Tuesday, January 30, 2024

no summary judgment on "non-toxic" and "Earth friendly"

Bush v. Rust-Oleum Corp., 2024 WL 308263, No. 20-cv-03268-LB (N.D. Cal. Jan. 26, 2024)

Bush alleged that Rust-Oleum falsely labeled of its “Krud Kutter” cleaning products as “non-toxic” and “Earth friendly.” Rust-Oleum sought summary judgment on the California consumer protection claims based on the argument that likely deception was refuted by disclaimers on the labels themselves and testimony from the plaintiff and his expert toxicologist. The court declined to grant summary judgment.

At the pleading stage in this case, the court held that the plaintiff’s definition of “non-toxic” — that “the product[s] did not pose any risk to humans, animals, or the environment” — was sufficient. But “the plaintiff and his expert toxicologist said during their depositions that risk can never be completely eliminated (for example, even water can be toxic in excess amounts),” so no reasonable consumer would believe the Krud Krutter products to be totally free of risk. Here, whether the plaintiff’s asserted definitions are reasonable were for the jury to decide as part of the overall reasonable-consumer test.

For the challenged claim “non-toxic,” Rust-Oleum argued that the plaintiff’s expert toxicologist’s theory of toxicity is disclosed on the front labels of its products, which say “Caution: Eye and Skin Irritant” next to the words “Non-Toxic.” But the expert opinion goes beyond eye and skin irritation, creating a genuine dispute of fact.

For the challenged claim “Earth friendly,” the rear of the product labels provide a definition of the claim. “But the definition is in small type and the defendant’s own surveys provide evidence that most consumers do not read it.” Again, a fact issue, and “Earth friendly” was not so general or nonspecific as to make it “extremely unlikely” that a consumer would rely on it. The defendant’s own surveys suggested as much, and California law did as well. White v. Kroger Co., No. 21-cv-08004-RS, 2022 WL 888657, at *2 (N.D. Cal. Mar. 25, 2022) (“California view[s] terms on the label or container of a consumer good like ... ‘earth friendly’ ... to mean that the product is not harmful to, or is beneficial to, the natural environment. While ... [this] California statute [does not] directly create[ ] a private cause of action, [it] do[es] undermine any argument that ‘reef friendly’ can be dismissed as mere puffery.”) (cleaned up).

Thursday, January 25, 2024

New article: Three Sizes Fit Some: Why Content Regulation Needs Test Suites

 At the Berkeley Tech LJ: 


The European Union’s Digital Services Act (DSA) offers a new model for regulating online services that allow users to post things. It uses size-based tiers to delineate the different levels of obligation imposed on various services. Despite the tiers of regulation in the DSA, and very much in its copyright-specific companion Article 17, it’s evident that the broad contours of the new rules were written with insufficient attention to variation. Instead, regulators assumed that “the internet” largely behaved like YouTube and Facebook. Using three examples of how that model is likely to be bad for a thriving online ecosystem—counting users, providing due process, and implementing copyright-specific rules—this Article concludes that, to improve policymaking, regulators should use test suites of differently situated services to ensure that they are at least considering existing diversity and properly identifying their targets.

Friday, January 19, 2024

Reading list: geolocation data increased incivility online

Civilizing social media: The effect of geolocation on the incivility of news comments

Yufan GuoYuhan Li, and Tian Yang, New Media & Society (2023)


Many social media affordances can affect the quality of online discourse, but such an effect remains understudied for the visibility of geolocation, which is available on most social media platforms. We looked at the event in which Weibo started to display users’ IP locations on 28 April 2022, which was supposed to reduce incivility as the deindividuation hypothesis predicted. Leveraging a natural experiment, we examined the effect of IP location visibility, with special attention to COVID-19-related news posts and location-based, uncivil name-calling. We found that displaying the IP location in the comments section increased location-based incivility, as geolocation can function as an effective cue that signals ideological affiliation and fuels conflicts between users holding different political positions on the Chinese Internet. Meanwhile, we characterized a moderating effect of audience size on this decivilizing effect. Our study suggests that diverse social media affordances can fuel group identification and facilitate intergroup behaviors.

mistaken calls from nonpurchasers have little weight in real estate TM case

Rampart Resources, Inc. v. Rampart/Wurth Holding, Inc., 2024 WL 195999, No. 23-6895 (E.D. La. Jan. 18, 2024)

The court, grappling with the Fifth Circuit’s rather inconsistent law on misdirected communications, denies a preliminary injunction (subject easily guessable by party names).

Rampart Resources was founded in 1989 in Baton Rouge. It provides land and real estate services including right-of-way acquisition, servitudes, real estate brokerage, permitting, land services, and property management across several industries including utilities, oil and gas, renewable energy, and public works. Most of its current business involves land use issues but does not involve any property management, though it previously managed residential properties for ExxonMobil and multifamily apartment units for the City of Baton Rouge. Its website states: “Our core services include right-of-way acquisition, surveying, permitting, project planning, and E&P land rights management.” Its clients are predominately corporate entities and municipalities. Nonetheless, as part of its business, Rampart Resources regularly interacts with non-client individual landowners to assist its clients in acquiring land. It has a registration for its logo, comprising stylized wording and a graphic of roads. (Note that the current website seems to have removed “resources” from the logo, which I expect is a big enough change that a new registration ought to be required.)

registered RAMPART RESOURCES mark

today's website
Rampart/Wurth Holding is a Louisiana-based real estate company offering: (1) multifamily management services for multifamily units; (2) commercial management services; and (3) single-family and small multifamily management and maintenance services, handled respectively by (1) Rampart Multifamily Management; (2) Rampart Commercial Management; and (3) Wurth Real Estate Services. Its principal clients are owners of commercial and residential real estate, but R/W also occasionally interacts with the residents and tenants of their clients. Its services include rent collection, resident/tenant placement and screening, property maintenance and repairs, property inspections, and eviction services. Its website states: “Property Management is All We Do.” It adopted the Rampart/Wurth branding in 2023. This lawsuit followed after a FedEx delivery driver told Rampart Resources’ president that “another Rampart” had recently opened in Baton Rouge and that she had mistakenly gone to Rampart/Wurth’s office instead of Rampart Resources.

