Courts Still Can’t Let Go of the Possibility that Individual Prices are Copyrightable–Rapaport v. Nivoda
In my previous blog post on this case, I wrote:
This lawsuit raises one of the venerable but surprisingly vexing copyright law questions: when is a price copyrightable? That might sound like a stupid question because “facts” aren’t copyrightable and a price seems like it should be characterized as a “fact.” And yet…the copyright law jurisprudence is littered with cases saying or implying that individual prices could be copyrightable (e.g., the abysmal CDN v. Kapes opinion), as wacky as that may seem. These cases have imperiled various important social activities, such as price comparisons.
This court, in contrast, treats the copyrightability of prices as the easy question I always thought it should be
Ideally, the clean lower court ruling would have finally ended the price copyrightability drama. Instead, the Second Circuit did a “well, actually….” and revives the case.
The court summarizes its views on when numbers are facts:
Where the expression takes the form of a number, “[t]he question then becomes the possible range of that number.” Determining this range allows the court to assess whether the number is an “original creation[],” deserving of copyright protection, or merely a “pre-existing fact[],” free to be copied.
What’s slowing down courts is when a number is the distillation of an estimator’s asset valuation. Valuations aren’t a precise science, so any numerical expression of a valuation isn’t a “fact” but an “opinion” and could very well be expressed as a range rather than a single number. Yet, as the lower court held, we can’t separate the estimator’s opinion from the fact that the estimator held this numerical opinion. That inseparability triggers copyright’s merger doctrine. This means anyone should be free to republish the estimator’s numerical “opinion” without copyright entanglements.
The Second Circuit, harvesting years of precedential doubt it has sowed on this point, cannot reach such a clean outcome:
Without evidence to establish how the Price List is created, however, the court lacked a basis for concluding that this idea is what is reproduced in the Price List. As pled, the Price List reflects Rapaport’s “opinion as to the market value of any diamond based on size, color, and clarity,” based on Rapaport’s “analysis and sophisticated expert industry opinion.” Rapaport does not allege, for example, that the List reflects nothing more than the average transaction prices for particular types of diamonds in a particular time span, based on an aggregation of reported sales. To the contrary, text accompanying printed versions of the Price List emphasizes that the List reflects Rapaport’s “opinion” of prices, which may differ “substantially” from “actual transaction prices.” Thus, for the limited purpose of our analysis, drawing all reasonable inferences in Rapaport’s favor based on the facts alleged in the complaint and other facts appropriately considered at this stage of litigation, we identify the “idea” at risk of merging with Rapaport’s expression, as “Rapaport’s opinion as to the value,” as opposed to the actual “market price” (meaning the price set by the market) of particular diamonds based on their size, color, and clarity….
No, this makes no sense. The process of manufacturing a number should be irrelevant to its copyrightability. Even if the valuation required the most creative process that has ever existed in human history, the estimator’s valuation becomes a fact about the estimator holding that opinion.
Still harvesting the decades-long bad precedent, the court continues:
In order to determine whether all possible expressions of a party’s opinion
regarding diamond price valuation are “substantially similar,” we must assess the possible range of these estimates. In doing so, we evaluate whether the number is better considered a “mechanical derivation[]” of a pre-existing fact, or an “original creation[],” resulting from “some minimal degree of creativity….the values included in Rapaport’s Price List, as alleged, reflect Rapaport’s “opinion” as to hypothetical products meeting certain specifications, based on Rapaport’s “analysis and sophisticated expert industry opinion.” On appeal, Rapaport emphasizes that the values included in the Price List do not express the “actual market price of any-real world diamond,” but rather its “opinion as to what should be the sales prices for different classes of hypothetical diamonds.” Perhaps most relevant, unlike the actual market price of particular futures contracts (which the settlement prices in New York Mercantile aimed to capture mostly by reference to a fixed set of trading data), the value of a diamond—like that of a used car—may depend on any number of facts beyond current trading data, such as subjective considerations of personal opinion and aesthetic preference. Rapaport alleges in its complaint that it arrives at the values included in the Price List by reference to factors beyond raw trading data, such as its “extensive market research and proprietary knowledge,” and its “decades of knowledge, investment, and industry relationships.” At this stage of the litigation, we are obliged to take them at their word. Only discovery can tell whether the sources and processes used by Rapaport in fact more closely resemble those used by NYMEX to determine settlement prices, and not (as alleged) those used by the Red Book editors for used-car valuations.
And what goal does this hack philosophizing advance? It suggests that every copyright claim over single numbers will not be resolvable on a motion to dismiss (and maybe not on summary judgment either). That’s a lot of socially wasteful litigation.
The court addresses other policy matters in wholly unsatisfactory ways:
the List attempts to value hypothetical products whose ultimate value is tied, at least in part, to subjective preferences. We are therefore not persuaded that failing to apply the [merger] doctrine will “seriously impair the policy of the copyright law that seeks to preserve free public access to ideas.”…
Nivoda does not suggest, nor does the present record indicate, that Rapaport would still have an incentive to produce the Price List if it lost its exclusive rights over that publication. The Price List “is available by paid subscription only” and, Rapaport asserts, is the product of “significant time, resources and expense,” resulting from the company’s longstanding “industry and customer relationships.” Accordingly, applying the doctrine appears unlikely, based on what we must assume to be true at this point in the proceedings, to serve copyright’s aim of encouraging the authorship of innovative works.
