Rationalizing (?) the Hart and Keller v. EA Sports Publicity Rights Rulings (Guest Blog Post)

By Guest Blogger Tyler Ochoa

[Eric's note: this is a long blog post from my colleague Tyler. It does a great job demonstrating that the interplay between the First Amendment and the publicity rights is complete anarchy, which isn't surprising given that we don't really understand what we're trying to accomplish with the publicity rights doctrines in the first place. The post also raises the unexplained paradox why courts are more solicitous of publicity rights than Lanham Act claims, even though the latter involves consumer confusion/deception while the former doesn't.]

The use of college athletes’ likenesses in sports-simulation videogames, such as Electronic Arts’ NCAA Football series, has spawned a number of lawsuits alleging that such use violates the athletes’ rights of publicity.  (These actions have been brought by retired college athletes, as the NCAA prohibits college athletes from commercially exploiting their rights of publicity while in college, as a condition of maintaining their “amateur” status.)  Two federal Courts of Appeals have now held 2-1 that the First Amendment does not protect Electronic Arts’ depiction of actual college players, so that EA may be held liable under state right of publicity laws.  The agreement between the two courts makes it considerably less likely that the Supreme Court will review either one of the cases when it resumes sitting in October.

EA’s NCAA Football game strives for realism and has achieved it to a large degree.  The in-game stadiums depicted look like the actual stadiums on college campuses; and the players for each team are modeled on the actual players who played for that team during the year depicted.  Each in-game player has the same uniform number, height, weight, attributes, and home town as the actual players.  EA does not use any names, in order to comply with its licensing agreement with the NCAA, in which it agrees not to use any actual players; but it is common for users to upload the names of the players from third-party sources.  EA also provides players with the ability to “customize” any player on the roster, so that the user can change the players’ height, weight, and other attributes.  Want to insert yourself as quarterback of your favorite team?  Go right ahead.

In Hart v. Electronic Arts, 717 F.3d 141 (3rd Cir. 2013), decided two months ago, [available at http://www2.ca3.uscourts.gov/opinarch/113750p.pdf ] former Rutgers quarterback Ryan Hart sued EA for using his likeness in its videogame.  As there is no real doubt that the player depicted was meant to be Hart, the prima facie case for liability was pretty straightforward: EA used players’ likenesses in a product for commercial gain.  The only significant question on summary judgment was whether EA’s depiction is protected by the First Amendment.  The District Court held that the videogame was protected by the First Amendment, a ruling that the Third Circuit panel reversed.

Similarly, in Keller v. Electronic Arts, No. 10-15387 (9th Cir. July 31, 2013), [available at http://cdn.ca9.uscourts.gov/datastore/opinions/2013/07/31/10-15387.pdf ] there was no real doubt that the player depicted was meant to be former Arizona State and Nebraska quarterback Sam Keller.  The Keller case was consolidated with a similar suit filed by former UCLA basketball player Ed O’Bannon and several others, which also challenged the NCAA’s restriction against players exploiting their rights of publicity while in college as an antitrust violation.  (Only the First Amendment issue was decided in this opinion; the antitrust issue remains pending.  Probably as a result, the NCAA has announced that it will no longer license EA after the forthcoming 2014 game, leaving individual colleges and universities to strike their own deals with EA.  EA has already signed an extension with the Collegiate Licensing Company, which controls trademark licensing for a large number of colleges and universities.)

Keller also involved a significant procedural complication.  The First Amendment issue was raised not in a motion for summary judgment, but in a special motion to strike under California’s anti-SLAPP law.  SLAPP is an acronym for “strategic lawsuits against public participation,” and such statutes are “designed to discourage suits that masquerade as ordinary lawsuits but are brought to deter common citizens from exercising their political or legal rights or to punish them for doing so [citation omitted].” California law allows a defendant in such a case to raise the First Amendment issue at the outset of the case, using a special motion to strike.  The standard, however, is lenient: although the statute requires the plaintiff to demonstrate that it has a “reasonable probability” that it will prevail, courts have watered down the statute by interpreting it as allowing the lawsuit to proceed unless the defendant would prevail under the First Amendment as a matter of law.  Applying this standard, the district court held that EA was not protected by the First Amendment as a matter of law, and the Ninth Circuit affirmed.  Thus, EA could theoretically still prevail at trial if the trier of fact found certain facts (discussed below) in its favor; but as a practical matter, the Ninth Circuit’s opinion is unlikely to be contradicted.

