[Sam Bayard, one of our friends down the street at the Citizen Media Law Project, has written the most detailed analysis I've seen of the Barclays v. TheFlyOnTheWall.com case. While focused on the work of financial analysts, the case could have serious impact on the ability of websites to aggregate and curate content. It also invokes the "hot news" doctrine that some news organizations have argued limits the kinds of linking other sites can do to their content. We're reprinting Sam's piece below; it's worth a read for anyone interested in how the new news ecosystem is evolving. —Josh]
In 2003, prolific legal scholar and 7th Circuit Judge Richard Posner published a law review article entitled "Misappropriation: A Dirge," which discussed — among other things — the continued viability of "hot news" misappropriation, a theory of unfair competition that dates back to the Supreme Court’s 1918 case, International News Service v. Associated Press, 248 U.S. 215 (1918), which involved unauthorized re-publication of wire service reports. Contrary to what Posner’s title might suggest, the article didn’t outright announce the death of the hot news doctrine, but it did paint a picture of a legal doctrine on the ropes — disdained by noted jurists, unwise as a matter of policy, and limited in practical significance. For better or worse, a decision issued last Thursday shows the doctrine to be very much alive and relevant. In fact, the case raises some disturbing prospects for news aggregation and sharing of information on the Internet more generally.
In Barclays Capital Inc. v. TheFlyOnTheWall.com, 06 Civ. 4908 (S.D.N.Y. Mar. 18, 2010), Judge Denise Cote of the United States District Court for the Southern District of New York issued a permanent injunction requiring the Internet-based financial news site FlyOnTheWall.com ("Fly") to delay its reporting of the stock recommendations of research analysts from three prominent Wall Street firms, Barclays Capital Inc., Merrill Lynch, and Morgan Stanley. The injunction requires Fly to wait until 10 a.m. E.S.T. before publishing the facts associated with analyst research released before the market opens, and to postpone publication for at least two hours for research issued after the opening bell.
The injunction is based on Judge Cote’s finding, after a bench trial, that Fly engaged in hot news misappropriation, "free-riding activity that is directly competitive with the Firms’ production of time-sensitive information, thereby substantially threatening their incentive to continue in the business." Barclays, slip op., at 87. Morgan Stanley and Barclays also succeeded on copyright infringement claims relating to Fly’s unauthorized copying and distribution of excerpts from their research reports for a few weeks in 2005, but the court awarded relatively minor damages on these claims and this doesn’t impact Fly’s current business practices, which no longer involve verbatim reproductions or close paraphrases of analyst research.
Like other Wall Street firms, Barclays, Merrill Lynch, and Morgan Stanley produce analyst research reports on stocks. The firms distribute these reports for a fee to their clients, usually large institutional investors. The firms often release these reports before the NYSE opens for the day, and the reports contain recommendations (buy/sell/hold) that, according to the firms, often spur investors into making trades, usually through the firm that issued the report. As a result, the release of a report often has a significant impact on the market price for the stock in question.
The firms’ paying clients gain access to the reports through several means, including the firms’ password-protected websites, licensed third-party distributors like Bloomberg and Thomson Reuters (presumably also using some sort of password protection), and email messages. In addition, the firms host private conference calls or webcasts in which their analysts discuss their research reports and recommendations with clients. Access to these calls and webcasts is restricted to those with the required passcode or login.
The firms take various precautions to ensure that the reports go only to paying clients. For example, they forbid employees from sharing the reports, their licensing agreements purport to forbid the clients from redistributing the research content, and licensed distributors like Bloomberg and Reuters contractually agree to maintain a "firewall" so that their media arms can’t obtain information from their research arms.
Inevitably, though, the research reports and the recommendations contained in them leak out, and Fly pioneered the business model of publishing this information for its own clients on a newsfeed over the Internet. The model has caught on, and, according to the court, presently "there is a crowded marketplace with small internet companies and major news organizations reporting the Firms’ Recommendations before and after the market opens." Barclays, slip op. at 35.
According to Judge Cote’s opinion, it looks like Fly’s operations have changed significantly over the last few years, largely in response to the firms’ lawsuit. Before 2005, Fly relied primarily on employees at the firms who emailed research reports to Fly after they were released to clients (this was pretty clearly a violation of the employees’ duties of loyalty and confidentiality to the firms). At that time, Fly staff would type the recommendation as a headline, sometimes accompanied by a verbatim reproduction or close paraphrase of a passage from the report explaining the basis for the recommendation. Id. at 32. Hence the copyright claims for Fly’s conduct in 2005.
