strip club photo litigation: a genre with rulings on false advertising and surveys

Skinner v. Tuscan Inc., No. CV-18-00319-TUC-RCC, 2020 WL 5946898 (D. Ariz. Oct. 7, 2020)

There’s a small cottage industry of right of publicity etc. claims against strip clubs and adjacent businesses, rivalling the timeshare rescue litigation industry; someone could do a public service by tracking these down. Defendant (Ten) used “risqué photos of Plaintiffs” to advertise its strip club. Plaintiffs brought claims for right of publicity/misappropriation of likeness and false light/invasion of privacy, as well as Lanham Act false advertising/false association/endorsement claims. As is becoming standard, plaintiffs do pretty well in the analysis.

Arizona false light/invasion of privacy: One-year statute of limitations; publication starts it running and continued posting online is not a “continuing wrong.” For photos published within that period, there were genuine issues of material fact on whether “the false light in which the plaintiff was placed would be highly offensive to a reasonable person in the plaintiff’s position.”

Ten argued that the ads made no false statements, but implications count and “[a] fact finder could decide that Plaintiffs’ images and the corresponding text in the advertisements falsely suggest the Plaintiffs were somehow affiliated with, promoted, or employees at Defendant’s strip club.” . Ten further argued that the underlying photos were already risqué, but “simply because a woman has modeled in risqué clothing (or even previously worked at a strip club) does not mean a reasonable person in a similar position could not be offended by the suggestion that the person is an exotic dancer at the defendant’s strip club.… Moreover, Plaintiffs’ have declared that the stripper lifestyle is disreputable and being impermissibly associated with such is offensive.”  However, a fact finder could also determine that any misrepresentation was minor and not actionable.

False light for a public figure also requires actual malice (knowledge of falsity or reckless disregard of the truth). This was a factual issue, though the court didn’t make clear exactly what Ten had to know/recklessly disregard: logically, it should be that Ten had to know or recklessly disregard the false implication of connection. Unfortunately, the court—possibly invited by the parties—focused on what Ten knew about the permission status of the photos, which is not the same thing, and cited a case holding that “failing to investigate the origin of models’ photos used in advertising and permission to use raised a genuine issue of fact as to actual malice.”

Arizona common law right of publicity: The right of publicity is rooted in property, not privacy, and thus subject to a two-year statute of limitations. (The court didn’t resolve whether the continuing wrong doctrine applied.) There were no genuine material disputes and plaintiffs were entitled to summary judgment: Ten received an advantage for uses of their photos, and failure to pay them for the use was cognizable harm.

Lanham Act false advertising: Plaintiffs don’t fall within the zone of interests for false advertising. While lost income from a missed photo shoot could be a financial injury, the alleged false advertising wasn’t harmful to their ability to compete.  “Not being hired by Defendant is not equivalent to not being able to compete with Defendant.” Direct competition isn’t required, but a bare assertion of overlapping commercial interests wasn’t enough under Lexmark. “Moreover, Plaintiffs make no allegations that their ability to obtain modeling jobs has been affected by the false advertisement.” Even if they were embarrassed from being wrongly affiliated with Ten’s, “they have produced no evidence that this association has damaged their reputation or their ability to compete in any fashion. Without direct competition or comparative advertising, Plaintiffs must show actual injury from the Defendant’s deception and have failed to do so.”

Likewise, even if plaintiffs were in the zone of interests for false advertising, they didn’t show causation. The only alleged injury, the loss of photoshoot income, “was neither caused by consumers withholding money, nor was it a result of consumer deception; the injury resulted from Defendant’s failure to pay for the photos. Defendant was not deceived by its own misleading advertising, and no consumer was hoodwinked into not paying Plaintiffs but instead giving Defendant his or her money.”

False association: material issues existed on likely confusion, despite the above analysis (which also implies lack of proximate causation of harm from false association). The court emphasizes the I-thought-deprecated “internet troika” of mark similarity, product/service relatedness, and marketing channels. The court equated copying photos to similarity of marks, implicitly holding that any image of the plaintiff is their “mark” (consistent with its idea that “recognizability is a measure of the strength of Plaintiffs’ mark, not the similarity”). The court enhanced the effect of its rounding-up in favor of plaintiffs by quoting prior language: “When the alleged infringer knowingly adopts a mark similar to another’s, reviewing courts presume that the defendant can accomplish his purpose: that is, that the public will be deceived.” But that logic depends on the alleged infringer knowing that there was a mark as opposed to a generic image of a pretty girl. On the intent factor specifically, though, it found that there was a factual issue of whether Ten was just looking for pictures of attractive women or intending to imply endorsement.