Several Rampart Resources employees told the president that they had received seven telephone calls in September and October 2023 from individuals attempting to contact Rampart/Wurth. The calls generally follow the same pattern: a call asks about services that Rampart Resources doesn’t provide; the caller is informed of their mistake. One caller stated that Jefferson Lakes Apartments had given her Rampart Resources’ phone number to refund a deposit.

Rampart Resources’ mark was legally protected and arbitrary. The evidence didn’t (at this stage) support a finding of geographic descriptiveness based on the prominence of Rampart Street in New Orleans, where Rampart Resources was not based. “Unlike, say, ‘Carondelet,’ ‘Tchoupitoulas,’ or a host of other well-known New Orleans streets, ‘Rampart’ has no inherent connection to New Orleans or to any specific geographical feature.”

Strength of mark: weighed in plaintiff’s favor, though not heavily so, given the arbitrariness of the mark balanced against substantial third-party use of “Rampart,” albeit perhaps only one in real estate. “Plaintiff’s sponsorship of certain events and promotion of branded items does little on its own to counteract Defendant’s evidence of widespread usage of the key portion of Plaintiff’s mark.”

Similarity: Defendant used “a key design with its branding of Rampart Multifamily Management and Rampart Commercial Management,” along with an unchallenged key design with two Rs and no other text. The only similarity between the two marks was the term “Rampart.” That was not substantial given the other elements of the marks. The total effect of the marks was dissimilar, favoring Rampart/Wurth.

not sure this is exactly what is litigated but it's what I found

Similarity of services: Where the respective services “are noncompeting, the [possible] confusion at issue is one of sponsorship, affiliation, or connection.” “The danger of affiliation or sponsorship confusion increases when the junior user’s services are in a market that is one into which the senior user would naturally expand.” Importantly, “[t]he actual intent of the senior user to expand is not particularly probative of whether the junior user’s market is one into which the senior user would naturally expand...Consumer perception is the controlling factor.”

Rampart/Wurth pointed out that Rampart Resources didn’t even bother to register its mark for “property management.” There was only a “minor overlap” in services. Property management constituted “only a small and infrequent portion of Plaintiff’s business.” Still, an expansion into the property management market was plausible, especially since Rampart Resources had done it in the past. “Moreover, the diverse array of services offered by the Plaintiff increases the likelihood of confusion among the consuming public.” This somewhat weighed in favor of finding confusion.

Similarity between parties’ consumers/clients: There was minimal overlap. “Plaintiff’s clientele includes municipal and corporate entities, particularly those in the utilities, oil and gas, renewable energy, and public works sectors. Meanwhile, Defendant’s customers are real estate developers and owners of commercial and multifamily property.” It was plausible, nonetheless, that there was some overlap.

Advertising campaigns: Rampart Resources alleged that it promotes its services primarily through its website, its “sponsorship of prominent charity and other events,” through branded marketing items such as shirts, cups, jackets, banners, signage at its offices, fliers, and folders, and, on occasion, print advertising. Both parties are findable online, and “both parties indicated that word-of-mouth advertising among their respective customer bases is perhaps their strongest form of advertising.” This factor wasn’t particularly probative, even though reliance on word-of-mouth might “diminish the importance of the dissimilarity between the trademarks.” Without much evidence, this factor was neutral.

Intent: There was no evidence of bad faith, making this factor neutral. Failing to stop use after receiving a C&D doesn’t mean bad faith.

Actual confusion: “[T]the Fifth Circuit requires more substantial evidence of confusion where the confusion does not result in swayed purchases,” and this should be weighed against the parties’ total sales volume. “[I]solated instances of confusion about the affiliation of two companies that do not result in redirected business are not enough to sustain a finding of actual confusion.”

The evidence here was entirely anecdotal. The FedEx driver wasn’t a customer/potential customer of either party and was only “briefly confused” about the names. “Proof of actual confusion requires more. For the seven phone calls, since they weren’t about overlapping services, Rampart Resources’ employees quickly identified the mistake. Maybe they were even calling about Rampart Apartments; there was limited evidence that they conflated the defendant with the plaintiff.

Even assuming they were all trying to reach Rampart/Wurth, this wasn’t particularly weighty, given the lack of evidence that actual customers were confused or swayed into doing business with Rampart/Wurth. Rampart/Wurth averred that it hadn’t received any mistaken inquiries or questions about association. Thus, the examples showed only a “fleeting mix-up of names” by persons who weren’t direct customers of either party. Also, when weighed against the volume of business conducted by the parties, the weight of seven phone calls was lessened. “[B]oth parties operate in several states, provide complex services to sophisticated clients, and interact with a wide swath of the public. Moreover, the Defendant claims to manage over 10,000 multifamily units and over eighteen million square feet of commercial real estate space.”

After all that, the court still weighed this factor slightly in Rampart Resources’ favor.

Consumer care: High, given the expense and sophistication of the relevant transactions. Rampart Resources argued that it “routinely works with property owners who are laypersons.” But none of these allegedly “unsophisticated” persons were actual clients or customers; rather they are persons whom Rampart Resources interacts with “in order for [its] clients to successfully acquire land.” This factor would become moot if this argument were accepted, given that “nearly every company, no matter how sophisticated their customers are, necessarily interacts in some capacity with unsophisticated members of the public.” The factor was the care exercised by potential purchasers, not the care exercised by anyone who interacts with the plaintiff. (This argument would be bolstered if the court were to talk about the reason we care about potential purchasers and not random people—the potential for harm.) Plus, even if landowners are unsophisticated, there was no evidence that they had been confused, even fleetingly.

Ultimately, the plaintiff showed only a “mere possibility” of confusion, not a “probability.” “The Court finds it is unlikely that the parties’ sophisticated clientele would confuse Plaintiff’s mark with that of the Defendant, especially given the limited similarity of the marks.”

Thursday, January 18, 2024

Reading list: Trademarks in an Algorithmic World

 "Consumers used to watch ads; now, ads watch them." So says Christine Haight Farley in her intriguing new article.

Christine Haight Farley, Trademarks in an Algorithmic World, 98 Wash. L. Rev. 1123 (2023). 