As a result of the lawsuit, however, Fly apparently changed its information-gathering process. According to testimony from Ron Etergino, Fly’s president and majority owner, he "no longer feels free to look at the research reports, even if someone should send them to him," id. at 33, and he now gathers information about the firms’ reports from other sources:
According to Etergino, he checks first to see what Recommendations have been reported on Bloomberg Market News. Then he checks Dow Jones, Thomson Reuters, and Fly’s competitors such as TTN, StreetAcount.com, and Briefing.com. Next, he visits chat rooms to which he has been invited to participate by the moderator. . . . Etergino also receives "blast IMs" through the Bloomberg, Thomson Reuters, or IMTrader messaging services that may go to dozens or hundreds of individuals. Finally, Etergino exchanges IMs, emails, and more rarely telephone calls with individual traders at hedge funds, money managers, and other contacts on Wall Street.
Id. at 34. In other words, Fly acquires information about the reports through a process that looks a whole lot like good-old fashioned journalism. And it largely relies on information that is publicly available through mainstream and Internet media reports, IM blasts, and what appear to be open chat rooms. The result is a headline like this: "EQIX: Equinox initiated with a Buy at FofA/Merrill. Target $110." Id. at 27.
Hot News and Copyright Law
As noted, the main dispute in the Barclays case was not about verbatim copying, but about Fly publishing time-sensitive facts from the firms’ research reports — essentially, the buy/sell recommendations. Facts are not protected by copyright law. Feist Publ’ns, Inc. v. Rural Telephone Service Co., Inc., 499 U.S. 340, 345 (1991). While the firms’ recommendations aren’t exactly facts in the same way as "hard news," the firms appeared to concede that they couldn’t stop Fly’s current reporting practices through resort to copyright law. Enter the hot news misappropriation doctrine, which is controversial precisely because it provides IP-like protection to facts despite copyright law’s bedrock policy that facts are in the public domain.
In International News Service v. Associated Press, 248 U.S. 215 (1918), the Supreme Court created the hot news misappropriation doctrine as a matter of federal common law, and some state courts, like those in New York, adopted it as part of state unfair competition law. The INS case arose after British and French censors barred INS from sending war dispatches to the United States because Hearst had offended the British and French by siding with Germany at the outset of WWI. See Posner, at 627. INS employees got around this problem by paraphrasing AP dispatches published in east coast newspapers and sending them by telegraph to the west coast for publication in Hearst newspapers. See INS, 248 U.S. at 231-32 (at issue was INS’ practice of "copying news from bulletin boards and from early editions of complainant’s newspapers and selling this, either bodily or after rewriting it, to defendant’s customers"); id. at 259-60 (Brandeis, J., dissenting) ("The means by which the International News Service obtains news gathered by the Associated Press is also clearly unobjectionable. It is taken from papers bought in the open market or from bulletins publicly posted.").
The INS Court acknowledged that AP had no copyright claim because it had failed to register and/or place notice on its news reports (no longer a requirement under U.S. copyright law), and because copyright law did not extend to the facts in the reports. But, the Court nonetheless enjoined INS from using AP’s news reports in direct competition with the news service, finding that the INS’s free riding "speaks for itself and a court of equity ought not to hesitate long in characterizing it as unfair competition in business." Id. at 240. Justices Holmes and Brandeis wrote powerful dissents, decrying the majority’s opinion as unprecedented, unnecessary, and unwise.
The main policy justification advanced by the majority, which remains the motivating principle behind hot news doctrine today, is that protecting hot-news-type information is necessary to preserve the incentives that drive economic actors to make the substantial investment required to produce a socially valuable product or service in the first place. Posner characterizes this policy impulse as protecting against the danger of "killing the goose that laid the golden eggs." Posner, at 628.
In the Barclays case, the idea is that Wall Street research reports are a social good — they help disseminate information important to the proper functioning of the securities markets that otherwise would not be produced. This may be a disputable proposition, but it’s one the court accepted. And, the theory goes, Wall Street firms like Barclays and Merrill Lynch won’t go to the expense of producing these socially valuable reports if companies like Fly can free ride off of them and undermine the money-making potential of the practice. Again, it’s disputable whether Fly’s conduct rather than other economic factors (like international economic meltdown) has hurt demand for the firms’ reports, but Judge Cote found as a matter of fact that Fly’s activities did create a substantial disincentive.