Background: Competing First Amendment Standards

An important question is what First Amendment analysis will be used in such lawsuits.  Courts facing First Amendment questions in right of publicity cases have set forth a large variety of different analyses.  In the only U.S. Supreme Court opinion to date, the Court held that the First Amendment “do[es] not immunize the media when they broadcast a performer’s entire act without his consent.”  Zacchini v. Scripps-Howard Broadcasting Co., 433 U.S. 562 (1977).  That case involved a 15-second clip of a human cannonball at a local county fair, shown on the local TV station’s news broadcast.  Unfortunately, the Court did not set forth any sort of general approach to deciding claims of this type, leaving lower courts to sort out the issues in later cases.  In particular, the Court did not adopt (or even refer to) either the “strict scrutiny” analysis typically used for content-based restrictions or the “intermediate scrutiny” analysis typically used for content-neutral restrictions that implicate First Amendment interests.  Interestingly, lower courts have declined to adopt either of these more general standards as well, and instead have largely preferred to use standards that are unique to the intellectual property context.

In two cases involving Clint Eastwood and the National Enquirer, in which Eastwood alleged that the tabloid knowingly printed false stories about him and advertised them on its cover in order to increase its circulation, courts have used the familiar standard from libel law, first set forth in New York Times v. Sullivan, 376 U.S. 254 (1964), under which a publisher may not be held liable for mere negligence in printing a false story about a public figure, but may only have damages imposed if it acted with “actual malice,” that is, if it knew the story was false or if it acted with reckless disregard as to its falsity.  Eastwood v. Superior Court, 149 Cal. App. 3d 409, 198 Cal. Rptr. 342 (1983); Eastwood v. National Enquirer, 123 F.3d 1249 (9th Cir. 1997).  Bizarrely, the same standard was used in a case in which Dustin Hoffman claimed that his likeness was used without permission in an article in L.A. Magazine, in which a still photograph of Hoffman in drag from the movie “Tootsie” was altered by superimposing another model’s body wearing a different dress, despite the fact that all parties agreed that L.A. Magazine did not try to “pass off” the photograph as real.  Indeed, the whole point of the story was to show famous scenes featuring Hollywood icons, digitally altered to wear modern fashion styles.  The Ninth Circuit held that the magazine article was protected by re-interpreting the test to inquire not whether the magazine knew that the depiction was false, but whether it had knowingly (and falsely) represented either that it was true, or that Hoffman had consented (finding a lack of evidence on both counts).  Hoffman v. Capital Cities/ABC, Inc., 255 F.3d 1180 (9th Cir. 2001).

In a notorious case, game show hostess Vanna White sued Samsung for an advertisement that depicted a robot in a blonde wig and evening gown, turning letters on a Wheel-of-Fortune game board.  Because the ad was “commercial speech,” a majority of the Ninth Circuit panel gave the First Amendment argument short shrift, dismissing it in a single short paragraph.  White v. Samsung Electronics America, Inc., 971 F.2d 1395 (9th Cir. 1992).  White was subsequently awarded $400,000 in damages.  Judge Kozinski’s opinion dissenting from the denial of rehearing en banc, arguing from stronger First Amendment protection, has become a staple of law-school casebooks.  989 F.2d 1512 (9th Cir. 1993).  White is arguably inconsistent with the later U.S. Supreme Court decision in Campbell v. Acuff-Rose, Inc., 510 U.S. 569 (1994), which held that an alleged rap parody of the popular song Pretty Woman could qualify as a parody protected by the copyright fair use doctrine.  Moreover, in a later case involving parody baseball cards, the Tenth Circuit expressly disagreed with White, holding that the First Amendment protected lampoons of active professional baseball players (such as the depiction of slugger Barry Bonds as “Treasury Bonds”).  Cardtoons, L.C. v. Major League Baseball Players Ass’n, 95 F.3d 959 (10th Cir. 1996).  There, the Court systematically analyzed the effect of a hypothetical injunction on the right of free speech, and balanced it against the effect of a hypothetical ruling against the players on the right of publicity.  The court examined seven rationales advanced for the right of publicity, and found either that each was unpersuasive or that it would not be significantly advanced by a ruling in favor of the players.

