The Canadian Government announced last week that it is amending the Copyright Act to extend the term of protection for performers and makers of sound recordings from its current 50 years to 70 years. This will bring Canada’s laws more into line with those of more than 60 counties which have protection of 70 years or more.
In the week following the government’s Budget announcement, certain critics of copyright term extension have advanced four criticisms of the proposed reform, claiming there will be:
- Heavy costs to consumers in royalty payments.
- Royalty payments sent out of the country.
- No additional incentive for creativity.
- Less creative material entering the public domain.
In a separate report (Common Myths), I review the evidence cited in support of each of these four contentions. In each case, I source the cited authority for each claim in reports circulating during the 2005-2009 U.K. and EU copyright reform debate, and the 2005 Australian debate on copyright term extension. I further identify fundamental errors in the economic analyses relied upon and claims made, and discuss how each claim was rejected by the respective governments.
Myth 1: Heavy Costs to Consumers in Royalty Payments.
It has been claimed that copyright term extension for sound recordings and performances will cost consumers millions of dollars. There is no economic study that actually estimates this in Canada. Furthermore, some critics (with reference to a “U.K. study”) have reported that the public cost associated with copyright term extension in the U.K. was estimated as high as €1 billion. Some have then implied that Canadian consumers might experience a similar order of cost from copyright term extension. There is, however, no such “U.K. study”. In fact, the U.K. Government’s “2006 Gowers Report on Intellectual Property, which commissioned and relied upon a report by the Centre for Intellectual Property and Information Law (CIPIL) at Cambridge University (CIPIL report), actually suggested that “the overall impact of term extension on welfare would be a net loss in present value terms of 7.8 per cent of current revenue, approximately £155 million.” This is a much smaller number than the €1 billion that has been cited by critics, and for reasons I outline below, CIPIL’s conclusion was in any event properly rejected by the U.K. Government.
Critics have also cited a Dutch study on intellectual property, claiming it reached the conclusion that “the extension would create significant costs for consumers and society as a whole”. This reference is to a 2008 article co-authored by a Dutch academic Bernt Hugenholtz, Director of the Amsterdam University Institute for Information Law, (AUIIL). It summarises a study undertaken by AUIIL in 2006 that was commissioned by the European Commission to evaluate the main arguments made in favour of a copyright term extension for sound recordings. What critics do not mention is that the study by AUIIL was rejected by the European Commission in a later full Impact Assessment, conducted by the Commission itself. The AUIIL report relied on similar assumptions as those contained in the U.K. CIPIL report mentioned above for the Gower Inquiry (which was also rejected by the Commission), and also contained no actual empirical evidence of the impact of copyright term extension on prices charged to consumers in Europe.
Thus, despite their characterization by critics, the actual studies cited do not make the claim that copyright term extension would cost the public in excess of €1 billion, The only two actual studies at the base of these claims were prepared by the CIPIL (2006) and AUIIL (2006), and were, in any event, rejected by the U.K. and EU Governments in 2008-2009. Neither of these studies estimated the costs of the European copyright term extension for sound recordings as €1 billion. At the risk of spreading misinformation, however, it turns out the €1 billion figure actually originates from a later report issued by a number of Law Centres (including CIPIL and AUIIL) (referred to as CIPIL et al. 2008). This joint report was CIPIL et al.’s response to the European Commission’s rejection of CIPIL’s and AUIIL’s original analyses.
As I show in my main report (linked to above) in more detail, the fundamental problems with these studies being relied on by term extension critics (i.e. the CIPIL, AUIIL and CIPIL et al. reports) are that they:
1. Ignore Digital Piracy: The reports ignore the historical context, which is the growth of digital piracy, and the fact most studies in top tier economics journals support a conclusion that digital piracy has been responsible for the decline in sound recording sales (70% in real terms) over the past 15 years. It is clear that lengthening the copyright term for sound recordings and performances is a rational and efficient policy response to the increased ease of pirating sound recordings, in that it will help to restore profitability and properly incentivise investment in the music industry.
2. Ignore The Free Rider Problem: The reports ignore the costs of downstream free riding, where public domain users try to avoid the cost of creating their own works by exploiting out-of-copyright creative works that have passed their currently limited copyright term. Thus, the CIPIL report tends to emphasize how copyright may entail costs for downstream creativity, but the CIPIL report nowhere mentions the free rider problem, where free riding by downstream public domain record labels may reduce creativity.
3. Assume the Creation of a Copyright Monopoly: One of the key (and incorrect) assumptions made by CIPIL and others critical of copyright term extension, however, is that copyright creates a monopoly. Instead, of course, copyright creates a property right, which, in the context of the music industry, supports a competitive market. The assumption that copyright creates a monopoly is clearly incorrect, as copyright only provides thin protection for the creative expression of an idea, and not the underlying idea, allowing competition from close substitutes (e.g., similar music).