I’ll leave to the economists the question of whether or not all this is wise economic policy. But from a legal perspective, the hot news doctrine creates an obvious tension with copyright law because, as noted above, it creates a pseudo property right in facts that copyright law says are in the public domain. This raises the specter of preemption: that is, a situation where federal law displaces inconsistent state law under the Supremacy Clause. Judge Cote’s opinion in Barclays does a very thorough job on this issue and determines — rightly, in my view — that federal copyright law does not preempt hot news misappropriation, or at least a narrow version of it. This result was a foregone conclusion for Judge Cote because the Second Circuit Court of Appeals had already said as much in National Basketball Association v. Motorola, Inc., 105 F.3d 841 (2d Cir. 1997), which is controlling precedent in the Southern District of New York.
Under NBA, the narrow version of hot news misappropriation that survives copyright preemption has the following elements:
(i) a plaintiff generates or gathers information at a cost; (ii) the information is time-sensitive; (iii) a defendant’s use of the information constitutes free riding on the plaintiff’s efforts; (iv) the defendant is in direct competition with a product or service offered by the plaintiffs; and (v) the ability of other parties to free-ride on the efforts of the plaintiff or others would so reduce the incentive to produce the product or service that its existence or quality would be substantially threatened.
Barclays, slip op. at 55 (quoting NBA, 105 F.3d at 845). Posner says that the "meat" of the test is in element (v), with (i) through (iv) describing a situation where (v) is likely to be satisfied. Posner, at 632. Therefore, "[t]he criterion appears to mean that states can protect fact gathering without running afoul of the preemption provision in the federal copyright statute only when unauthorized copying of the facts is likely to deter the plaintiff or others similarly situated from gathering and disseminating the facts that the defendant has copied." Id. The test is "alarmingly fuzzy once the extreme position of creating a legal right against all free riding is rejected, as it must be." Id. at 638.
In other words, hot news doctrine presents an inherently subjective and necessarily fact-specific standard, and one would expect courts to be cautious in finding it met, if for no other reason than to avoid the potential conflict with copyright law and to promote the public’s access to information. In Barclays, the firms convinced Judge Cote at trial that each element was satisfied, showing that, while it may take a unique set of facts, it’s not an impossible task.
What About the First Amendment?
Notably lacking from Judge Cote’s very thorough opinion is any discussion of how hot news misappropriation interacts with the First Amendment. This could be because Fly didn’t argue the point, at least not directly. While this post suggests that Fly’s lawyers "played the free speech card," it is hard for me to believe that Judge Cote would fail to address such an important argument if it were raised directly in the briefs. We have a student looking through the documents on PACER, which are pretty extensive, but so far we haven’t turned up any direct invocation of the First Amendment, except for an affirmative defense in the answer. As we’ll see below, though, Fly undoubtedly raised factual arguments that bear on the question.
The First Amendment issue is an important one because the Supreme Court didn’t address it in INS. Justice Brandeis’s dissent gives us a First Amendment tingle in his famous statement, "[t]he general rule of law is, that the noblest of human productions — knowledge, truths ascertained, conceptions, and ideas — become, after voluntary communication to others, free as the air to common use," 248 U.S. at 250 (Brandeis, J., dissenting), but even he didn’t seem to appreciate the constitutional implications of the case. It’s also an important question because First Amendment doctrine has developed considerably since 1918, and free speech concerns of which the Justices had only a vague inkling now have become an accepted part of the constitutional landscape.
The First Amendment issue raised by the case is one I’ve addressed before. A long line of Supreme Court cases hold that the First Amendment protects truthful speech on matters of public concern. See, e.g., Bartnicki v. Vopper, 532 U.S. 514, 527-28, 533-35 (2001) (First Amendment barred imposition of civil damages under wiretapping law for publishing contents of conversation relevant to matter of public concern); Florida Star v. B.J.F., 491 U.S. 524, 534 (1989) (First Amendment barred imposition of civil damages on newspaper for publishing rape victim’s name); Smith v. Daily Mail Publ’g Co., 443 U.S. 97, 103-06 (1979) (First Amendment barred prosecution under state statute for publishing names of juvenile offenders without permission of court); Landmark Communications, Inc. v. Virginia, 435 U.S. 829, 841-42 (1978) (First Amendment barred criminal prosecution for disclosing information from a confidential judicial discipline proceeding). Therefore, “if a newspaper lawfully obtains truthful information about a matter of public significance then state officials may not constitutionally punish publication of the information, absent a need to further a state interest of the highest order.” Smith, 443 U.S. at 103; accord Bartnicki, 532 U.S. at 527-28.
