taking control: Honeywell TM, ad claims lose, (c) continue

Honeywell International Inc. v. ICM Controls Corp., No. 11–569, 2014 WL 4248434 (D. Minn. Aug. 27, 2014)

Honeywell sued ICM for patent infringement, copyright infringement, violation of the Lanham Act, and violation of state law.  (I will ignore the patent issues.)  The claims relate to ICM’s sales of allegedly “direct knockoffs” of four Honeywell combustion control products used for controlling heating appliances.  The products are two “oil primary controls,” which control the fuel oil for an oil-based furnace, water heater, or boiler; and gas ignition controls, ditto for gas.  Honeywell sells oil primary controls in residential and light commercial markets as OEM and replacement units; the gas ignition controls are only replacement units.

ICM received summary judgment on Honeywell’s false advertising claim, which challenged ICM’s advertising of its combustion control products as “Made in the USA.”  ICM’s expert reports indicated that a significant part of the manufacturing of the relevant products—including the final assembly—occurs at a facility in Syracuse, NY. So does conceptualization, design, testing, and other production activities also occur there. Some foreign components are incorporated into the products, but their value ranged from between 10.4 and 23.8 %, depending on the product.  The foreign materials cost ranged between 33.6 and 80.2 %, and with labor and other costs added in, ICM’s expert placed the foreign content of various ICM products at between 19.4 and 52.9 %.  Honeywell argued that “Made in the USA” (along with  “100% American Made” and “American Made”) was a false designation of origin.

“The standard for proving literal falsity is rigorous” and “only an unambiguous message can be literally false.” (Ambiguity, the court determined, was a question of law, whereas falsity was a question of fact; falsity by necessary implication wasn’t at issue.)  In the context of this case, the court found “Made in the USA” not unambiguous enough to be literally false.  The FTC has guidelines for US origin claims, but Lanham Act plaintiffs have to show literal falsity/misleadingness, not merely violation of FTC guidelines, so the court wasn’t going to look at whether the products violated the FTC guidelines.  (This finesses an extremely difficult question: when the government sets a standard, aren’t consumers likely to believe that producers are complying with the government standard, whatever it is, even if they don’t know the details of the standard?  They could then be deceived by an idiosyncratic definition.)  Plus, the FTC’s own policy statement showed that “Made in the USA” was not unambiguous enough to support a literal falsity theory.  The statement requires, at a minimum, that “the final assembly or processing of the product must take place in the United States,” because that’s so important to consumers.  But beyond that threshold the FTC considers other factors, “including but not limited to the portion of the product’s total manufacturing costs that are attributable to U.S. parts and processing; and how far removed from the finished product any foreign content is.”  The FTC doesn’t have a fixed proportion, but says it’s evaluated on a case by case basis. Thus the phrase was not sufficiently unambiguous to find literal falsity here, where there was no dispute that final processing and assembly occurred in the US.  (Plus, other laws use different definitions to identify a country of origin.)

Without literal falsity, Honeywell lost because it couldn’t show misleadingness; it didn’t offer consumer or market research showing the relevant consumers’ understanding of the phrase. Honeywell offered the testimony of a professor of marketing, who opined that “contractors,” who the parties treat as the relevant consumers, “are likely to believe” that a “Made in the USA” claim “implies that 100% of the product and its components are manufactured” in the United States.  But Honeywell didn’t show that the professor conducted any research on actual contractors or had any relevant experience with them.  That wasn’t enough. 

Nor could Honeywell rely on FTC studies about general consumers’ perception of the phrase, performed in 1991 and 1995. Those studies “do not address what a contractor in the present-day understands by the phrase and whether that understanding is inconsistent with ICM’s use of it.”  Plus, the FTC itself acknowledged that many consumers just had a general sense of what the phrase “Made in the USA” means rather than a highly specific view of how to evaluate costs, processing, etc.  Beyond the final processing requirement, “a customer’s understanding of the accuracy of a U.S. origin claim would depend on the product and the type of content that was foreign sourced”—and Honeywell lacked evidence about that.

ICM also prevailed against Honeywell’s trade dress infringement claims.  Honeywell claimed an interest in a combination of identified elements.  For one product, and representatively, they were: 

• a gray case;

• a raised and ridged circular bezel formed out of the case, located on the left side of the “face” of the product;

• a red circular reset button set inside the raised/ridged bezel;

• a black-and-white label with an orange warning bar immediately to the right of the red button;

• an LED indicator light nestled in the housing and located between the thermostat and flame-sensor connections along the right side of the unit; and

• the location and order of the thermostat and flame sensor quick-connect terminals on the right side of the unit.