4. Assume Deadweight Costs: A further problem with the CIPIL study is its assumption that for copyright goods that already exist, there is no benefit from a retrospective term extension, only deadweight costs. Generally speaking, however, property rights over other existing goods (e.g. land) last forever – or in perpetuity – and are seen to be beneficial – so why not for copyright, which covers existing creative goods? Copyright under the common law also lasted in perpetuity, at least until the right was taken away by legislation. CIPIL et al. claim that while normal or scarce goods like land can become congested if they are over-used, copyright goods are different – they are not rival, nor scarce. It is claimed many people can use copyright goods without congestion cost, so there is no need for conservation or management of access to copyright goods. CIPIL et al. rely on Landes and Posner (1989) [1] for this assumption.
Landes and Posner (2002) [2], however, have since rejected the notion, recognizing instead that there are costs to over-use of valuable copyright works. Thus, they note:
If …anyone were free to incorporate the Mickey Mouse character in a book, movie, song, etc., the value of the character might plummet. Not only would the public rapidly tire of Mickey Mouse, but his image would be blurred, as some authors portrayed him as a Casanova, others as cat meat, others as an animal-rights advocate, still others as the henpecked husband of Minnie. … until Mickey Mouse’s commercial value was zero.
Thus, the CIPIL report’s analysis is based on a fundamentally wrong and since rejected premise, and consequentially its recommendations are flawed.
5. Make Modeling Errors: There are basic errors in the formulae used in the various reports critical of copyright term extension for sound recordings. Thus, in the CIPIL report’s theoretical model, one inaccuracy alone led CIPIL to overestimate the overall welfare loss by 44% [3]. Even after these errors were corrected, however, the model was still sensitive to changes in assumptions about a number of parameters, which were largely “guestimates”, and not based on any empirical evidence. Thus, even minor changes to parameters generated benefits to consumers from copyright term extension – the exact opposite conclusion reached by CIPIL. As outlined in my full report (Common Myths), extreme assumptions and arithmetic errors were also present in the CIPIL et al. report calculating the €1 billion cost.
6. Ignore The Facts – For example on what happens to consumer prices: none of the Studies relied on by critics of copyright term extension for sound recordings have collected any data to test their analyses. By comparison, PricewaterhouseCoopers (PwC) looked at 129 albums recorded between 1950 and 1958, and found no clear evidence that records in which the related rights have expired were sold systematically at lower prices than records which are still protected. Thus, the European Commission’s full Impact Assessment (IA) of the proposed copyright term extension concluded: “A term extension would have no negative impacts on consumer prices and would have a positive impact on the quality of services offered to consumers as well as on consumer choice.” (Impact Assessment, p. 36)
As I discuss in my full paper (Common Myths), the studies being relied on by Canadian critics have adopted the most extreme possible assumptions and have made basic errors including faulty arithmetic. The studies were ultimately rejected by the U.K. and EU governments. Having been exposed for making basic errors and rejected by the very governments that commissioned their work, the authors of the rejected studies then orchestrated a public campaign, involving various open letters, statements, conferences and articles, that railed against the governments’ decisions and propagated their misleading analyses. We are still hearing the echo from this today.
Myth 2: The Royalty Payments Will Be Sent Out of the Country.
A further myth or exaggerated claim is that not only will consumers pay a heavy price in greater costs for sound recordings, but the amounts paid by consumers will be sent out of the country. For this it has been claimed that “one scholar estimated that the copyright extension resulted in Australians sending an extra $88 million per year in royalties overseas” (the “Dee report”) [4]. Again, however, the study was speculative. It merely extrapolated from the fact that Australia was a net importer of copyright in 2002-2003, and assumed that this deficit would increase by an arbitrary amount, leading to an assertion that extending copyright term limits for sound recordings would push up these net payments by $88 million per annum. No analysis was done of what actually happened to net payments from Australia after the reform.
The extension of copyright term limits would have increased revenues for Australian creators, both in Australia – which would not figure in the balance of payments – and overseas. Thus, Australian creators were not entitled to an extended copyright term in other countries unless and until Australia extended its laws. The paper also failed to explore the distortion that existed between Australia’s relatively low term of protection and the higher term in other countries, and the effect this had on net payments from Australia in 2003. In all likelihood, the fact that the Australian copyright term was lower than that in the US probably exacerbated the balance of payments deficits, and its alignment to the US restored better balance.
Myth 3: No Additional Incentive for Creativity.
The third economic myth is that artists are not incentivized by longer copyright terms to create new works (or in this case, sound recordings). It has been claimed that “[l]ong copyright terms are a poor recipe for compensating creators, who generally receive low royalties from their works.” On the contrary, one of the only empirical studies based on a unique dataset, conducted by Stan Liebowitz in 2006 [5] , indicates that the proposed change in the sound recording copyright term is likely to have considerably larger financial implications than has previously been assumed. It was noted that by increasing the copyright term of sound recordings from 50 to 95 years would likely increase nominal revenues by almost 70%. Discounting these future nominal revenues, as is proper, lowers their present value to a range from 3% to 10%, depending on the discount rate chosen.[6] The discounted value is of sufficient size to allow an economically consequential increase in the production of new works.