In Bartnicki v. Vopper, members of a teachers union sued a radio announcer under state and federal wiretapping laws after he played an unlawfully recorded telephone conversation on the air. The radio show host had received the recording from a third party who himself had received the tape in the mail from an anonymous source. The Supreme Court held that the First Amendment prohibited the recovery of damages against the radio show host for publishing the tape, explaining that “a stranger’s illegal conduct does not suffice to remove the First Amendment shield from speech about a matter of public concern.” Id. at 535. The constitutional principle in Bartnicki and other Supreme Court cases is not limited to traditional forms of media like newspapers and radio broadcasters. See Mary T. Jean v. Massachusetts State Police, 492 F.3d 24 (1st Cir. 2007) (First Amendment barred criminal prosecution for posting illegally recorded video online when recording made by third party, even if knowing receipt of the recording constituted a crime under Massachusetts law).
In Barclays, Judge Cote considered it unimportant that Fly obtained the information it published from other news services that were publishing the firms’ recommendations on the Internet in advance of Fly’s own publication. The court said that "the conduct of third parties is simply of no moment in finding Fly liable for hot-news misappropriation," and "it is not a defense to misappropriation that a Recommendation is already in the public domain by the time Fly reports it." Barclay, slip op. at 61. This may be a faithful application of the INS case itself — recall that INS involved taking facts from publicly available bulletin boards and published newspaper accounts — but INS never considered the First Amendment, so it can’t resolve the issue.
Under Bartnicki and the cases mentioned above, if Fly obtained the information in question through lawful means, then the First Amendment protects its right to publish that information. There is nothing inherently unlawful about Fly reading about a stock recommendation on a newsfeed provided by another news service or participating in a public chat room where Wall Street "rumors" are discussed (accessing a passcode-protected conference call would be another matter). The court says that Fly has engaged in "illegal conduct" by publishing the information it did, Barclays, slip op. at 61, but this label begs the question — that is, whether the state may constitutionally penalize publication of truthful information relating to a matter of public concern that was not obtained in violation of any other applicable laws.
To be sure, the person who originally leaks a firm research report to a news service or chat room participant may violate a legal duty owed to one of the firms, but "a stranger’s illegal conduct" is not sufficient to remove First Amendment protection under Bartnicki. The question is closer for Fly’s pre-lawsuit-era publication of reports received directly from firm employees who violated a duty of loyalty and confidentiality. It might be independently "unlawful" in the constitutional sense to knowingly induce a breach of these duties, but even in the trade secrets sphere this question has not been resolved with any clarity. Furthermore, I’m not aware on anything that would make it "unlawful" for Fly to communicate by email or telephone with firm clients who are willing to convey the substance of the recommendations, though this probably violates the client’s license agreement. Regrettably, the court did not differentiate between Fly’s different information-gathering tactics, and it enjoined publication of information obtained through at least some practices that clearly aren’t "unlawful" in any meaningful sense.
The court might well respond to all this by arguing that the firms’ reports are not facts related to a matter of public concern like ordinary news, but rather "subjective judgments based on complex and imperfect evidence." Id. at 78. There may well be a constitutionally significant distinction between reporting the subjective recommendations generated by these Wall Street firms and objective, external facts that are discovered "out there" in the world. On the other hand, these subjective judgments have objective, real-world consequences, and the announcement of a recommendation is itself a newsworthy event because it may cause a change in a stock’s price. It strikes me as difficult, and potentially hazardous, to try to distinguish between reporting the "subjective" recommendations versus reporting the "objective" fact that they were made, especially when the publication in question looks like this: "EQIX: Equinox initiated with a Buy at FofA/Merrill. Target $110."