In Rogers v. Grimaldi, 875 F.2d 994 (2d Cir. 1989), actress Ginger Rogers sued the producers of a movie entitled “Ginger and Fred,” about a fictional pair of Italian dancers who were compared to Rogers and Astaire and acquired the nicknames “Ginger and Fred.”  Her suit was based on Section 43(a) of the Lanham Act (false endorsement) as well as Oregon right of publicity.  The Second Circuit rejected a proposed standard drawn from a real property case, Lloyd Corp. v. Tanner, 407 U.S. 551 (1972), in which the court held that the First Amendment did not require states to allow distribution of handbills on private property unless there were “no alternative avenues of expression.”  Instead, the Second Circuit held that the Lanham Act was outweighed by the First Amendment “unless the title has no artistic relevance to the underlying work whatsoever, or, if it has some artistic relevance, unless the title explicitly misleads as to the source or the content of the work.”  The court applied a similar standard to the right of publicity claim, holding that the First Amendment prevailed “unless the title was wholly unrelated to the movie or was simply a disguised commercial advertisement for the sale of goods or services.”  Both the Second Circuit and the Ninth Circuit have subsequently expanded the reach of the Rogers test to the use of trademarks in the content of expressive works.  Cliffs Notes, Inc. v. Bantam Doubleday, Inc., 886 F.2d 490 (2d Cir. 1989); Mattel, Inc. v. MCA Records, 296 F.3d 894 (9th Cir. 2002); E.S.S. Entertainment, Inc. v. Rock Star Videos, Inc., 547 F.3d 1095 (9th Cir. 2008).  The Rogers test is very protective of free speech.  The only appellate case in which the defendant failed to satisfy the Rogers standard is Parks v. LaFace Records, 329 F.3d 437 (6th Cir. 2003), in which Rosa Parks successfully sued over the use of her name as the title of a rap song which used the phrase “Everybody move to the back of the bus,” but not as a reference to segregation.  The court held that the song was “wholly unrelated” to the civil rights icon.

Two cases involving works of art introduced two additional First Amendment analyses.  In Comedy III Prods v. Gary Saderup, Inc., 25 Cal. 4th 387, 106 Cal. Rptr. 2d 126 (2001), the California Supreme Court held that reproductions of a charcoal drawing of the Three Stooges were not protected by the First Amendment.  Despite rejecting six proffered reasons why the lithographs and t-shirts might not be protected by the First Amendment, the court held that the First Amendment interest was outweighed by the right of publicity unless the use was “transformative,” a term borrowed from the copyright “fair use” doctrine.  “When artistic expression takes the form of a literal depiction or imitation of a celebrity for commercial gain, . . . the state interest in protecting the fruits of artistic labor outweighs the expressive interests of the imitative artist.  On the other hand, when a work contains significant transformative elements, . . . First Amendment protection of such works outweighs whatever interest the state may have in enforcing the right of publicity.”  The Saderup court added two additional inquiries.  “Another way of stating the inquiry is whether the celebrity likeness is one of the ‘raw materials’ from which an original work is synthesized, or whether the depiction or imitation of the celebrity is the very sum and substance of the work in question. . . . Furthermore, in determine whether a work is sufficiently transformative, courts may find useful a subsidiary inquiry . . .: does the marketability and economic value of the challenged work derive primarily from the fame of the celebrity depicted . . . [or] from the creativity, skill, and reputation of the artist[?]”

