The (Non)Finality of a Fair Use Opinion

Today marks the beginning of Fair Use Week, a celebration of the doctrine of fair use in copyright law. Fair use allows a judge to decide – using a set of four factors articulated by the Copyright Act to guide the analysis – that a person can use another’s copyrighted work without permission or payment, despite the copyright holder’s normal ability to control the use.

If you chat about fair use these days, the word that gets tossed around more than any other is “transformativeness.” The word scarcely exists outside of copyright, but has overtaken discussions of fair use, and in particular the “first factor” of the statutory four-factor analysis, following a landmark law review article by Judge Pierre Leval from 25 years ago, where he wrote:

I believe the answer to the question of justification [in fair use] turns primarily on whether, and to what extent, the challenged use is transformative. The use must be productive and must employ the quoted matter in a different manner or for a different purpose from the original. A quotation of copyrighted material that merely repackages or republishes the original is unlikely to pass the test; in Justice Story’s words, it would merely “supersede the objects” of the original. If, on the other hand, the secondary use adds value to the original – if the quoted matter is used as raw material, transformed in the creation of new information, new aesthetics, new insights and understandings – this is the very type of activity that the fair use doctrine intends to protect for the enrichment of society.

(Citations omitted, hyperlink added.) The Supreme Court adopted Judge Leval’s approach in the 1994 case Campbell v. Acuff-Rose, which remains the last word from the Court on the substance of fair use. It took some time to take root, but today courts seem especially enamored by the “transformativeness” heuristic when deciding fair use cases. Prof. Neal Netanel published an article a few years ago that analyzed current fair use caselaw, built upon three earlier studies of fair use opinions from Profs. Barton Beebe, Pamela Samuelson, and Matthew Sag. In his article, Netanel notes an earlier period where courts focused predominantly on the “fourth factor” of the four-factor test, inspired by an article by Prof. Wendy Gordon and best represented by the Supreme Court’s Harper & Row decision, but that this approach has now been largely replaced:

[F]air use doctrine today is overwhelmingly dominated by the Leval–Campbell transformative use doctrine. … [J]udicial adoption of the transformative use paradigm increased measurably during 2006–2010, even if it was already high previous to that period. During 2006–2010, 85.5% of district court opinions and 93.75%, or all but one, of appellate opinions [that considered fair use] considered whether the defendant’s use was transformative.

He later went on to add that, “in contrast to the Harper & Row regime in which the fourth factor was undoubtedly the most important, today it is largely the first factor, particularly whether the use is held to be transformative, that drives fair use analysis.” Netanel notes that of the cases he surveyed, when a use was found to be transformative, the defendant won 89% of the time from 1995–2000, and every time from 2001–2010.

The dramatic rise of “transformativeness” has lead to concern amongst some scholars (and judges) that this consideration has improperly reduced the traditional multi-factor balancing of fair use to a simple yes/no test. If one is being true to the doctrine, not all fair uses should have to be transformative, and not all transformative uses should be fair. And so there’s a certain hunger amongst scholars to find contemporary examples of nontransformative fair uses. (Precisely what counts as a “transformative” use is a whole other discussion; see Anthony Reese’s article on that.)

The Second Circuit had such a case last January in Swatch Management v. Bloomberg. The facts of the case are straightforward: the Swiss watch and jewelry company Swatch held a conference call with about a hundred investor financial analysts to discuss its annual earnings report. The conversation was recorded at Swatch’s behest, and shortly after the call was complete the news company Bloomberg obtained a copy of the recording and posted it on the website for its “Bloomberg Professional” service, so that potential investors could hear how Swatch was describing its own financial situation. Swatch sued Bloomberg for infringement of its sound recording.