 According to the sole normative foundation for trademark protection—“search costs” theory—trademarks transmit useful information to consumers, enabling an efficient marketplace. The marketplace, however, is in the midst of a fundamental change. Increasingly, retail is virtual, marketing is data-driven, and purchasing decisions are automated by AI. Predictive analytics are changing how consumers shop. Search costs theory no longer accurately describes the function of trademarks in this marketplace. Consumers now have numerous digital alternatives to trademarks that more efficiently provide them with increasingly accurate product information. Just as store shelves are disappearing from consumers’ retail experience, so are trademarks disappearing from their product search. Consumers may want to buy a product where the brand is the essential feature of the product such that the brand is the product, but they no longer need the assistance of a trademark to find the product. By reflexively continuing to protect trademarks in the name of search costs theory, courts give only lip service to consumer interests without questioning whether trademarks are fulfilling any useful information function. In many cases, trademarks may actually misinform consumers by masking the identity of the producer or its distanced relationship with the trademark owner. Without having deliberately decided to do so, trademark law is now protecting “brands as property” without any supportive normative rationale. Removing the veil of search costs theory will enable courts to consider whether trademark protection is justified in particular cases. 

prospective injunctive relief for consumers hangs on in 9th Circuit for now

Clark v. Eddie Bauer LLC, 2024 WL 177755, No. 21-35334 (9th Cir. Jan. 17, 2024)

This unpublished opinion has a dissent from Judge Bea indicating further disruptions in standing may be coming.

Clark appealed the dismissal of her putative class action complaint based on a “fake discount” theory, alleging that Eddie Bauer never sold the relevant items at the “normal” prices. After a question was certified to the Oregon Supreme Court, it ruled that

[A]n “ascertainable loss” within the meaning of the [Oregon] UTPA can, under some circumstances, flow from a consumer’s decision to purchase a product in reliance upon the retailer’s misrepresentation as to price history or comparative prices. Thus, plaintiff’s purchase price theory is a viable theory of ascertainable loss even in the absence of a showing that the seller misrepresented some characteristic or quality of the product sold.

Thus, the panel reversed Clark’s claims for money damages. Clark failed to state a claim for retrospective equitable relief because her complaint didn’t allege the absence of an adequate remedy at law for her disgorgement and restitution claims. But prospective injunctive relief was still possible because she alleged future harm (the failure to be able to rely on Eddie Bauer’s advertising). TransUnion didn’t clearly reject that circuit precedent.

Judge Bea dissented on the prospective relief part, reasoning that Clark hadn’t identified a sufficiently close common-law or historical analogue for her asserted injury. Inability to trust Eddie Bauer wasn’t enough. The closest historical analogue was misrepresentation, but “[f]or centuries, misrepresentation torts have required a showing of justifiable reliance and actual damages.” (Just imagine if courts treated trademark harm theories this way!) And Clark wasn’t justified in relying on Eddie Bauer’s prices because she knew the truth; plus, she didn’t have actual pecuniary damages. Prior circuit precedent relied on cases finding informational injuries sufficient for standing, which the Court has now disavowed: TransUnion said that “receipt of inaccurate information” wasn’t itself an injury where there was no duty to disclose and no resulting monetary harm.

I have to admit, I thought that TransUnion was the Supreme Court arrogating control over what constitutes an injury away from legislatures. But, once we’ve defined a good enough injury (harm from false advertising), the question of standing for injunctive relief seems to me to be a different type of question. Perhaps the Court will also ultimately ditch 9th Circuit precedent on this point, but it’s not logically required.

call for submissions: Harvard/Stanford/Yale Junior Faculty Forum, June 3-4, 2024

 Request for Submissions

Harvard/Stanford/Yale Junior Faculty Forum

June 3-4, 2024, Stanford Law School

Harvard, Stanford, and Yale Law Schools are soliciting submissions for the 2024 Harvard/Stanford/Yale Junior Faculty Forum, to be held at Stanford Law School on June 3-4, 2024. Twelve to twenty junior scholars (with one to seven years in teaching) will be chosen, through a double-blind selection process, to present their work at the Forum. A senior scholar will comment on each paper. The audience will include the participating junior faculty, senior faculty from the host institutions, and invited guests. The goal of the Forum is to promote in-depth discussion on the selected papers and more general reflections on broader methodological issues, as well as to foster a stronger sense of community among American legal scholars, particularly by strengthening ties between new and veteran professors.


TOPICS: Each year the Forum invites submissions on selected topics in public and private law, legal theory, and law and humanities topics, alternating loosely between public law and humanities subjects in one year, and private law and dispute resolution in the next. For the upcoming 2024 meeting, the topics will cover the following areas of the law:

- Antitrust

- Bankruptcy

- Civil Litigation and Dispute Resolution

- Contracts and Commercial Law

- Corporate and Securities Law

- Intellectual Property

- International Business Law

- Private Law Theory and Comparative Private Law

- Property, Estates, and Unjust Enrichment

- Taxation

- Torts

A jury of accomplished scholars will choose the papers to be presented. There is no publication commitment. Stanford Law School will pay presenters’ travel expenses, though international flights may be only partially reimbursed.QUALIFICATIONS: Authors who teach law in the U.S. in a tenured or tenure-track position and have not been teaching at either of those ranks for a total of more than seven years are eligible to submit their work. American citizens or permanent residents teaching abroad are also eligible provided that they have held a faculty position or the equivalent, including positions comparable to junior faculty positions in research institutions, for less than seven years and that they earned their last degree after 2014. We accept jointly authored submissions, but each of the coauthors must be individually eligible to participate in the Forum. Papers that will be published prior to the Forum are not eligible. There is no limit on the number of submissions by any individual author. Faculty from Harvard, Stanford, and Yale Law Schools are not eligible.

PAPER SUBMISSION PROCEDURE: Electronic submissions should be sent to Corissa Paris cparis@law.stanford with the subject line “Junior Faculty Forum.” The deadline for submissions is February 23, 2024Remove all references to the author(s) in the paper. Please include in the text of the email and also as a separate attachment a cover letter listing your name, the title of your paper, your contact email and address through June 2024, and the topic under which your paper falls. Each paper may only be considered under one topic. Any questions about the submission procedure should be directed to Prof. Norman W. Spaulding, nspaulding@law.stanford.edu.FURTHER INFORMATION: General iquiries concerning the Forum should be sent to Norman Spaulding (nspaulding@stanford.law.edu) at Stanford Law School, Christine Jolls, (christine.jolls@yale.edu) or Yair Listokin (yair.listokin@yale.edu) at Yale Law School, Matthew Stephenson (mstephen@law.harvard.edu) or Rebecca Tushnet (rtushnet@law.harvard.edu) at Harvard Law School.