Critics have raised the argument that University of Montreal economist Abraham Hollander found that the economic value of a copyright term extension to the recording industry was very small. The Hollander study notes:
[Sound recordings] are protected for a period of 50 years from fixation. Adding 20 years of protection would contribute 2.3% to the present value of royalties under a 7% discount rate, assuming that the flow of royalties remains unchanged during the whole period.
As noted by Liebowitz, however, this present value of 2.3% (while it may seem small) can provide considerable incentive for new works and creativity. Moreover, the 7% discount rate is probably high, given that the music industry can spread its risk through large portfolios of sound recordings. A lower discount rate would imply a higher NPV. So despite the arguments from term extension critics to the contrary, Hollander’s study in fact supports the likelihood of term extension increasing creativity.
A further problem with the work cited by critics of copyright term extension for sound recordings is that it has failed to recognize that music companies use current income to invest in new artists, and why this is rational behaviour. Record companies tend to be better judges of returns on investment in current music than outside investors. This means internal financing out of current income, rather than external financing, is at a lower cost. Thus, increased revenues from older works under an extended copyright term will be used by music companies to expand investment into new sound recordings and in the development of new and existing artists. This is confirmed by the way in which record companies actually act in practice. Thus, an analysis of Artist and Repertoire (‘A&R’) spending to find and develop new recording artists of each of the major record companies, showed the data was consistent with the hypothesis that record companies finance new material from the sales of existing material. Therefore, a retrospective increase in the copyright term is likely to have a positive effect on the level of recorded music being created. This analysis was consistent with a survey by the BPI of independent record company members in 2007, which showed that independent record companies rely heavily on internal funds for financing.
The clear conclusion is that, contrary to the many assumptions made by its critics, copyright term extension for sound recordings and performances will both increase the incentive, and the capacity to invest in new sound recordings.
Myth 4: Less Entering the Public Domain.
Finally it is claimed that term extension will simply leave Canadians with 20 additional years of no new works entering the public domain. Despite this myth, the exact opposite is true. Ultimately there will be more sound recordings created, and therefore more supply into the public domain, with a longer copyright term. Also, whether works go to the public domain or not is a choice that resides with artists. After all, it is possible to put work into the public domain under copyright. The problem, however, is that under a public domain law, it is not possible to bring works into the private domain to protect investment and maintain quality.
Conclusion
As summarized above, in a separate more detailed report I have examined and debunked four myths about the likely economic consequences of copyright term extension following the Canadian government’s recent Budget announcement on April 21, 2015. In each case, economic theory and data is shown not to support the claims made. These common mistakes in economic analysis have, however, been revived from the 2005-2009 U.K. and EU debate, and the 2005 Australian debate on copyright term extension for sound recordings. In each of those cases, these arguments were considered and rejected by policy makers in deciding to extend the term of copyright. The same should be true for Canada. Term extension is an efficient way to restore returns to investment in creativity, which was devastated by the growth of digital piracy from 2000. By increasing revenues, and therefore the incentive to invest in creativity, copyright term extension will help increase the supply of new creative goods, enhance consumer choice, competition, and quality, and lower prices in the long run. It will also help to enhance incentives to invest in, and market existing creative goods, and to maintain and enhance their quality, safeguarding our cultural past and musical legacy, while enriching both the present, and the future.
Dr. George Barker is Director of the Centre for Law and Economics at the Australian National University, and past President of the Australian Law and Economics Association. He was awarded the Olin Fellowship in Law and Economics at Cornell University in 2000, was Visiting Fellow at Oxford University Law School 2008, and is currently a visiting Fellow at the British Institute of International and Comparative Law London and Centre for Law and Economics and Society, University College London.
[1] Richard A. Posner & William M. Landes, “An Economic Analysis of Copyright Law,” 18 Journal of Legal Studies 325 (1989), p. 362.
[2] Richard A. Posner & William M. Landes, “Indefinitely Renewable Copyright” (John M. Olin Program in Law and Economics Working Paper No. 154, 2002), p. 12-13.
[3] In its response to a critique by LECG, CIPIL acknowledged that it failed to use the relevant full formulae in its calculations, and instead used an approximation, and that this led them to overestimate the welfare loss from term extension by 44%. With the full formulae their estimate of overall welfare loss falls to 5.4% of present revenues rather than 7.8% of present revenues see CIPIL (2007) “LECG Critique of the CIPIL Review: Response”, Cambridge: Centre for Intellectual Property and Information law p8. http://www.cipil.law.cam.ac.uk/policy_documents/response_lecg(1).pdf
[4] See the “Dee report”, available at: https://www.eff.org/deeplinks/2012/08/all-nations-lose-tpps-expansion-copyright-terms.
[5] Liebowitz (February 2006), “What are the Consequences of the EU extending Copyright Length for Sound Recordings, report prepared for IFPI.
[6] As Liebowitz notes “Much of this gain from an extra 45 years comes from the early years, due to present value discounting. For example, using a 5% interest rate, more than 70% of the gain is derived from the first 20 additional years.” Liebowitz (2006) p17 footnote 11