Feist eliminated sweat-of-the-brow justifications for copyright. Apparently the Second Circuit prefers a pre-1991 world. Oops, I’m not sure I can reference the number 1991. I might have infringed on someone’s copyright.
On remand, I hope the lower court embraces the full spectrum of ways this case should resolve in favor of the defense. Ultimately, I hope this opinion becomes another mockable tangent in courts’ decades-long quixotic quests to manufacture copyrights for single numbers.
Case Citation: Rapaport USA, Inc. v. Nivoda USA, LLC, 25-1065 (L) (2d Cir. Sept. 4, 2026). Judge Nardini, a Trump appointee, authored this opinion.
BONUS: The Construction Specifications Institute, Inc. v. Zerodocs.com, Inc., 2026 WL 2600412 (C.D. Cal. Sept. 1, 2026). The court summarizes the plaintiff’s work:
The work at issue in this copyright and trademark case—the CSI MasterFormat—is a classification system that has been, for decades, the national standard for identifying construction bids. CSI’s 2020 edition of its MasterFormat contains a taxonomy system for classifying and organizing technical information and project details in architectural, engineering, and construction projects. Specifically, each component of a project can be identified by a standardized six-digit number, which can be expanded to eight digits when greater detail is required, accompanied by a standard title that is usually a descriptive word or short phrase…
CSI’s MasterFormat Maintenance Task Team (the “MFMTT”) is tasked with conducting ongoing reviews of the MasterFormat, and the MFMTT has adopted a biennial revision process. Users of the MasterFormat can submit commentary and proposals to the MFMTT for review and approval. The MFMTT bases its approval and revision process on several broad, flexible criteria. The MFMTT draws on user feedback and proposals to make its revision and addition determinations. All proposals must pass through a two-stage process before they can be approved by the MFMTT.
The court’s description of the defendant’s activities:
Zerodocs distributed a product called “SimpleSpec” that is composed of a collection of pre-written word processing document templates. The SimpleSpec templates were created in 2017 to help users write bid proposals that comply with the MasterFormat classification system. Each SimpleSpec template is directed to one entry in CSI’s MasterFormat system.
Zerodocs used CSI’s “CSI” mark on its website to refer to CSI’s MasterFormat, but it did so without CSI’s authorization. Those references accurately reflect the source of the standards to which the Zerodocs SimpleSpec templates were written—the MasterFormat standard…
Zerodocs has adopted the structural style of MasterFormat into its Vendor Specs, and it uses CSI’s MasterFormat framework. Specifically, Zerodocs’s SimpleSpec offers at least 457 templates organized in the MasterFormat framework. CSI further identifies that, in March 2024, Zerodocs’s website displayed at least 199 SimpleSpec specifications that are identical to the 2020 edition of CSI’s MasterFormat. CSI shows that SimpleSpec contains templates that use specific number and title combinations from the MasterFormat—e.g., “03 30 00—Cast-In-Place Concrete.”
Though Zerodocs copied something, there was no copyright infringement. The “divisions, numbers, and titles of the MasterFormat are short phrases, and, therefore, they are not protected elements.”
The taxonomy isn’t copyrightable either:
CSI’s descriptions are factually concise, and they identify “specific work results and construction practices.” For example, in the section for Concrete, there is no room for creative expression because CSI must necessarily be as concise as possible and identify the specific work result—i.e., Concrete. In describing the sections and title of the taxonomy, CSI is “knuckling under” to facts about materials used or jobs performed in construction projects. Furthermore, whether a description is updated or adopted by the MFMTT’s revision process is a rote, four-step process that asks if the new description better defines the subject matter, is more useful, or is more appropriate. Those are utilitarian queries, not creative endeavors. Even if the MFMTT wanted to change a description, it prioritizes established industry terms and introduces new terms only when necessary. No “blood is shed” in the MFMTT’s revision process…
CSI assigns descriptors as concisely as possible by using words that best describe the subject.
I don’t think knuckles or blood have a lot to do with the copyrightability of taxonomies, but it might make a good Netflix series. Also, I’m not sure an author’s utilitarian aim dictates whether a work is expressive. Still, this court is at least doing better than the Rapaport court.
Zerodocs argued Baker v. Selden resolved the case. I do love topical citations in the 2020s to the nineteenth century classic, but the court points to rogue decisions from the 3rd and 7th Circuit that found taxonomies copyrightable despite the Selden case. Instead, “the Court is unpersuaded that CSI’s MasterFormat is wholesale uncopyrightable simply because it is a system. Rather, as previously discussed, the Court concludes that CSI’s MasterFormat did not meet the threshold for copyrightability of a taxonomy, as guided by American Dental and Southco.”
The court dismisses the trademark claims due to the nominative use doctrine.