Relatedness of goods tipped “slightly” in Ten’s favor; they both use social media, but then so does everyone. There was also not enough evidence on consumers’ degree of care. There was a genuine issue of fact as to whether the plaintiffs “are recognizable to members of the community for which Defendant’s advertising is focused.” Plaintiffs submitted a survey “indicating a small percentage of those surveyed felt they recognized Plaintiffs” and of strong social media findings; Ten didn’t have its own survey, though it identified evidence that plaintiffs’ current recognizability is questionable.

Plaintiffs submitted a survey “indicating that approximately 66 percent of interviewees believed that Plaintiffs would likely participate in the strip club activities at Ten’s, 65 percent thought Plaintiffs were in some way affiliated with Ten’s, and 87 percent felt that it was very or somewhat likely that the Plaintiffs were representative of those employees that performed at Ten’s.” But a fact finder could agree with Ten’s criticisms of the survey methods and the clarity of the questions.

Given the existence of many similar cases, Ten sought to have nonmutual defensive collateral estoppel apply to some of the plaintiffs who’d lost similar Lanham Act/right of publicity claims against other defendants. The court declined to do so; it wasn’t bound by the prior cases’ holdings on surveys/strength of the mark, and the survey expert’s testimony has been allowed in some cases albeit excluded in others. And one case involving a cowboy bar wasn’t as offensive a use.

Skinner v. Tuscan, Inc., 2020 WL 5946897, No. CV-18-00319-TUC-RCC (D. Ariz. Oct. 7, 2020)

This opinion refuses to exclude three experts: two of plaintiffs’ (survey and damages) and one of defendant’s (damages).

Plaintiffs’ survey expert, Buncher, used 600 respondents in two groups; “selected” participants, “half of whom were men, resided within a certain radius of Ten’s and had attended strip clubs in the previous two years.” As noted above, the survey found a high degree of confusion over affiliation/sponsorship. Around 15% of participants believed they recognized the plaintiffs (which is still not the same thing as having a trademark, as the Tiger Woods case made clear). The court found that numerous criticisms of the survey went to weight rather than admissibility: The lack of a control group wasn’t fatal, nor was alleged distorting effects of the questions, nor was the absence of a don’t know option/anti-guessing instruction, nor was the mismatch between respondents and Ten’s clientele. Ten noted that Buncher’s survey evidence had been excluded in other similar cases.

Lack of a control group: Buncher testified that “the survey was a communications study designed to evaluate what messages Defendant’s advertisements communicated to the audience. Unlike a causal study, he claims, communications studies do not require a control group.” [Um. All Lanham Act surveys, so defined, would be “communications studies.” Plus, Lanham Act claims are causal claims: the ads caused confusion of a relevant sort. This is just bad reasoning.] Instead, he used a control question that excluded plaintiffs’ images from the ads and asked respondents how the exclusion affected the perception of the ad. This, he claimed, was “consistent with the logic of the Diamond research standard.” [I can hear Prof. Diamond exclaiming in horror from here. Ten is right that this is not a control question; it is instead a biased comparison drawing attention to the absence of a photo that will predictably elicit “oh, the ad is worse without a picture,” and it fails to test whether the difference is from the absence of a picture versus the absence of a picture of plaintiffs, the only relevant legal question. Even a control question that asked about a different ad with a picture of a non-plaintiff would have done better at measuring reactions to the presence of the plaintiffs—which, one might infer, is why Buncher didn’t use a real control.]

Anyway, objections based on an expert’s “methodology [and] survey design ... go to the weight of the survey rather than its admissibility.”  [I teach my students that, while this is often true, a bad enough survey can just be excluded, especially if there are also problems with the respondent selection; I would have put this one in that category, but the court finds each of these decisions to be just fine and so it doesn’t.]

As for the no response/no opinion issue, Buncher testified that the literature supported the conclusion that “permitting these non-responses would actually increase the amount of guessing.” This is the only decent defense of the survey in the bunch; it seems to be a legitimately contested issue.