Christine Jolls

Yair Listokin

Matthew Stephenson

Rebecca Tushnet

Norman Spaulding

Wednesday, January 17, 2024

over dissent, 6th Circuit holds that large player in fragmented market could show proximate cause under Lexmark

Campfield v. Safelite Gp., Inc., --- F.4th ----, 2024 WL 164976, Nos. 22-3204/3225 (6th Cir. Jan. 16, 2024)

Over a dissent in relevant part, the court revived plaintiff Ultra Bond’s Lanham Act claim relating to vehicle glass repair and replacement (VGRR). Safelite provides windshield repair and replacement services, while Ultra Bond supplies proprietary bonding resin to repair windshield cracks. Ultra Bond alleged that Safelite violated the Lanham Act by falsely advertising that windshield cracks longer than six inches could not be safely repaired and instead required replacement of the entire windshield. Safelite counterclaimed that Ultra Bond stole trade secrets. The district court granted summary judgment against all claims, finding no valid causes of action. I won’t discuss the trade secret issues, though the court of appeals revived some as not time-barred.

Safelite is the VGRR market leader: in 2016, it had 35.4% of the market; its closest competitor had just 3%. Safelite won’t repair windshield cracks that are longer than six inches (“long cracks”). “Windshield replacement is where Safelite makes its money, while its repair business operates at break-even or at a loss. Ultra Bond makes patented products for vehicle glass repairs, specifically for long cracks, and performs those repairs; it accounts for over 50% of national long-crack repair product sales.

Safelite promoted its policy of repairing only cracks six inches or shorter under a marketing campaign of “the dollar-bill rule”—if the crack is shorter than the length of a dollar bill (approximately six inches), Safelite can repair it. In 2007, the American National Standards Institute (ANSI), in a process that included both Safelite and Ultra Bond, conducted a safety study and concluded that cracks up to fourteen inches could be safely repaired without requiring windshield replacement. It set that as the best practice nationally, and Safelite voted to support that standard, but continued to market the “dollar-bill rule” as the safety standard for windshield repairs and continued to tell consumers that cracks longer than six inches require windshield replacement.

The vast majority of Safelite’s sales come from insurance reimbursement. “And while insurance companies ultimately set the standards for what kinds of damage it will cover, Safelite knows that its dominant market position meant that it can set the standard for insurance companies.” It told its insurance company clients that crack repair could not be safely performed on cracks longer than six inches. “And although it told insurance companies otherwise, Safelite never conducted its own technical study on the safety of long crack repairs to support its assertions.” (Monopoly power has many defects; here the anti-innovation face of monopoly also appears as misleading advertising.) “Safelite made these statements to insurance companies knowing that its insurance clients’ policyholders would choose long crack repair if it were covered,” and it admitted that “its insistence on setting as strict a crack repair standard as possible was tied to protecting its higher-margin windshield replacement business.”

The district court found that Safelite’s statements to insurers and directly to customers counted as commercial advertising or promotion, but that statements made by Safelite through ghostwritten insurance brochures and to customers purely in its capacity as a third-party administrator for insurers did not. It found that there was no genuine issue of material fact on whether the false advertising harmed Ultra Bond, and that laches partly barred the claims because Ultra Bond delayed nearly two decades before suing.

Faced with this story, the court of appeals found that Ultra Bond’s Lanham Act claim should have survived Safelite’s motion for summary judgment. Laches “bars only recovery of pre-filing damages; it does not prevent Ultra Bond from obtaining injunctive relief or post-filing damages.”

Commercial advertising or promotion: The court of appeals upheld the ruling that Safelite’s statements in ghostwritten brochures and statements in its capacity as a third-party administrator for insurance companies weren’t commercial advertising or promotion. They didn’t have any indication that they were made with “the purpose of influencing customers to buy the defendant’s goods and services.” Instead, it was the statements to insurance companies and agents that led insurance companies to set their standards, which were then set forth in the brochures and statements made by third-party administrators. “Thus, Safelite, when functioning as a [third-party administrator] or providing ghostwritten informational brochures to insurance companies, was simply acting on the success of its allegedly misleading or false earlier statements.”

Causation: Under Lexmark, proximate cause can be alleged by “a supplier of a company’s direct competitor where the decreased demand caused by false advertising directly harms the supplier.” Proximate cause doesn’t require any one specific fact pattern or theory. “In this case, the structure of the market suggests that there is unlikely to be a more directly injured commercial victim than Ultra Bond. (Safelite’s direct competitors are VGRR businesses, often small shops, that provide both crack repair and windshield replacement, so false statements that favor one service over the other would not necessarily harm them.)”

Consumer affidavits and expert evidence also supported the causal relationship between Safelite’s statements and decreased demand for Ultra Bond products:

First, nine commercial customers stated that they have experience with customers hearing from Safelite that long crack repair is not safe, educating those individuals that such repair is safe, and having those individuals choose long crack repair, which these customers perform using Ultra Bond products. Second, when misleading ads regarding crack repair were ordered to be removed from the marketplace in New Zealand, Ultra Bond’s direct sales and distribution sales to the country doubled. Third, Ultra Bond’s second expert … conducted a consumer survey and estimated that 24.5% to 30.6% of respondents who replaced windshields would have had them repaired but-for Safelite’s allegedly false statements.

Safelite argued that the declarations from commercial customers were conclusory and repeated claims about demand for Ultra Bond with only slight variations across the declarations: “[I]f customer demand for long crack repair were to increase as a result of customers being informed that long crack repairs can be safely done ... up to 14 inches, I would most certainly have to compete for this increased customer demand by buying more ... Ultra Bond, Inc. products[.]” But this statement wasn’t presented alone: the VGRR shop owners “explain how they have consistently met customers who learned from Safelite that their long cracks could not be repaired. Some shop owners have successfully reeducated customers and completed a long-crack repair using Ultra Bond products, but others have detailed how they have lost customers who called their insurance, were directed to Safelite as the TPA, and were told that long crack repair is unsafe or that the windshield must be replaced.”