In ETW Corp. v. Jireh Publishing, Inc., 332 F.3d 915 (6th Cir. 2003), however, the Sixth Circuit held that reproductions of a painting of Tiger Woods winning the 1997 Masters golf tournament were protected by the First Amendment.  The court found that the Saderup “transformative” standard was satisfied; but it also applied the Rogers v. Grimaldi standard, and a third standard, taken from the Restatement (Third) of Unfair Competition, § 47, cmt. c, which states that “the use of a person’s identity primarily for the purpose of communicating information or expressing ideas is not generally actionable as a violation of the person’s right of publicity,” unless the name or likeness is used solely to attract attention to a work that is not related to the identified person, or if the work contains substantial falsifications.  In a meandering opinion, the Court held that the poster was protected under any of the three standards.

Cases involving the use of real people as models for comic-book characters have also reached divergent results.  In Winter v. DC Comics, 30 Cal. 4th 881, 134 Cal. Rptr. 2d 634 (2003), singers Johnny and Edgar Winter were depicted in a Jonah Hex comic book series as “Johnny and Edgar Autumn,” two half-worm, half-human villains, or “vile, depraved, stupid, cowardly, subhuman individuals who engage in wanton acts of violence, murder and bestiality for pleasure and who should be killed.”  The California Supreme Court unanimously held that the comic book met the “transformative” standard of Saderup and was protected by the First Amendment.  But when hockey player Tony Twist’s name was used by writer Todd McFarlane as the name of a Mafia don in the comic book series Spawn, the Missouri Supreme Court rejected both the Restatement and the “transformative” standard, adopting instead a “predominant use” test (proposed by a prominent right-of-publicity plaintiff’s attorney in a law review article), which looks at whether the product “predominantly exploits the commercial value of an individual’s identity, . . . even if there is some ‘expressive’ content in it that might qualify as ‘speech’ in other circumstances.”  Doe v. TCI Cablevision, Inc., 110 S.W.3d 363 (Mo. 2003).  [McFarlane was significantly damaged by his apparent attempt to ward off a defamation claim by claiming that his Mafia don was not a comment on the real Tony Twist, and by the fact that he occasionally marketed Spawn and related merchandise to hockey fans.]  Not surprisingly, given its origin, this standard is most protective of celebrities and most hostile to free speech.

Choosing a Standard

So, which of these competing standards should be used in the videogame cases?  Both Cardtoons and White are pretty clearly limited to parody cases, and Eastwood is pretty clearly limited to false news reports masquerading as true (despite the Hoffman departure).  That still leaves three (or four) standards: Rogers, the Restatement standard (to the extent it differs), the Saderup “transformative” standard, and the Doe “predominant use” standard.

The Hart opinion criticized the “predominant use” test as “subjective at best, arbitrary at worst,” and for requiring courts “to analyze select elements of a work to determine how much they contribute to the entire work’s expressiveness.”  Moreover, the court said, adopting this test “would suppose that there exists a broad range of seemingly expressive speech that has no First Amendment value.”

Both Hart and Keller rejected application of the Rogers test.  The Hart court was concerned that the Rogers test was “a blunt instrument, unfit for widespread application in cases that require carefully calibrated balancing,” and suggested that “adopting this test would potentially immunize a broad swath of tortious activity.”  In particular, the court was concerned that the “wholly unrelated” standard would be easily met by any product targeted at sports fans.  “It cannot be that the very activity by which [Hart] achieved his renown now prevents him from protecting his hard-won celebrity.”