Honeywell controllers

The court began by acknowledging the striking similarity between the Honeywell and corresponding ICM products. But copying is often ok, and the Supreme Court has cautioned against overprotecting trade dress, especially product designs and configurations.  Honeywell had the burden of showing nonfunctionality. If the feature affects the cost and quality of the article, it’s functional; if the feature would put competitors at a significant non-reputation-related disadvantage, it’s functional—but competitive necessity isn’t required in non-aesthetic functionality cases.

Comparison between Honeywell and ICM

For each product, the elements identified by Honeywell were part of components on the exterior of the controllers, each of which served a function.  “[T]he case serves to protect the internal components, the reset button is used to reset the device, the raised bezel protects the button, the label provides information (including manufacturer identity) or warnings, and the LED light operates as an indicator.”  However, Honeywell argued that its trade dress covered “the gray color of the case, the red color of the reset button, and the location relative to each other of the various components on the external body of the unit”—which reduced to shape, color, and configuration. 

The court found that a reasonable jury couldn’t find that Honeywell met its burden of overcoming the presumption of functionality. Honeywell sought to show that there were other design choices and thus no competitive necessity.  But “TrafFix makes clear that treating the competitive necessity test as the sole and comprehensive gauge—without consideration of the purpose, use, cost, and quality aspects relevant to the primary inquiry—is error.”

Though Honeywell’s witnesses called the shape, color, and configuration “arbitrary,” they weren’t using it in the trademark sense, but merely meant that multiple options existed for the particular shape, color, or configuration at issue.  “Critically missing from the evidence that Honeywell puts forward in its opposition to ICM’s motion is a demonstration that its claimed trade dress is not essential to the use or purpose of the products and does not affect their cost or quality.”  Though it emphasized the feasibility of alternative designs, it did not show that those designs didn’t have cost or quality consequences compared to the Honeywell designs.  Nor did it show that the design elements were selected with an eye to identification and individuality rather than utility. This failure of proof warranted summary judgment.

The court also commented that “common experience confirms the functionality of many of the elements that make up Honeywell’s claimed trade dress. For example, although alternatives may be possible, black and gray are common choices for housings for technical equipment, especially equipment meant to be in the background, so that they do not stand out.”  The use of common use of red and orange to draw attention suggests that “their use for an important button or warning label serves a function. Common sense also suggests manufacturing, storage, and transportation efficiencies of simple shapes, such as a generally rectangular shape for the overall product or for a round button.”  Plus, some design constraints come from the locations where the controllers are mounted, the wiring needed, and the install space.  But Honeywell didn’t provide evidence showing its product configurations and shape didn’t stem from those requirements.

Other competitors' designs

Given that the controllers were “little more than a set of ‘individual functional components’ that have been “combined in a nonarbitrary manner to perform an overall function,” Honeywell couldn’t claim that the overall trade dress was nonfunctional.  Moreover, had Honeywell shown that the claimed elements didn’t affect cost or quality, it still wouldn’t have shown that control over these “basic colors and shapes, as well as their nondescript combination” didn’t put competitors at a non-reputation-related disadvantage. This wasn’t like the “eccentric” use of green-gold on a laundry press pad, “the unexpected placement of red on a shoe’s outsole that contrasts with the upper part,” or the “purposefully-striking contours of a Ferrari.” Honeywell didn’t show that the possible variations were “sufficiently great in number—especially in light of the size, installation space, and wiring constraints for controllers—that they would not significantly disadvantage competitors.”  Though Honeywell showed alternative designs from other competitors, “some do not differ so significantly as to make it clear that no concern of an infringement claim would exist.”

If the court accepted Honeywell’s configuration as valid trade dress, “the overall appearance of practically any everyday appliance or piece of generic equipment would meet the nonfunctionality test, because as long as the item is large enough or has at least a few external components, some variation in the configuration is bound to be possible.”  Other producers would be entitled to the same protection, and new entrants would quickly face an “undue burden” to steer clear of predecessors.  For example, if Honeywell could protect having “thermostat and flame sensor quick-connect terminals on the right side,” competitors would be limited to the other three sides, which would then run out.  “That burden seems especially onerous when imposed on suppliers of technical equipment for mundane applications that are purchased almost exclusively, if not entirely, for the functionality that they provide.”

