar to a French rule which treats book authors as the co-author of their film adaptations. See Article L 113-7 of the Intellectual Property Code (France).

101 See e.g. What About Royalties? A Sound Guide for Musicians (November 2006) Venture Navigator <> (unavailable from original source but accessible via; copy on file with editors) (‘Often session musicians and backing vocalists are asked to sign a standard consent from (drafted by the Musicians Union and Equity respectively). This waives their rights to be paid each time their performance on a recording is used. Instead they get a one-off payment for the session.’); Brecknell, What singers and musicians need to know <> (‘To avoid ambiguity, and in accordance with standard industry practice, my session musicians will be asked to sign a waiver of rights in the creative work.’).

102 See e.g. Ricketson, ‘The Copyright Term’, above n 2, 768–769.

103 Kimberlee Weatherall develops this point in ‘A reimagined approach to copyright enforcement from a regulator’s perspective’, this volume.

104 Email from Heidi Bond to Cyberprofs mailing list, 17 May 2014.

105 See e.g. Stuart Dredge, ‘Billy Bragg: labels not Spotify deserve streaming music payouts scrutiny’, The Guardian (online), 7 November 2013 <>; David Byrne, ‘The internet will suck all creative content out of the world’ The Guardian (online), 12 October 2013 <>; Stuart Dredge, ‘Thom Yorke calls Spotify “the last desperate fart of a dying corpse”’ The Guardian (online) 7 October 2013 <>. Generally, see also Eddie Schwartz, ‘Code: fair trade music: letting the light shine in’ in Susy Frankel and Daniel Gervais (eds), The Evolution and Equilibrium of Copyright in the Digital Age (Cambridge University Press, 2014).

106 The report is published at SNEP, Le Marche De La Musique Enregistree-Bilan 2014 (3 February 2015) <>, see English-language précis at Tim Ingham, Major Labels Keep 73% of Spotify Premium Payouts – Report (3 February 2015) Music Business Worldwide <>.

107 David Falkayn, A Guide to the Life, Times and Works of Victor Hugo (University Press of the Pacific, 2001) 29. The proposal was made to a worldwide audience in a speech opening the 1878 International Literary Congress in Paris. The Congress ultimately passed a resolution recommending the adoption of a domaine public payant. Ricketson, The Berne Convention For the Protection of Literary and Artistic Works: 1886–1986, above n 2, 47.

108 For details of these nations and the outlines of their schemes see generally: United Nations Educational, Scientific and Cultural Organization – World Intellectual Property Organisation, Committee of Non-Governmental Experts on the ‘Domaine Public Payant – Analysis of the Replies to the Survey of Existing Provisions for the Application of the System of ‘Domaine Public Payant’ in National Legislation, UNESCO/WIPO/DPP/CE/I/2 (10 March 1982) <>; Peter Schonning, ‘Survey of the term of protection of authors’ rights: Toward More Universal Protection of Intangible Cultural Property’ (2000; 11(4) Entertainment Law Review 59; Séverine Dusollier, ‘Scoping Study on Copyright and Related Rights and the Public Domain’, Committee on Development and Intellectual Property (CDIP), 4 March 2011, 39–40; Cathryn A Berryman, ‘Toward More Universal Protection of Intangible Cultural Property’ (1993–1994) 1 Journal of Intellectual Property 293. See also Ryszard Markiewicz and Janusz Barta, ‘The new Polish Copyright Act – standards and particularities’ (1995) 26(3) International Review of Intellectual Property and Competition Law 337, 342 (describing the 1994 introduction of a paying public domain into Polish law, which required 5–8 per cent of gross income from the sale of copies of expired works to be paid into a ‘creativity promoting’ fund).

109 That was the reason for the German proposal being derailed in 1965. See Davies, above n 5, 201. For a fuller history of this proposal see Paul Katzenberger, ‘Die Diskussion um das “domaine public payant” in Deutschland’ in Festschrift für Georg Roeber zum 10 Dezember 1981, Schriftenreihe der UFITA – Edition 63, Freiburg 1982, 193–230.

110 Dusollier, above n 108, 40; Carlos Mouchet, ‘Problems of the “Domaine Public Payant”’ (1983–1984) 8(2) Columbia VLA Art & Law 137.

111 Whitford Committee, Report of the Committee to Consider the Law on Copyright and Designs (March 1977) The Whitford Committee Report [41].

112 Buccafusco and Heald, above n 58, 10–11.

113 Landes and Posner, above n 20, 471.

114 Ibid 517–518.

Previous Next