The Southern District of New York found this to be a fair use, but not a transformative one. In its short discussion of transformativeness, the court noted that the presence or absence of a transformative use is not dispositive, and despite a lack of transformation Bloomberg’s use was still favored under the first factor because it is purpose. Bloomberg “advanced the public interest of furthering full, prompt and accurate dissemination of business and financial news.” The court noted that Swatch sought additional discovery as to whether the use was transformative, but it rejected this request, stating “any such issues of fact are irrelevant to my analysis of Defendant’s use for I accepted that Defendant’s use was not transformative.”

On January 27, 2014, the Second Circuit issued an opinion agreeing with the district court. The court held that Bloomberg clearly intended to communicate this to the investing public, that securities law strongly encourages this communication (and requires it of American companies), and that “this important public purpose underlying Bloomberg’s use overcomes the countervailing weight we would otherwise give to Bloomberg’s clandestine methods and the commercial, nontransformative nature of its use.” The court went on to argue that a nontransformative use was, perhaps, preferred when considering the use of a work in a news reporting context:

In the context of news reporting and analogous activities … the need to convey information to the public accurately may in some instances make it desirable and consonant with copyright law for a defendant to faithfully reproduce an original work rather than transform it. In such cases, courts often find transformation by emphasizing the altered purpose or context of the work, as evidenced by surrounding commentary or criticism. [Citing two cases where uses of marginally-modified works were found fair.] Here, Bloomberg provided no additional commentary or analysis of Swatch Group’s earnings call. But by disseminating not just a written transcript or article but an actual sound recording, Bloomberg was able to convey with precision not only what Swatch Group’s executives said, but also how they said it. This latter type of information may be just as valuable to investors and analysts as the former, since a speaker’s demeanor, tone, and cadence can often elucidate his or her true beliefs far beyond what a stale transcript or summary can show. … [I]n light of the independent informational value inherent in a faithful recording of the earnings call, the fact that Bloomberg did not transform Swatch’s work through additional commentary or analysis does not preclude a finding that the [first factor] favors fair use.

So while there was social value added by the use and the use was ultimately found fair, in the eyes of the Second Circuit the use was not transformative.

Or at least the court thought that if you read the opinion in the winter of 2014. If you waited and read the opinion that summer, you would find that the use was held to be both fair and transformative. The court amended its opinion in late May, and changed its analysis of this factor substantially. Many small and large edits to this section shifted it towards this finding. For example, the court previously defended nontransformative uses in news reporting by saying “the need to convey information to the public accurately may in some instances make it desirable and consonant with copyright law for a defendant to faithfully reproduce an original work rather than transform it.” That sentence now ends “…to faithfully reproduce an original work without alteration,” suggesting that the lack of alteration can still transform a work in the fair use sense of the term. Whole new paragraphs to the opinion are added, including:

Furthermore, a secondary work “can be transformative in function or purpose without altering or actually adding to the original work.” [Citing two additional cases where such a use was found that were not in the original opinion.] Here, notwithstanding that the data disseminated by Bloomberg was identical to what Swatch Group had disseminated, the two works had different messages and purposes. To begin with, while Swatch Group purported to convey true answers to the analysts’ questions and to justify the propriety and reliability of its published earnings statement, Bloomberg made no representation one way or another as to whether the answers given by Swatch Group executives were true or reliable. Nor did Bloomberg purport to support the propriety or reliability of Swatch Group’s earnings statement. Bloomberg was simply revealing the newsworthy information of what Swatch Group executives had said. Bloomberg’s message— “This is what they said”—is a very different message from Swatch Group’s— “This is what you should believe.”

Analytically, I think the amended opinion is in better harmony with transformativeness doctrine to the extent that it bases its decision on the fact that physical alternation is not necessary to find transformation. There were times in the original opinion where a reader could get the impression that “transform” was used strictly in a physical sense, instead of the transformation of purpose contemplated by Leval and Campbell. But even under that proper framing it is a bit of stretch to say that Bloomberg’s use actually recontextualized Swatch’s recording in a notable way, when Bloomberg does not seem to do more than present Swatch’s recording without challenge or context. The court is absolutely right to point out that providing such primary sources is an important part of how the public consumes news today, but I can’t help but wonder whether the Second Circuit had the better approach the first time: the use doesn’t transform the work much, but it still has great public value, and on balance it should be allowed. (I find all three Swatch opinions to have the better side of the argument than their counterpart case from the Ninth Circuit, Monge v. Maya Magazines, which rejected a broader “public interest” consideration in fair use.)