This created a genuine issue of material fact on injury causation. “While Lexmark itself involved an alleged 1:1 ratio between sales gained by the defendant and sales lost by the plaintiff, it does not hold that § 1125(a) requires such a ratio in order to establish causation.” Plus, the New Zealand evidence was “additional evidence” that would allow a jury to find causation. Because reasonable minds may differ “as to the foreseeability of a particular risk or the character of an intervening cause, the question is one for submission to the jury under proper instructions as to proximate cause.”

The court also upheld the dismissal of Safelite’s unfair competition claim based on statements that, e.g., insurance companies “dupe[ ]” customers “into paying $350 for a $20 windshield,” and that Safelite’s repair tool was intentionally imperfect “as there is no way they could not know when it appears to not work on two out of three repairs.” But Safelite didn’t show falsity, or sales diversion.

Judge Bush dissented only on Lanham Act causation and would have found no proximate cause. The dissent said that proximate cause usually only allows the most direct victim to sue, and that Lexmark created a “narrow” exception where there was a one-to-one decrease in sales for every increase for the false advertiser. (Lexmark reasoned that the victims of false advertising are always harmed by third parties withholding their business because of the false advertising, and thus proximate cause encompasses their injuries—which seems exactly the scenario here.)

Judge Bush wasn’t willing to attribute customer beliefs about the safety of long crack repair to Safelite’s advertising, making the declarations speculative. It’s always possible to linguistically extend a causal chain, and Judge Bush did:

One must assume first that there are customers with long cracks in their windshields who would seek to get them repaired rather than replaced but decided not to because of Safelite’s statements; second, that those customers would choose one of Ultra Bond’s commercial customers for repair services; and third, that this untapped customer base is so substantial that commercial customers, who ostensibly use Ultra Bond for repairs under six inches, would have to buy additional product from Ultra Bond. And, for the injury caused by Safelite’s statements to insurers, Ultra Bond’s theory additionally requires one to accept without proof that, absent Safelite’s statements, the insurers would opt to change their policies to cover repair for cracks longer than six inches for their customers. This extended inferential chain is a far cry from the “automatic” injury at issue in Lexmark.

Also, most of the declarants state that long crack repair is only a small part of their business, so the causal chain also requires that Ultra Bond’s direct customers could and would have absorbed their customers’ added demand for long crack repairs.

Ultra Bond’s own expert explained that the crack repair industry is “highly fragmented,” which “makes assessing the particular impact of Safelite’s actions to Ultra Bond difficult,” especially since Ultra Bond’s products can be used to repair both long cracks and cracks under six inches. The New Zealand evidence was only “a bare assertion from Ultra Bond’s owner to [an] expert and, more importantly, does not translate into a triable issue that Safelite’s advertisements in the United States ‘more or less automatically’ cause an injury to Ultra Bond as in Lexmark.” (Seems to me that the empirical evidence takes the place of logic in showing injury, and that logic is not the only or indeed, given judicial fallibility, the best way of showing injury.) Thus, Ultra Bond wasn’t a “direct victim.”

The real weakness of the dissent, it seems to me, is the “if not them, then who?” question. Maybe insurance companies, but given standard principal/agent problems, they don’t necessarily have the right incentives either. Ultra Bond seems like a good candidate!

Tuesday, January 16, 2024

after experts excluded, slack fill class action fails

Krause-Pettai v. Unilever United States, Inc., --- F.Supp.3d ----, 2023 WL 6429805, No. 20-cv-1672-AGS-BLM (S.D. Cal. Sept. 30, 2023)

This case is about “nonfunctional slack fill,” or useless empty space, inside underarm-deodorant sticks. The court rejected the claim on summary judgment despite not finding it completely preempted.

Both the FDCA and California’s Sherman Food, Drug, and Cosmetic Law set the same baseline requirements for drugs and cosmetics: an item is “misbranded” if “its labeling is false or misleading in any particular” or if its “container” is “filled as to be misleading.” Because the Sherman Law’s standard “is identical to” the FDCA’s, it is not preempted. But the California Fair Packaging and Labeling Act (CFPLA) is explicitly preempted because it states that any opaque container is “misleading” if “it contains nonfunctional slack fill,” with some exceptions. Such a per se rule was not identical to the FDCA. That is, a state legislature can’t determine that something within the FDCA’s scope is misleading as a matter of law.

But the preemption wasn’t complete. The FDA’s failure “to issue specific regulations” about nonfunctional slack fill does not mean manufacturers may make cosmetics with “any” amount of it, even refrigerator-sized deodorant sticks that are 99% empty. “[M]ere deliberate agency inaction—an agency decision not to regulate an issue—will not alone preempt state law.”

On to the merits. The court excluded plaintiffs’ experts. The expert on product packaging didn’t adequately explain his testing methodology or resolve apparent contradictions in his reports about how many different products he tested; he didn’t have written records of what he did.

As for the consumer perception expert, plaintiffs offered his opinions that: (1) consumers “spend limited time examining package labeling information” and generally “assume that larger packages contain a larger quantity of a product”; (2) “[n]et weight labeling information on product packages is rarely examined (or understood) by consumers”; and (3) due to these consumer tendencies and “general unfamiliarity with the concept of slack fill, the relevant Unilever product package features suggest” that Unilever consumers got “less product than they might have anticipated.” But he didn’t seem familiar with the facts; plaintiffs’ central allegation was that Unilever’s sticks were in larger packaging than competitors’ with the same net weight, but he claimed that Unilever’s products “are roughly the same size (or larger) than those used by its competitors,” but “contain less actual product.” He also apparently never examined the deodorant and antiperspirant market, focusing his research mainly on food and beverages; his two studies of grocery-shopping habits that looked slightly beyond that didn’t provide much help, because one of them lumped everything together, while, in the other, the two non-food items—shampoo and toothpaste—had the longest average selection times, with shoppers spending twice as long choosing shampoo as bananas. “If anything, these results suggest that food-buying habits don’t apply equally to other commodities.” Thus, the adequacy of his data was concerning, and plaintiffs didn’t explain why his results could be generalized to deodorant buyers.

With the experts gone, plaintiffs’ deposition testimony that they were deceived was insufficient to show that reasonable consumers were likely to be deceived, because they were “unaccompanied by other pertinent evidence of deception. By contrast, the testimony of even a single individual may suffice if paired with patently false marketing or relevant extrinsic evidence.” Here, each product was labeled with its actual net weight, which could be used for value comparisons, and there were no relevant surveys.