Both Hart and Keller suggested that the Sixth Circuit’s use of Rogers in the Parks case was less persuasive because the same court, in deciding ETW a few months later, did not clearly use Rogers, but also relied on the Restatement and Saderup.  However, both courts also indicated that the Rogers standard should continue to be applied to false endorsement cases brought under the Lanham Act.  Indeed, the Ninth Circuit so held in a companion case brought by former professional football player Jim Brown, in which the state-law right of publicity claims were dismissed for lack of diversity jurisdiction after the federal Lanham Act claim was decided.”  Brown v. Electronic Arts, No. 09-56675 (9th Cir. July 31, 2013) [available at http://cdn.ca9.uscourts.gov/datastore/opinions/2013/07/31/09-56675.pdf ]  The Keller court noted that Rogers was designed “to protect consumers against the risk of consumer confusion — the hallmark element of a Lanham Act claim,” whereas “[t]he right of publicity protects the celebrity, not the consumer,” and does not require any showing of confusion.

This double standard makes little sense.  Serving the public interest by protecting against consumer confusion should be a stronger, more compelling government interest than protecting the purely private interest of the celebrity, yet both courts would apply a much more speech-protective First Amendment standard in cases involving the former interest than in cases involving the latter.  Moreover, in Rogers itself, the Second Circuit applied a nearly-identical analysis to the right of publicity claim as it did to the Lanham Act claim. Yet Keller says that it did so only because “Oregon courts . . . [had] not determined the scope of the common-law right of publicity in that state,” whereas four California courts have already applied the Saderup “transformative” standard. Thus, the Keller court bizarrely suggests that the Rogers “wholly unrelated” standard was merely a federal court’s Erie prediction about a matter of state tort law, rather than a federal First Amendment limitation on state tort law.

Instead, both Hart and Keller applied the Saderup “transformative” standard.  The Keller court did so essentially by default, whereas the Hart court said that the “transformative” standard “appears to strike the best balance because it provides courts with a flexible — yet uniformly applicable — analytical framework,” and because it thought it was consistent with most of the previously decided cases.

Applying the Standard

A fundamental question when applying the Saderup standard is: what is it that has to be “transformative”?  Does the celebrity likeness itself have be “transformative”?  Or is it sufficient if the work as a whole is “transformative”?  Saderup itself consistently indicates that the inquiry is whether the “work” is transformative.  [That is also true in Campbell v. Acuff-Rose Music, the U.S. Supreme Court case discussing the copyright fair use doctrine from which the California Supreme Court borrowed the “transformative” inquiry.]  Yet both Hart and Keller hold instead that what matters is whether the depiction of the celebrity is transformative, rather than the work as a whole.  Instead of looking to Saderup, Keller instead relied on a California Court of Appeal case, No Doubt v. Activision Publishing, 192 Cal. App. 4th 1018, 122 Cal. Rptr. 3d 397 (2011), in which a rock band successfully sued the makers of the videogame Band Hero, for including avatars of the band members in a videogame that allowed those avatars to perform songs other than those made popular by the band, a flexibility that exceeded the scope of the license that the band members had granted.  The Keller court thus (somewhat strangely) abdicated its role as a federal court to construe federal law, and instead deferred to a state court’s view of federal law.  The Hart court also relied on No Doubt, but it further justified its decision on the ground that otherwise, “[a]cts of blatant misappropriation would count for nothing so long as the larger work, on balance, contained contained highly creative elements in great abundance.”

A further question is: of what significance it is that avatars can be altered by the users?  This fact potentially distinguishes Hart and Keller from No Doubt, in which the court specifically noted that the avatars could not be altered.  Nonetheless, the Hart court held that the ability to alter avatars was not material, because the realism associated with real players was an important factor in getting consumers to buy the product.  This reasoning suggests a possible work-around, in which EA distributes the game with “generic” avatars at each position, and lets users modify the avatars themselves.  (There is little doubt that a market for user-generated content supplying the various “attributes” would immediately arise, but then litigation might have to be directed at the suppliers of such content, rather than at the EA videogame itself.  Surely the fact that EA’s avatars are capable of being modified to represent people of different height, weight, ethnicity, etc., cannot itself be a basis for liability.)

