Delivered in Parliament on 13 September 2021
Mr Speaker, before I continue, I wish to declare that I am a research-oriented academic that both utilizes as well as generates intellectual property (which, admittedly, I often proceed to sign away), and that my family had previously inherited a number of musical copyrights from my late father, which we have since allowed to lapse.
I will begin by stating my support for the Bill. However, I will point out a number of residual concerns I have, notably on the duration of copyrights, copyrights associated with sound recordings, default ownership of copyrighted works, and fair use provisions. I will conclude by taking a step back, and asking the broader question of whether intellectual property (IP) law has fulfilled its intended purposes.
How long is too long?
Clauses 114, 122, and 125 introduce a number of revisions to copyright duration. To the good, it streamlines the expiry of copyrights, to a common 70 years starting either posthumously, or from the date of creation or publication, depending on the nature of the work. Such harmonization and simplification are certainly welcome.
Inherent in this simplification, however, are a number of implicit choices that, to my mind, raise some questions. Limiting the duration of copyright for anonymous works to 70 years is certainly better than perpetuity, which was the case before. But—setting aside minimum terms stipulated in international conventions—why would such a long duration of protection be necessary in the first place? With no identified author, who would be the beneficiaries of a 70-year copyright, and more importantly, would conferring such protection generate net benefits for society, rather than releasing its use to the public domain? I shall return to this more general point—about the benefits and costs of granting intellectual monopolies—at the end of my speech.
Similarly, it is worth observing that the decision to limit copyright protection to 70 years following the creator’s passing—as opposed to the date of publication, as was the case previously—amounts to an extension of the period of protection. Again, while I understand why such consistency between clauses may be attractive, we should be aware that this inhibits the free usage of such works for the purposes of teaching and research, which yields benefits of their own, instead extending additional rent to copyright beneficiaries.
Before I move on, I will just add that while the comparatively longer copyright duration of 70 years is certainly the norm in most high-income countries, this is by no means unanimous. Canada, Hong Kong, and New Zealand have also chosen the lower threshold of life-plus-fifty years instead. And it is worth noting that the
Before I move on, I will just add that while the comparatively longer copyright duration of 70 years is certainly the norm in most high-income countries, this is by no means unanimous. Canada, Hong Kong, and New Zealand have also chosen the lower threshold of life-plus-fifty years instead. And it is worth noting that the minimum term in the Universal Copyright Convention is a mere 25 years, while Berne Convention stipulations are either 25 or 50 years, depending on the nature of the work. Will the Ministry share how the 70-year duration was chosen?
Ministry of sound
Clause 121(b) of the Bill provides for copyright in sound recordings, and were included to respect Singapore’s obligations under the EU-Singapore FTA. However, I should point out that there are consequences to both affording copyright for sound recordings, as well as taking EU intellectual property standards as a benchmark.
One concern with sound recordings has to do with a practice that is a foundation of modern hip hop music: sampling. These apply, especially, to drum breaks and modified soundscapes. There is now a large body of precedent cases that have resolved on either side of infringement or fair use. But the point I wish to make here is more general: by admitting sound recordings into copyright protection, we run the risk of shutting down whole genres of emergent creative output. The pertinent question, then, is whether the fair use provisions outlined in Part 5 are sufficient to preclude such overzealous applications. Has the Ministry considered how such potential abuse may be managed?
Moreover, strict adherence to EU standards for intellectual property may be a slippery slope. Members of this House will well be aware of the stringency of EU geographical indications and traditional specialties; it is the reason why we cannot, even if we had the skills, knowhow, and wherewithal, to produce camembert or scotch whisky. Taken to the extreme, even suggestive terms such as parmesan-style cheese would be prohibited.
Now, while these protections almost always apply to agricultural products—and hence are largely only of academic interest in the Singapore context—we should not be so dismissive of unintended concessions. Singapore has recently embraced artisan beer and spirit manufacture, and there is a risk is that designations such as whiskey (with an “e”), ouzo, or grappa cannot be applied to the work of our craft producers.
Who owns what and why
Division 8 strikes me as self-conflicted. On one hand, Clause 135 grants certain authors automatic ownership of copyright over commissioned works, thereby strengthening the legal standing of creators over their intellectual property. But on the other hand, Clause 134 defaults ownership of copyright for sound recordings and films to employers, so long as the work was produced over the course of employment.
One could argue that Clause 134 is simply an extension of the current position, to cover other types of works. And one could further contend that these are two distinct circumstances; in the former, the author is a contractor performing a one-off service—and hence should retain the rights to their creations—while in the latter, the employee is engaged in a regular employment contract, and hence is being fully compensated for their energies in producing works on behalf of their employer.
But this is an artificial distinction. In both cases, there is payment for services rendered. An employment contract is simply a series of repeated one-off commissions, and payment for regular employment is paid in lump-sum form during payday, as opposed to at the end of each commission. Hence, these are merely operational distinctions. Whether the rights to a creation resides with the author or the employer should therefore be a matter of private negotiation between the two parties; there is no reason why the law should necessarily confer copyrights to one party versus the other. To see this, it is simply sufficient to observe, for example, that photographers routinely offer package deals where the client may either purchase all images (and their associated rights), or they may pay a nominal amount for a set of printed images, without the rights to the images or their soft copies.
As a result, the two sets of clauses come across as self-contradictory, at least from a logical point of view. In my view, it is better for the default to vest the rights to the creator, and have any rights transfer to the employer—along with associated compensation—be explicitly drafted into the employment contract. This preserves the transparency of the process, while also ensuring that the weaker party in the employment relationship—typically the employee—holds the initial bargaining power for their intellectual creations.