“With so little positive proof, plaintiffs cannot make their case. And that’s before taking stock of the countervailing evidence that undercuts their theory of deception.” First, there was no comparative evidence to corroborate plaintiffs’ claims that Unilever is an outlier and that its competitors suffer lost sales due to their more aboveboard packaging. Unilever “introduced compelling comparative evidence” that its “Dove and Degree sticks are generally in line with competing products.” Second, Unilever’s evidence that “from 2016 to 2022, there were zero complaints from California consumers concerning the empty space in the products at issue” was “highly relevant,” and it wasn’t hard to discover given that all four plaintiffs became suspicious because Unilever’s sticks seemed top-heavy. Third, Unilever’s expert testified that any slack fill in the sticks at issue was functional.

new article w/Mark Lemley: First Amendment Neglect in SCOTUS IP Cases

First Amendment Neglect in Supreme Court Intellectual Property Cases

Mark A. Lemley & Rebecca Tushnet (forthcoming, Supreme Court Review)


The Supreme Court decided two cases of central importance to free speech during the 2022 term – in both cases without addressing the First Amendment implications. In Andy Warhol Foundation v. Goldsmith, the Court upheld a ruling that Andy Warhol’s reworkings of Lynn Goldsmith’s photograph of the artist Prince into highly stylized silkscreens and drawings were not transformative, and thus were unfair, at least when images of the artworks were licensed to illustrate articles about Prince. In Jack Daniel’s v. VIP Products, the court found that a parody dog toy in the general shape of a Jack Daniel’s bottle, with the label “Bad Spaniels,” deserved no special protection for its parody against Jack Daniel’s trademark claim. The Court reached these results using ideas about the lesser status of profitable speech that it flatly rejected in other cases the same term, and with rationales that seem directly at odds with its First Amendment jurisprudence.

 In this article, we show that the Court’s decisions cannot be reconciled with its approach to any other area of speech, and that they are already having pernicious effects in the lower courts. We consider some possible explanations for the inconsistency: the possibility that the Court just doesn’t see First Amendment issues in IP cases; the possibility that a political realignment has left conservative justices less enchanted with speech in the marketplace; and the possibility that this is part of a broader trend away from holding courts to the same constitutional standard as the other branches of government, combined with statutes that leave room for substantial judicial discretion in individual cases. Whatever the explanation or explanations, the decisions in Warhol and Jack Daniel’s to cut back dramatically on judicially-created speech-protective rules may have the ironic effect of forcing the Court to confront directly the constitutional fragility of much modern IP law.


Monday, January 15, 2024

Tea Rose flour ads

 Thanks, Library of Congress! My search for depictions of the dueling Tea Rose flours in Hanover Star Milling Co. v. Metcalf, 240 U.S. 403 (1916), had previously been futile, but not any more.

Friday, January 12, 2024

"carbon neutral" plausibly misleading because consumers don't understand it

Dorris v. Danone Waters, 2024 WL 112843, No. 22 Civ. 8717 (NSR) (S.D.N.Y. Jan. 10, 2024)

Plaintiffs alleged that advertising Evian as “carbon neutral” violated the consumer protection statutes of New York, Massachusetts, and California, and constituted breach of express and implied warranties, unjust enrichment, and fraud. Most of the claims survived, though the NY and breach of implied warranty claims failed.

Carbon Trust "Carbon neutral" logo
back of bottle with logo highlighted

multipack packaging with logo highlighted

Plaintiffs alleged that consumers are willing to pay more for environmentally sustainable products. Carbon neutral” is technically defined as “having or resulting in no net addition of carbon dioxide to the atmosphere.” Plaintiffs alleged that manufacturing the products still causes CO2 release, and even if “carbon neutral” referred to offsetting emissions and complying with the Carbon Trust standard, Danone fails to disclose “how it calculates its carbon neutrality, the meaning of the Carbon Trust standard and how Defendant complies to that standard, and whether the standards themselves are ‘carbon neutral’ in that any pollution output is truly offset by other projects.”

NY: None of the plaintiffs alleged that they were deceived in New York, as required by the law.

Danone argued that (1) no reasonable consumer would understand carbon neutral to mean the Product emits no carbon dioxide; (2) Defendant accurately represented that Carbon Trust certified the Product “carbon neutral”; and (3) Plaintiffs cannot challenge the Carbon Trust certification as false or misleading. The court could not resolve these issues on a motion to dismiss.

First, Danone argued, no reasonable consumer could reasonably believe that Evian is transported from their factories in the French Alps to California and Massachusetts without emitting any carbon at all. Indeed, it continued, “(1) no carbon zero products exit, (2) the dictionary definition of ‘carbon neutral’ describes the use of offsets to balance emissions, and (3) the Product’s website explains evian® water’s approach to reducing and offsetting carbon emissions.”

The court began with Merriam-Webster’s definitions, “both of which lack specificity and may be difficult to comprehend”: carbon neutral means (1) “having or resulting in no net addition of carbon dioxide to the atmosphere” or (2) “counterbalancing the emission of carbon dioxide with carbon offsets.” The term “carbon neutral” is

more technical and scientific, unfamiliar to and easily misunderstood by the reasonable consumer. Consumers thus may reasonably become confused with the term “carbon neutral” if it has not previously been explicitly defined for them—as in, before seeing it on the Product’s label. It is plausible then that the ambiguous term “carbon neutral,” a technical word not within an average consumer’s common parlance and carrying multiple meanings, could mislead a reasonable consumer.

While “carbon neutral” “may be understood by manufacturers, distributors, and other entities within the industry, the common consumer may attach a layperson’s understanding to the term.” What reasonable consumers would think was for a factfinder.  

The FTC’s Green Guides supported this conclusion. Massachusetts Chapter 93A incorporates FTC regulations, and the Green Guides warn:

Unqualified general environmental benefit claims are difficult to interpret and likely convey a wide range of meanings. In many cases, such claims likely convey that the product, package, or service has specific and far-reaching environmental benefits and may convey that the item or service has no negative environmental impact. Because it is highly unlikely that marketers can substantiate all reasonable interpretations of these claims, marketers should not make unqualified general environmental benefit claims.

