In 2012, a report was published in the online journal LAWS entitled A Counterfactual Impact Analysis of Fair Use Policy on Copyright Related Industries in Singapore. I know. Sounds like a real page-turner for the general reader, right? To be sure, most of us are not schooled in the arcana of statistical economic analysis, but suffice to say the report, written by Roya Ghafele and Benjamin Gibert, concluded (or at least implied) that expansion of the fair use doctrine in the Singapore Copyright Law in 2005 resulted in economic growth in what the researchers identified as the “copying technology industries” (e.g. disk drives, CDs, etc.) and with no detrimental impact to the copyright industries. When the report was published, it was predictably seized upon by Mike Masnick as proof of one of Techdirt’s core tenets, namely that “less copyright is always economically beneficial”.
But an analysis published yesterday by George S. Ford, PhD at the Phoenix Center for Advanced Legal & Economic Public Policy Studies, has called the Singapore Study a work of “stunningly poor quality”. Citing numerous flaws in methodology, he insists that the report’s conclusions should not be considered instructive to copyright law in Singapore or anywhere else. In fact, the reason I qualified the study’s conclusion in the previous paragraph is that apparently Ghafele and Gibert themselves do not claim to identify a causal relationship between Singapore’s revised fair use doctrine and an increase in sales in “copying devices”. Ford argues this flaw alone is sufficient to label the entire study as “worthless to policymakers”. Naturally, one must be careful about taking sides among economists, whose stock and trade is critiquing one another’s methods that the rest of us don’t really understand. Nevertheless, Ford’s critiques ought to at least raise questions among us laymen when he says unequivocally …
“While evidence on fair use policies is welcome and critical to informed policy reform, Ghafele and Gibert’s empirical analysis is so poorly done that it fails to shed any light on copyright laws. Governments reviewing their copyright laws should dismiss the Singapore Study as junk science.”
Ford’s criticisms include Ghafele and Gibert’s failure to employ a proper control group, to account for differences in scale among the technology businesses aggregated into the study, and to exclude from their analysis catalytic factors other than changes to fair use—not the least of which were other 2005 amendments to Singapore’s Copyright Law. As stated, I cannot presume either to critique or defend the computations applied by Ford, but what I can comment on is this aspect of his conclusion:
“… the expanded fair use policy was incorrectly interpreted by consumers as a license to pirate and distribute intellectual property without consequence. Less than a decade after the new fair use policy was implemented, Singapore amended its copyright law to address widespread digital piracy.”
Indeed. Fair use has been so chronically misrepresented in the public dialogue that the principle has been broadly interpreted as the antithesis of copyright, which is simply incorrect. In fact, this issue points to one of the reasons I find the hypothesis of the Singapore Study a bit odd in the first place. An attempt to quantify the extent to which Singapore’s fair use revisions acted as a market catalyst at all seems to treat the doctrine as though it were a universal exception to copyright rather than a narrowly defined, case-by-case, limitation—one that is in fact expected to spawn some portion of new copyrightable works. For instance, if 100 fair uses are made and half are for new works that have their own copyrights, that’s 50 new copyrights supported by fair use doctrine. So, would that be a net win for copyright or for fair use? The question is absurd because fair use is a part of copyright law. If fair use were indeed the opposite of copyright and without reasonable limitations, then it would simply nullify copyright, taking the concept of fair use with it across the event horizon into irrelevance.
More specifically, fair uses in Singapore, which they call fair dealing, are conditional just like they are in the US. The first four of five factors Singaporean courts consider in a fair dealing defense are modeled almost verbatim on the four factors applied in our courts. Like our Copyright Act, Singapore’s fair dealing statutes seek to define the specific conditions under which there are limitations on exclusive rights; and most encouragingly, those statutes appear to have the same intent to protect free speech, which was the original reason we codified fair use in the 1976 law. So, the decision to examine the effect of Singapore’s fair dealing doctrine on the market for devices and media used for data storage seems inscrutably haphazard.
After all, any number of factors may increase the sale of disk drives, recordable CDs and DVDs, etc., including rampant piracy itself, which is outright infringement and not a fair dealing. More acutely, as Ford mentions, 95% of Singapore’s electronics production is exported. Singapore is a tiny market (pop. 5.5 million) whose economy includes a robust wholesale and retail sector shipping to foreign markets and catering to a very large volume of tourist/shoppers from the region. In other words, Australians traveling to Orchard Street to buy hard drives tells us nothing at all about Singapore’s fair dealing statutes in its copyright law. So, if Ford and other critics are right that the Singapore Study does not account for this, that is a considerable flaw in the research.
Additionally, an analysis in the sales of electronic hardware used for copying and storing digital media does not appear to address any of the questions being asked about fair use doctrine in the age of the Internet. If there is merit to that conversation, it would rationally involve—and in fact does involve—a discussion of platform-based uses like YouTube, blogs, or fanfic sites that encourage remix; or we may consider the casual sharing of content via social media; but this line of investigation would seem to consider fair use’s initial intent to protect free speech more than an inquiry into broader economic benefits. I am skeptical that new, platform-based uses—however common and ubiquitous they may be—provide a rationale for “expanding” our own doctrine; but at least these types of uses do represent changes in the nature of how works are used, which is not the case with regard to storage media.
Meanwhile, the general consumer in our market appears to be moving away from a paradigm of storing media at all, thanks to the convenience and low cost of streaming and the availability of cloud-based options in lieu of local devices. (Note the lack of disk drive in your new computer.) So, if sales of certain recordable media were to decline over the next five years, what would that tell us about our fair use doctrine, which has been law for 40 years? Not much, I think. It’s simply an odd metric to examine—a bit like measuring bottled water sales in order to determine how many Americans are going to the gym and then to draw conclusions about our overall cardio-vascular health.
While limitations on copyright’s exclusivity, including fair use, can produce market benefits, it is always necessary to seek a balance. As the market evolves, the contours of fair use may indeed shift, though the more those contours cease to define boundaries at all, the more the doctrine is stripped of its significance. As such, neither professional nor amateur analysis should confuse the policy discussion by asking the wrong questions.