What accounts for the Second Circuit’s change of heart? It’s not clear. The public docket of the case doesn’t shed any light, though there are some missing document numbers between the January judgment and the May amended opinion. Perhaps in those missing entries there was a motion for reconsideration or en banc petition that prompted a change, but searches online find no mention of it, and its undisclosed existence would raise more questions than it would answer. There’s no dissent or concurring opinion that suggests a January strife between the judges that is resolved by May. A quick search reveals no discussion in the media about the change – in fact, most of the posts about the case online were published before the amended opinion.

There’s a deeper question here, and that is the change itself. Fair use is an area of law that is especially dependent upon the dozen-or-so cases that come out each year further defining and reforming the doctrine to new types of uses and new sets of facts. Lawyers advising clients on fair use matters are put at a tremendous disadvantage if their building blocks are made out of shifting material. This is compounded by the fact that revisions in opinions are often released without any fanfare or news coverage at all. I only learned of this change thanks to David Hansen at UNC, who discovered it a little over a week after it happened and shared his finding on a copyright listserv. Though I’m not sure if this is where he found it, the Second Circuit decision in Authors Guild v. HathiTrust came out that day, and cited to the freshly-revised Swatch opinion – ironically, in a section where the court found a nontransformative fair use.

The surprising impermanence of judicial opinions received very little discussion until last year, when Prof. Richard Lazarus released a draft of his article The (Non)Finality of Supreme Court Opinions, and Adam Liptak covered it in the New York Times. In his article, Lazarus goes into considerable depth about the history of the Supreme Court and its reporter of authority, the United States Reports, from the early days where decisions were declared orally and documented by individual reporters working with the Justices (such as Henry Wheaton and Richard Peters, of the first Supreme Court copyright case Wheaton v. Peters, a case that Peters both won and reported) to the modern, elaborate system of publication and revision we have today.

Lazarus notes innumerable examples where changes to Supreme Court opinions were made – the overwhelming of which are small, but with some notable exceptions – and how many of these changes occur without any observable record-keeping or public scrutiny. Some of the examples he cites are quite substantive, although few are as long as the numerous new paragraphs in Swatch. And as Lazarus notes, the long-term effects of this system of revision can be severe:

[S]erious practical problems arise when the version of the Court’s opinion upon which lower courts, other branches of government, and scholars and teachers rely can change, without notice, as many as five years after initial publication. Not only do those relying on the Court’s opinions not know of the need to correct their own work, let alone have any practical way to discover the changes made, but their own writings — whether a judicial opinion, casebook, or treatise — can unwittingly perpetuate the error, long after the Court itself has changed its opinion.

The damage in Swatch appears to be contained. The original decision was cited in two cases before the amended report was issued, in the Second Circuit’s Wadsworth v. Allied Professionals Insurance Co. and the District of Connecticut’s Garcia v. Hebert, and both cite it for unrelated points. And Prof. Lazarus’s work may help usher in greater transparency for this often-overlooked part of the development of a judicial opinion. It’s already inspiring tools to help detect these changes, and there’s signs that the Supreme Court may be modifying its behavior.

But the Swatch story serves as an important reminder for lawyers that work in areas of law, like copyright, where opinions get considerable public attention and discussion the week they’re released. Our attitudes toward the outcomes in these cases may form in January, but before citing the case in May, it’s probably a good idea to make sure the opinion still says what you think it does.

This entry was posted in Uncategorized. Bookmark the permalink.