The complaint plausibly alleged that the average American consumer does not know the term’s technical definition, as “nearly sixty percent ... do not understand what the term ‘carbon neutral’ means” and that “reasonable consumers often mistake ‘carbon neutral’ for ‘carbon zero or carbon free,’ ” even if “carbon zero” products do not currently exist. As for the explanations on Danone’s website, which links to the Carbon Trust website, “[r]easonable consumers are not expected to look beyond misleading representations on the front of the container” or “to do research.” The requisite research was not “minimal,” either—“reasonable consumers should not be expected to visit two separate websites and read several pages to fully understand the meaning of ‘carbon neutral’ and ‘certified by Carbon Trust.’”

Danone argued that, if consumers don’t understand “carbon neutral,” they couldn’t be misled; the survey cited by plaintiffs showed that lots of people didn’t know exactly what it meant. “[A]lthough respondents may not understand the precise definition of the term, they could still be persuaded by the environmentally friendly sounding representation.” Indeed, such consumers “are exactly the type of consumers who would be reasonably swayed by misleading marketing practices.”

Monday, January 08, 2024

always-available, effortless discount plausibly makes higher "regular" price misleading

Vizcarra v. Michaels Stores, Inc., --- F.Supp.3d ----, 2024 WL 64747, No. 23-cv-00468-PCP (N.D. Cal. Jan. 5, 2024)

Vizcarra alleged that Michaels deceptively advertises its products as discounted when in fact they are always available for at least 20% less than the purported “regular” price. The court dismissed her unjust enrichment claim but otherwise allowed her consumer protection claims to proceed.

Allegations: On Michaels.com, Michaels’ entire inventory is always available at a discount of at least 20% off of the “regular” listed prices, and these discounts are prominent in stores and on its webpages. E.g., in January 2023, in search and product pages, the text “Save 20% with code 22MADEBYYOU” appeared in red text immediately below list prices. At least one sitewide discount code offering at least 20% off of all merchandise is always offered (screenshots covered Jan. 2021-Feb. 2023). Similar discounts are offered in stores via coupons that are available both online and in stores. Thus, Michaels’ products are always available—in store and online—for at least 20% off the prices Michaels characterizes as “regular.” For Vizcarra’s in-store purchase, her receipt indicated she saved $11.65.

California’s FAL specifies: “No price shall be advertised as a former price of any advertised thing, unless the alleged former price was the prevailing market price as above defined within three months next immediately preceding the publication of the advertisement or unless the date when the alleged former price did prevail is clearly, exactly and conspicuously stated in the advertisement.”

Michaels argued that it “advertises current—not former—prices,” so it wasn’t covered by the former price rule. But current prices can be former prices—or, more to the point, consumers can receive the message that they are former prices, and that the product was previously available only at the non-discounted price. (And, since reference prices matter a lot to consumers, there’s an obvious incentive to advertise prices no consumer ever actually had to pay, which is where the restrictions on discount advertising have their genesis.) The question wasn’t whether all discount codes or coupons were covered “former price” advertising, but whether it’s possible to offer such a scheme in a way that presents the higher current price as a former price, and whether Michaels did so. By contrast, club memberships or “[c]oupons offered in exchange for receiving something from a consumer, like sharing personal information or repeat purchases, are clearly distinguishable.” The law covered situations where “former” was implicitly conveyed to a reasonable consumer, rather than only explicitly. As another court wrote, “the requirement that a consumer enter a coupon code to obtain the advertised discount is merely a routine, procedural step in the purchase transaction and is not material.”

Drawing all inferences in Vizcarra’s favor, the court found it plausible that “most consumers, when confronted with two prices including a lower price that can be obtained with negligible additional effort, will opt for the lower price.” That would mean that the prevailing market price was the lower, discounted price, whether the products were exclusive to Michaels or otherwise.   

The other consumer protection claims also survived because this conduct was plausibly misleading to consumers, as did claims for intentional misrepresentation and breach of warranty.

Friday, January 05, 2024

where cross-examination exposes lack of TM confusion, out-of-court confusion "evidence" becomes less credible

Florida Virtual School v. K12, Inc., 2024 WL 22039, No: 6:20-cv-2354-GAP-EJK (M.D. Fla. Jan. 2, 2024)

Some interesting comments on when individual instances of “confusion” don’t count, as well as their relevance to evaluating out-of-court social media etc. statements as evidence of confusion.

In 2011, Plaintiff sued defendants for using the marks “Florida Virtual Academy/Program” and the associated acronyms, “FLVA/P.” The parties settled in 2015; defendants agreed to cease their use of those marks and to avoid the words “Florida” and “Virtual” together in a mark. The agreement listed four examples of prohibited marks as well as acceptable marks, but did not require defendants to use the acceptable examples. In 2019, defendants launched the “Florida Online School,” abbreviated “FLOS.” About a year later, plaintiff demanded that defendants stop using that too, which they did, but plaintiff sued anyway, seeking (by the time of trial) an injunction and disgorgement.

Plaintiff has seven registered trademarks involving Florida Virtual School or FLVS; two of the registrations are incontestable.

registrations with graphical elements

The “Florida Online School” mark included an image of a Florida panther.

blurry panther mark

Strength: Plaintiff’s marks at issue were all highly descriptive. Acknowledging that the Eleventh Circuit has said that circuit precedent is probably wrong about this, the court accorded artificial strength to the incontestable ones, in the form of secondary meaning, but then—also following circuit precedent—suggested that this presumption of strength was easily rebutted with real marketplace evidence. And defendants did rebut it—all the marks were commercially weak. (Also, the incontestability was procured under “dubious circumstances”; although defendants’ cancellation counterclaim was precluded by the prior settlement agreement, the plaintiff admitted that it made at least one material misrepresentation to the PTO in its registration application, and there was evidence of further misrepresentations “in its effort to establish secondary meaning and overcome the initial rejection of its application for these marks.”)

The evidence of weakness was substantial:

Apart from the geographic, descriptive nature of its marks, Plaintiff’s own internal materials tend to illustrate their inherent weaknesses. While multiple witnesses testified as to Plaintiff’s significant marketing and advertising efforts, that alone is not indicative of strength. Plaintiff’s Director of Marketing … testified that changing a logo and using it in different ways “can dilute a brand” in the same breath as she acknowledged that, in only twenty-five years of existence, Plaintiff has changed its logo six times. [Another witness] discussed a nearly $5 million effort to rebrand Plaintiff’s global operations as recently as 2020.

… Plaintiff argues that “Florida consumers consistently recognize FLVS significantly more than they recognize K12 and other[s].” However, upon review, the survey Plaintiff cites shows that its superiority is marginal—often within ten percentage points ….