Ten also pointed out that the survey was flawed “because participants were not given the opportunity to specify Plaintiffs’ true names and could not express uncertainty. Because of this, Defendant claims, no Plaintiff has been truly identified and the survey results cannot measure identifiability.” But the court disagreed because “it is possible to recognize a person without recalling their name. In fact, as the undersigned is learning all too well, with age this occurs more and more frequently.” That went to weight, anyway, as did criticisms of the ambiguity of the survey, e.g. asking “whether the participants felt the Plaintiff enjoyed the lifestyle portrayed by the strip club or participated in Ten’s events.” “ ‘[T]echnical inadequacies’ in a survey, ‘including the format of the questions or the manner in which it was taken, bear on the weight of the evidence, not its admissibility.’ ” [Note: wording is not format or manner.]

Sample: Buncher testified that the survey used 50% women in order to isolate gender to determine whether the message portrayed in the ads differed by sex. “As Defendant did not provide a clientele list to Plaintiffs, it is difficult to say how a more accurate representative sample of Ten’s clientele could be obtained.” Good enough.

Ultimately, Buncher indicated that he created the survey to conform “with the generally accepted standards and procedures in the fielding of surveys set forth by the American Marketing Association, Marketing Research Association, CASRO and ESOMAR” and “[t]he survey was designed to meet the criteria for survey trustworthiness detailed in the Federal Judicial Center’s Manual for Complex Litigation, Fourth.” Plaintiffs showed by a preponderance of the evidence that he used acceptable methods.

The survey was also relevant despite its failure to make respondents identify the plaintiffs. As for prior exclusions of his evidence, “[w]hile the Second Circuit may exclude survey evidence upon a finding that the methodology is lacking, in this circuit questionable methodology goes to the weight, not the admissibility.” And Buncher’s evidence has “been unilaterally permitted in the District of Arizona in similar cases.” [Uniformly?]

Plaintiffs’ damage expert Chamberlin estimated actual damages at an aggregate amount of $435,000. He first calculated each plaintiff’s day rate—how much they would have been paid to produce the photographs used by Ten’s under a hypothetical negotiated contract—and then multiplied the day rate by the number of Ten’s distinct uses of each photograph. [I really don’t understand how that could be a credible methodology. The day rate is about how long it takes the plaintiff to pose for the photo; the number of uses by the defendant is completely independent of that, even though it could plausibly be the basis for a different damages calculation. It’s like multiplying lost work days from an injury by the cost of treating that injury.]  

It was ok to use each plaintiff’s most lucrative contract to estimate the day rates based on his testimony that models get different rates based on the product they’re promoting. More troublingly, it was ok to use an image multiplier even though none of the disclosed modeling contracts use the kind of “usage multiplier” for distinct usages like advertisements, branding, and social media; rather, Defendant states, the contracts reflect a flat rate per job. Plaintiffs [though it seems not Chamberlin himself] argued that “any modeling agent would consider [various different uses] when negotiating a modeling contract.” [OK, but then wouldn’t the contract be use-based rather than day rate-based? That is, the testimony above indicates that the day rate itself would change based on the use. Are multipliers standard in the industry? If Ten is correct that they weren’t used in any of the contracts on which Chamberlin based his opinion, that seems like a problem.]

But “a party’s disagreement with the sources upon which the expert bases his or her conclusions goes to weight of the evidence and not admissibility,” so his choice of high-value contracts didn’t itself render his testimony unreliable. [I wish the court would have addressed the criticism of “multiply day rate by number of uses” specifically. I just … don’t get it.] The fact that other courts have accepted the same criticisms wasn’t dispositive because the Second Circuit, where those courts were, does things differently than the District of Arizona.

Defednant’s expert Einhorn also got in. He calculated the day rate (including a 50% premium for similar risqué photo shoots) and divided it by the number of images likely to be produced in a one-day photo shoot. He subtracted a 20% modeling agent fee and multiplied the final amount by the number of uses of the photograph. He concluded that the total actual damages ranged from $1,990 to $3,980 per plaintiff. Even though he lacked experience with the modeling industry specifically, his professional experience as a forensic economist and as an expert witness in cases involving “intellectual property, media, entertainment, technology, trademarks, publicity rights, and product design” was sufficient. Even though he relied on other models’ modeling contracts, and used an internet search of the phrase “working day rate, models” to support his day rate calculations, among other criticisms, that went only to weight and not admissibility.