Not surprisingly, given this narrow focus on the avatars themselves and discounting all of the other contributions of EA’s artists and programmers, including the ability of users to alter the avatars, both courts held that the avatars were not transformative, and therefore they were not protected by the First Amendment.  In both cases, however, one of the three judges dissented on exactly these two issues: that in deciding whether the works were “transformative,” the court should consider the game as a whole, and especially the ability of users to alter the avatars to their own liking.

A further irony is that all of the attributes that “identify” the football players: their numbers, positions, height, weight, ethnicity, hometown, etc., and all of the statistics on which the “random” algorithms controlling the videogame are based, are publicly available information.  The Eight Circuit has already held that the use of such publicly available information in providing commercial fantasy sports products to the public is protected by the First Amendment.  C.B.C. Distrib. & Mktg., Inc. v. Major League Baseball Advanced Media, L.P., 505 F.3d 818 (8th Cir. 2007).  The only difference between using statistics to run a fantasy football game and using statistics to run a fantasy football videogame is the visual element.  Essentially, the court holds that the visual element — the depiction of a body type similar to a celebrity, imitating the celebrity — is enough to take the game out of realm of fully protected speech and to put it into the same category with plastic action figures, coffee mugs, and other celebrity merchandise.  If taken literally, the opinions in Hart and Keller could be read to bar the literal depiction of a celebrity in a work of historical fiction, such as Forrest Gump, or a “docudrama” such as the movie 42 (about Jackie Robinson).  I am confident that this will not come to pass, and that courts will continue to give First Amendment protection to depictions of real people in “traditional” entertainment media, such as movies and television, even when lightly (or heavily) fictionalized.  See, e.g., Guglielmi v. Spelling-Goldberg Prods., 25 Cal. 3d 860, 160 Cal. Rptr. 352 (1979) (fictional movie about Rudolf Valentino); Tyne v. Time Warner Entertainment Co., 901 So.2d 802 (Fla. 2005) (suit by heirs of the fisherman depicted in the movie The Perfect Storm); Ruffin-Steinback v. dePasse, 267 F.3d 457 (6th Cir. 2001) (TV miniseries about singing group The Temptations).  But the fact that courts cannot yet articulate a consistent First Amendment standard that distinguishes between the literal depiction of a celebrity in a sports-simulation videogame and the literal depiction of a celebrity in a more traditional work of entertainment strongly suggests that courts simply do not place the same value on the videogame medium as they do on more traditional media.  For a extensive rebuttal to this apparent discrimination, see William K. Ford and Raizel Liebler, Games Are Not Coffee Mugs: Games and the Right of Publicity, 29 Santa Clara Computer & High Tech. L.J. 1 (2012) [available at http://digitalcommons.law.scu.edu/chtlj/vol29/iss1/1 ]

Although there was a split at the district court level, the fact that both the Third Circuit and the Ninth Circuit reached the same result, using almost the same reasoning, significantly decreases the likelihood that the U.S. Supreme Court will grant review.  That possibility is further diminished by the procedural posture of the Keller case: the Supreme Court prefers to review final judgments, rather than intermediate rulings in cases in which significant court proceedings are still to come.  EA would be better off filing a petition for rehearing en banc in the Keller case.  (It has already filed such a petition in Hart).   Only two Ninth Circuit judges voted in the Keller case, and they split on the result.  The deciding vote was cast by a Senior U.S. District Judge from the Western District of Michigan, sitting by designation. (Senior Ninth Circuit Judge Tashima formed part of the majority in Hart while sitting by designation on the Third Circuit, but judges with senior status are not eligible to vote on petitions for rehearing.)  If a majority of the 27 active judges of the Ninth Circuit vote to grant review, the case would be reheard by a panel of 11 judges, including Chief Judge Alex Kozinski, a videogame aficionado who has twice previously expressed support for a robust First Amendment test in right of publicity cases, and who used the Rogers standard in his opinion in Mattel, Inc. v. MCA Records, 296 F.3d 894 (9th Cir. 2002).  If either en banc court were to reach a different result, then the parties would have a much better chance of getting the U.S. Supreme Court to review the case.