Note: Honeywell also styled its infringement claim as a false advertising claim, alleging that the similar appearance created the false impression that ICM’s products were affiliated with Honeywell and violated §43(a)(1)(B).  The court didn’t think that trade dress infringement should be styled as a false advertising claim, and questioned the validity of “construing a product’s appearance as a ‘statement of fact’ that can form the basis for a claim under subsection (a)(1)(B).” Color or shape can absolutely be statements of fact—for example, that’s what it means to say that green on a soda can signifies that it contains lime/lime flavor and thus green is descriptive for lime soda—but I agree that restating the trade dress claim as one for false advertising doesn’t help.  The standard for “does this configuration constitute a statement that this product comes from Honeywell?” should be the same as “is this configuration a valid trade dress?”

Honeywell also sought to exclude the testimony of ICM experts, including that of Adam Vaczek, who would opine on the copyright infringement claims that remained in the case.  Honeywell alleged that the installation manuals and product labels of ICM’s accused products infringed copyrights.  Vaczek had over 30 years of experience in the HVAC industry as a service technician and service manager for a large contractor, with considerable exposure to installation manuals and product labels. He opined that Honeywell’s manuals and labels “contain little, if any content that is unique to Honeywell or that is not determined by the features or functions of the controls themselves. The organization of the information contained in the manuals and labels is standard industry information, common directives, fact-based communications, methods of operation and universally shared graphics. Many of the instructions and methods, use simple ways to convey factual messages.”  He concluded that the manuals’ chapter headings and organization followed a typical format; that the manuals’ content was factual, dictated by the features of the control, and used industry-standard lanuage; and that the labels too were functional and standard in the industry.  In the second part of his report, he identified differences between the parties’ works and opined that the similarities resulted from product features or industry standards.

Honeywell argued that Vaczek wasn’t a copyright expert and thus his testimony wasn’t relevant.  Substantial similarity requires both extrinsic and intrinsic analysis, the first of ideas and the second of expression.  (Pause to note how weird and wrong this is.)  Factual works may have thinner or thicker copyrights; and any copyrighted work includes unprotectable elements.  Merger and scenes a faire, for example, identify standard features that everyone is free to use.  Given this law, the court found that Vaczek’s report was not admissible in its entirety, to the extent that it went to the intrinsic test.  Expert opinion evidence may be considered in connection with the extrinsic test for substantial similarity, because it depends on objective criteria, but it can’t be used to show or rebut similarity of expression under the intrinsic test.  Vaczek’s conclusions such as “The visual impression of these two manuals is different,” “The chapters in the S8610U manual and sections in the ICM290 manual differ in length, width and layout,” and “The ICM290 and Honeywell S8610U manuals contain different content and overlapping content is presented differently,” went too far.  (Even critics of the extrinsic/intrinsic divide for expert testimony might not disagree with this conclusion as long as the factual discussion comes in; Mark Lemley’s criticisms, I think, go to other problems with the division.)

The first part of Vaczek’s report was admissible and covered most of the same facts.  That part was relevant to the strength and scope of the protected elements in Honeywell’s works, including the application of merger and scenes a faire.  Vaczek’s discussion of information that needed to be in the manuals and labels in light of the products’ functions and industry expectations was “especially relevant as some of the registrations of the items at issue reflect that the copyright interest is in the compilation of the information.” Vaczek’s experience gave him specialized knowledge about the functional, technical, and other requirements for content included on the works, which would help the jury. However, Vaczek was not a linguist and didn’t have experience drafting manuals or labels, so his claims that there were “limited ways” to say something, the words used are “simple ways to convey the message,” or the language uses the “most basic words” were not appropriate opinions. The court would carefully oversee the presentation of his testimony.

The proposed expert testimony of Professor Justin Hughes didn’t fare as well, despite his impressive qualifications and experiences. The report explained, among other things, that (1) facts and data do not prevent copyrightability; (2) analysis of “concepts” is not relevant to copyrightability; (3) “uniqueness” is not relevant to copyrightability; and (4) use of common expressions and terms do not prevent copyrightability. Honeywell argued that his report merely stated his understanding of the law, which guided his analysis of Vaczek’s report, and didn’t instruct the jury on the law.  “But the analysis that Professor Hughes proffers is inextricably linked with his articulation of the nature of copyright law and his opinion about its correct application. The entire thrust of his analysis is an identification of the copyright law that should apply and an assessment that Mr. Vaczek’s opinions disregard the law or are irrelevant in light of it.”  Allowing him to testify would impermissibly invade the province of the court and the jury. Honeywell didn’t show that he had technical or industry expertise that would enable him to challenge Vaczek’s opinions about how much functional and other requirements drive the content of the manuals and labels at issue.