Fair is fair
As a routine consumer and producer of intellectual property, I regard Part 5—governing fair use provisions—as central to the business of ideas. This is truly where the rubber hits the road: permitting usage of IP is crucial not just from an educational perspective, which is dear to my heart as a teacher. But—as any researcher will tell you—the entire intellectual edifice stands on the shoulders of earlier discoveries and developments. Set against this is the need to avoid abuse—one only need to be reminded of the American painter and photographer Richard Prince’s notorious appropriation art4—and the risks that an excessively permissive regime may discourage innovation altogether.
That said, the limitations defined in the Bill, as it stands, remain unclear. Clause 194(1)(b) allows for a “reasonable portion” of a work to be copied, which Clause 197(2)(c), for instance, specifies as 5 percent or 5 pages (depending on the total length) of a literary or dramatic work. But 5 pages may simultaneously be too much and too little. 5 percent of a volume of collected works or 5 pages of an integrative scientific periodical such as Science or Nature may well encompass entire articles, while 5 pages of an involved mathematical article may not even include all the pages of a single proof.
And while I appreciate the many exemptions the Bill makes for learning purposes in educational institutions, the text is still unclear if this applies uniformly to research exercises as well, and whether such use would remain acceptable in a non-educational (but knowledge-based) nonprofit, such as a privately-registered think tank or research institute. Similarly, school-based plays often do charge to cover overheads and possibly raise funds, but are nevertheless nonprofits in the sense that student actors and backstage crew are unpaid amateurs.
Fair use provisions also raise an important meta-question that has plagued academic publishing. Members in this House may be aware that a number of international academic publishing houses have been accused of exploiting free labor from civic-minded academics and subsequently turning around and burdening libraries with expensive subscriptions. It took the University of California—the largest public university system in the world—two whole years of bruising contract negotiations to finally arrive at a deal that would allow its faculty and students open access to Elsevier journals. The European Union, in an effort to prevent paying double for publicly-funded knowledge—once to fund the research, another to access it behind a copyrighted paywall—has even gone as far as to mandate that the output of any project that benefits from public funding must remain freely accessible. All these examples underscore the potential for copyright to be abused and circumvented, even when fair use provisions are carefully crafted. Such rent-seeking undermine the scholarly enterprise, which rely on the free exchange of information. In my view, it is generally better to err on the side of permissiveness, insofar as educational endeavors are concerned. Has the government has considered advancing similar legislation here?
Against intellectual monopoly
Allow me to close with a somewhat more philosophical reflection, on the entire presumption that intellectual property rights are a necessary extension of physical property rights. The conventional case claims that, since the transmission of ideas is free, we must therefore confer them protective rights, because the failure to do so would lead to a collapse in incentives necessary to support innovation and invention.
Mr Speaker, it is a fallacy that just because ideas are costless to transmit, they are consequently costless to replicate. Ideas are ultimately embedded in knowledge, which—as anyone in this House who has slaved over textbooks will know—must be acquired at a cost. By the same token, the ability to develop and embed these selfsame ideas into profitable goods and services—the “development” part of R&D—requires a nontrivial (and certainly non-free) amount of human capital to effect the required transformation, and a similar costly process of embedding said knowledge into economically-valuable physical capital. Put another way, just because the technical specifications for, say, the latest microchip may be freely available, it doesn’t meant that you or I would be able to launch the next Intel or Motorola; we still need to hire engineers—who are indisputably not going to work for free—to be able to realize the final product (and there’s no guarantee they’ll do a good job, either).
Indeed, economic research has shown that innovation could still occur under competitive conditions, in the absence of IP rights. This would occur if technological improvements enable the quick recovery of sunk costs from the market—for example, when digital-rights management tools can be applied to restrict unabashed replication of books and music, even without the threat of legal force—or when a growing economy reduces the need for intellectual property protection, because growth itself can generate short-term rents that more than compensate for the opportunity costs incurred in innovation.
The case does, admittedly, become more complicated when we extend our mental model to include not just a home country, but also a foreign one. For instance, protecting foreign IP may even turn out to encourage domestic innovation. However, while targeting IP at goods produced in poorer countries may well raise the innovation rate, the converse would result if IP were directed targeted at rich-country goods.
Far be it that I should be perceived as suggesting that we go ahead and eliminate intellectual property rights and withdraw from the World Intellectual Property Organization. Such a move would be, to say the least, extreme. The Worker’s Party supports the recognition of intellectual property in the work of authors and performers. Rather, my contention is simply that—as much as intellectual monopolies may provide incentives for innovation—they may also encourage socially-costly rent-seeking behavior, and we should be aware of this inherent tradeoff. In this speech, I have provided a number of examples of potential abuse and overreach.
In my view, the only way such overreach may be contained is to ensure that those called upon to resolve contending claims remain sensitive to these tradeoffs. To this end, a deferral to the courts, on a case-by-case basis (as provided by in this Bill), of what constitutes fair use may well represent a sound way forward. However, I should point out that this is only the case if there are mechanisms in place to ensure that those adjudicating these cases possess the capacity to evaluate potentially highly technical presentations. The risk, so amply demonstrated by the U.S. patenting system, is that highly paid attorneys and technical specialists run rings around time- and knowledge-constrained public officers. To this end, how would the Ministry ensure that there is sufficient technical capability among judges to evaluate the merits of fair use provisions in cases of this nature?
Members of this House that listen to Taylor Swift’s music may be aware that she recently sought to re-record many of her earlier songs, initially released under the Big Machine Records label, which still holds the copyright to those original recordings. But the need for TayTay to re-record her own music was to recover control over what she believed to have been a blatant abuse by the record company of the extant IP regime. This episode is a reminder to us that, at some point, even the biggest megastars were once obscure, and risk exploitation by unscrupulous industry insiders. The onus is on us, as legislators and policymakers, to try to get the balance between the two right.