In a 2018 survey of parents with school-aged children that Plaintiff commissioned while researching its brand effectiveness, its mark had only 15% more awareness than Defendants’ mark and, moreover, only 30% of respondents recognized Plaintiff’s brand, even when prompted. Similarly, in a 2020 commissioned survey, without prompting, only 1% of respondents could name Plaintiff as an online education provider. While Plaintiff’s full-name marks garnered around 36% awareness among prospective families, the acronym marks had less than 15% awareness among prospective and current families. Indeed, in an internal marketing presentation from January 4, 2022, Plaintiff itself used words like, “plain…bored…uninspired…nondescript…[and] sterile” to describe the brand identity of its acronyms. This is strong evidence of the commercial weakness of these marks.

Also, Florida county school districts often incorporate the phrase “VIRTUAL SCHOOL” into the brand for their online educational offerings, in partnership with both plaintiff and defendant. A plaintiff witness testified that these school districts are “using their county name[s] to distinguish [themselves] from us or anyone else,” e.g., the Broward Virtual School, which weakened the mark. “Though use of a mark by licensees supports its strength, the use of Plaintiff’s hybrid marks throughout Florida’s 67 counties to cover services that are actually provided by both Plaintiff and Defendants weakens Plaintiff’s marks significantly.” The court thought it was ridiculous to argue that “Virtual” & “School” together were not infringing, while “Florida” & “School” together were. “Though some of these third-party users operate as Plaintiff’s franchise partners, the fact that these franchise relationships also allow Defendants to provide substantial services (e.g. an entire elementary school program) under the same marks significantly undercuts the strength of Plaintiff’s marks.”

Plaintiff also failed to provide evidence of actual confusion. There’s a lot of general confusion about online education, and about the fact that, in order to provide their services, plaintiffs have to associate with public school districts, which are overseen by the Florida DOE, of which defendant is a subagency.

But trademark confusion? No. The only live evidence of actual confusion was the testimony of two parents who mistakenly enrolled their children with Defendants’ Florida Online School. But the initial testimony that the names were confusing “readily disintegrated under live cross examination.” One witness “testified on multiple occasions that her confusion stemmed from her misconception that there was only one online education provider available to her.” And it was plaintiff’s own delay in assigning a teacher that led her to switch back to a brick and mortar school. The other, who was vision-impaired, testified that she enrolled one of her sons with defendants’ Florida Online School, “not realizing the schedule, and then immediately realized that wasn’t the right thing for us as a family and put him into [Plaintiff’s school].” Her “repetitive emphasis on her son’s need for a flexible schedule as the reason for unenrolling him undercut[] the relevance of that guided testimony.”

Given how the supposed evidence of confusion deteriorated under cross-examination, the court also mostly rejected twenty-one emails from the parties’ employees, parents, students, and other third parties. (Portions were likely admissible as party statements.) For purposes of trial, the emails were not trustworthy evidence of confusion, even as state of mind: “Was the author really confused? What was the nature of the confusion? Who caused the confusion? Was Plaintiff harmed by the confusion? The answers to these questions require cross-examination.” The court pointed to the fact that the two live witnesses’ emails could support an inference of confusion, but upon cross-examination, “it became clear that the source of confusion was not Defendants’ name.” Thus, these out-of-court statements weren’t reliable.  “[I]f anything, they support the fact that online educational service providers exist in a muddled marketplace replete with generically and descriptively named participants.”

In this context—seeking millions of dollars of disgorgement—the absence of confusion evidence, including survey evidence, weighed heavily against finding likely confusion, especially given that potentially millions of consumers were exposed to the alleged infringement over two years, which was plenty of time for any such evidence to emerge.

Intent: The Eleventh Circuit says: “If it can be shown that a defendant adopted a plaintiff’s mark with the intention of deriving a benefit from the plaintiff’s business reputation, this fact alone may be enough to justify the inference that there is confusing similarity.” This requires a “conscious intent to capitalize on [its] business reputation, w[ere] intentionally blind, or otherwise manifested improper intent in adopting [the Florida Online School] name and acronym.”

Plaintiff’s “strongest evidence” of intent was that defendants continue to use the words, “Florida online school,” on their website — “contending, incredibly, that any use of those words in any context constitutes trademark usage. It was not enough that Defendants completely rebranded their entire online school; Plaintiff now insists that Defendants—who operate an online school in Florida—must not use those words anywhere on their websites.” No. Although defendants did change some website language during trial, the court’s review found that this just reflected “ill- informed, non-attorney employees scrambling to avoid the swinging arms of a bully rather than any kind of concerted effort to alter evidence.” Non-trademark usage of the words “Florida online school” was ultimately irrelevant to the analysis.

“The complete dearth of evidence of any ill intent on behalf of Defendants is enhanced by their testimony that the Florida Online School name was never of particular importance …—they simply chose a descriptive name that was not on the list of marks prohibited by the Settlement Agreement.” And, when plaintiff complained in August of 2020—one year after the school had begun operations as Florida Online School—defendants began “instantly” working with its contracted school district to change the name. “The fact that it took in excess of one year to accomplish a complete rebranding of a school name, including updating email addresses and all school literature, is not unreasonable.”

Ultimately, there was no evidence of intent to trade on plaintiff’s good will, weighing strongly against infringement.

Mark similarity: Textual similarity in a crowded field with a limited number of descriptive terms wasn’t probative, and the visual elements were very distinct.

Consumer sophistication: “Though the expense associated with a college education is largely absent from these providers, the nature and importance of a parent’s choice of where to educate their child is comparable to that decision. The fact that Plaintiff presented two individuals who experienced confusion with this marketplace—one of whom freely admitted that she undertook no research while the other plainly changed her mind based on scheduling concerns unrelated to any mark confusion—does not suggest that the thousands of other customers of these parties are not sophisticated.”

Similarity of services slightly favored plaintiff; the trade channels were similar, but defendants’ customers weren’t individual parents and students, but rather a school district. And they targeted school districts in their marketing. Plaintiff focuses on parents and students. Defendants market only their K12 brand nationally. “Indeed, any parent who stumbles onto the Florida Online School website would be unable to enroll and would be directed to the main K12 website if they were interested in enrollment.”

Balancing: Defendants win.