The court may have ameliorated some of the First Amendment concerns by clarifying that the scope of its injunction, like the scope of hot news misappropriation, is narrow:
[T]o the extent Fly alters its business and begins to engage in actual analysis of market movements, and refers on occasion after the market opens in New York to one of the Firms’ Recommendations in the context of independent analytical reporting on a significant market movement that has already occurred that same day, such conduct will not run afoul of the injunction.
Id. at 87-88. But, this description of speech activity (the court doesn’t frame it in terms of speech) that won’t be enjoined displays an obvious preference for original/sweat of the brow/"analytical" content-creation over the free transmission of facts and information, which is a lot of what happens on the Internet. This is a preference that hot news doctrine’s anti-free-riding purpose surely calls for, but I don’t believe the First Amendment shares this ideal. (Copyright sure doesn’t. See Feist, 499 U.S. at 359-60.) As I’ll touch on more below, the court’s logic here also has foreboding connotations for news aggregators and others who supposedly "free ride" by transmitting information to others over the Internet without engaging in "independent analytical reporting."
News Aggregators, Bloggers, and the Like
The $75,000 question is what the Barclays case means for other online news aggregators, as well as social media more generally. Will the major newspapers be able to use this case to revive a robust hot news misappropriation doctrine that will kill the news aggregators and lock down facts on the Internet? I have no doubt that AP lawyers are smiling to themselves this week, but I don’t think this decision spells doom for the Internet as we know it.
The bad news for aggregators, bloggers, and those who like to share news is that this is a detailed, thoroughly reasoned (with the First Amendment exception noted above) decision from a respected judge in one of the most prestigious federal district courts in the nation. And, the decision is the product of a full-blown trial, giving it a concreteness and specificity that other, Internet-related hot news decisions, like Associated Press v. All Headline News, 608 F. Supp. 2d 454, 458-61 (S.D.N.Y. 2009), lack. This will give the decision credibility and make it useful in the hands of future judges looking for direction.
Worse, there are moments when reading the opinion where one feels like Judge Cote might as well be talking about news aggregators or bloggers free riding on "original reporting" instead of equity research. The court’s concept of free riding (element iii of the NBA test) certainly sounds like it would apply to news aggregation or acts of curation more generally:
To the extent that Fly adds value through its collection and aggregation of information, however, the value reflected in that act of aggregation does not controvert the fact that Fly expends no effort to produce the Recommendations and does not contribute to the underlying research and analysis process.
Barclays, slip op. at 60. It’s not a huge logical jump to say that all news aggregators are "free-riding" because they "expend no effort" to produce original reporting, and therefore "do not contribute to the underlying [journalistic] process." But this logic vastly understates the social benefit contributed by news aggregators, as well as bloggers who curate and comment on the news without expending effort to create it, and it automatically tilts the scales in favor of content producers at the expense of informational services and commentary, without any real justification.
Also potentially troubling is the court’s willingness to attribute the firms’ disincentive to produce equity research to Fly’s online activities as opposed to global financial meltdown, a willingness we can only hope won’t be reproduced when it comes to evaluating the alleged contribution of news aggregators and social media to newspapers’ current financial plight. Courts need to take a very close look at what is causing newspapers to suffer hard times; increased competition and loss of monopoly advertising rents explain a lot more than headlines and ledes with a link back, but that’s a topic for another day.
In any event, on the all-important fifth element (killing the golden goose), the Barclays case is easily distinguishable because the firms made a good (if not bullet-proof) case that production of high-quality equity research implicates a special need for time-sensitive exclusivity so that firm clients can feel they uniquely benefit from the recommendations and so that these clients can place trades with the firms based on them. Most regular news doesn’t share this rivalrous character, and it may be extremely difficult for newspapers to show that news aggregation or blog commentary ultimately hurts their bottom lines.
As for blogs, Twitter, and other types of social media, Barclays is further distinguishable because of the direct and obvious competition between Fly and the firms, which will be lacking in all but the most unusual cases. When it comes to Google News, which may be a real competitor, the ability of news organizations to opt out using robots.txt makes it extremely difficult to argue that Google is free riding, much less that it is destroying all incentive to engage in original reporting.
Finally, I suspect that the move from the financial sector to the general news sector will brighten and clarify the First Amendment issue discussed above, making it harder for courts to ignore that hot news doctrine plainly contemplates restricting the publication of truthful information on matters of public concern, regardless of how that information is required.