Regression damages model fails to convince court

Reed Const. Data Inc. v. McGraw-Hill Companies, Inc., --- F.Supp.3d ----, 2014 WL 4746130, No. 09–CV–8578 (S.D.N.Y. Sept. 24, 2014)

 

Reed sued McGraw-Hill for violations of the Lanham Act, the Sherman Act, and various state law torts. The parties are the only two competitors in the business of providing construction product information (CPI), which allows subscribers in the building trade to bid for jobs.  They sell subscriptions to “nationwide searchable databases that can filter projects based on the user’s preferences. For example, a user can search for library projects in Topeka, Kansas, worth more than three million dollars, that need plumbing in the next two months.” The CPI services provide plans, bidding information, and contact information for the planner, architect, or general contractor on the job. Reed alleged that McGraw-Hill surreptitiously accessed Reed’s database (Connect) and used that access to generate false or misleading product comparisons with McGraw-Hill’s Dodge Network that it distributed to prospective Reed customers.

 

CPI customers prefer a service that lists more projects over one that lists fewer, so the parties compete to have the most projects in their databases. Their user agreements limit permissible use of the information, and the agreements don’t include “creating comparisons with competing CPI providers.”  (That prohibition of comparisons seems anticompetitive and against public policy, as opposed to a prohibition on scraping data, which has different justifications.)

 

Around 2004, McGraw-Hill began to access Reed Connect in order to create favorable comparisons; to do so, it needed to know how many projects were listed on Reed Connect.  It also wanted to be aware of changes in the marketplace and to ensure that Reed was not listing significant projects that it had missed. McGraw–Hill paid consultants—“referred to internally as ‘spies’”—to subscribe to Reed Connect. They would sometimes falsely claim that the fake entities they created to subscribe were associated with actual builders and contractors. McGraw–Hill paid these consultants $3.45 million in cash and personal checks and listed the expenses on its books as “Stationery and Supplies,” or “Magazines and Books.”

 

McGraw-Hill hired Roper to generate product comparisons, but, according to Reed, Roper wasn’t independent, as it claimed. Rather, Roper “did little more than send someone to sit in a room and watch a McGraw–Hill employee run searches on the two services,” without ensuring that the two searches were fairly comparable. McGraw–Hill allegedly used one of its other  products in the tests but said that it had used the Dodge Network.  The searches were selected so as to emphasize McGraw–Hill’s strengths and minimize Reed’s by limiting comparisons to projects worth more than $1 million, whereas Reed was stronger below $1 million.  In addition, McGraw-Hill allegedly ran searches to get projects that needed to be completed expeditiously (ASAPs) from its database but not from Reed’s database. The result was a report in which McGraw–Hill boasted “71% more planning projects, 78% more bidding projects, and 71 % more digitized plans and specifications.”

 

McGraw-Hill also made ad hoc comparisons of the services in response to questions from customers. McGraw–Hill frequently advised customers to search for a particular project in both services, knowing that the suggested project would be found only in the Dodge Network, as well as suggesting state and local comparisons that were generally similar to the Roper reports in both content and methodology. McGraw–Hill touted a five-to-one advantage in projects “exclusive” to McGraw–Hill. Reed alleged that the true ratio was closer to 2.6–to–1.

 

“On at least a few occasions, McGraw–Hill used its access to Reed Connect to find new projects.” McGraw–Hill said these were “isolated potential violations of McGraw–Hill’s rules in which McGraw–Hill may have used Reed Connect to obtain a source of project leads.” The parties agree that McGraw–Hill broke its own rules at least a few times and used its access to Reed Connect for purposes other than generating comparisons.

 

Reed sued in 2009; its RICO claims were dismissed, but the other claims proceeded. At the motion to dismiss stage, Reed alleged that no fewer than 231 customers reported noticing the Roper Reports and were influenced by their contents. “Discovery has not borne out that claim,” though Reed had one customer declaration showing that the Roper reports influenced purchases.  Reed also argued that it was injured because it was forced to price its services lower than it otherwise would have absent the misconduct. It offered Dr. Frederick Warren–Boulton’s testimony in support of this claim; McGraw-Hill moved to exclude his testimony.

 

Dr. Warren-Boulton opined on four questions: Was there a distinct national market for CPI sufficient to trigger § 2 of the Sherman Act? Did McGraw–Hill exercise power in that market? Did McGraw–Hill’s misconduct allow it to keep its market power? Did McGraw–Hill’s misconduct damage Reed? To support his opinions, he conducted statistical regression analyses of the parties’ pricing and service data in an attempt to isolate the effect of the variable at issue here (McGraw-Hill’s alleged misconduct).

 

In order to isolate the price effects of the misconduct, Warren-Boulton compared the parties’ prices for national services during the relevant period with the parties’ prices for local services during the relevant period.  This was based on the assumption that national pricing was affected by McGraw–Hill’s misconduct significantly more than local pricing, and that the effects of McGraw–Hill’s misconduct would grow weaker over time (because the misconduct ceased in approximately 2008). If the difference between each party’s price index declined over the relevant period, and that decline couldn’t be attributed to any other observable factor, then Warren-Boulton would consider that proof that McGraw–Hill’s malfeasance worked a price effect.

 

McGraw-Hill objected to the assumptions of the model.  Warren-Boulton acknowledged that Reed presumably had been becoming a more effective competitor, though the local market had always been effective; if that were true—and the evidence suggested it was—McGraw would have to cut its national prices but not its local prices, adequately explaining the narrowing gap without the presence of any misconduct. This was a significant flaw that, coupled with other flaws, rendered the model inadmissible.

 

McGraw-Hill also argued that Warren-Boulton’s model had to be wrong because he found a price effect with no corresponding quantity effect: he found that “the misconduct differentially affected the prices that customers were willing to pay for each of the two competitors’ services, but had no effect on how much customers chose one over the other.”  That contradicted standard microeconomic theory, which predicted that in almost all markets (excluding perfectly inelastic goods, Giffen goods, and Veblen goods, none of which were involved here), increased price decreases consumption.

 

Warren-Boulton responded that the CPI market had negotiated prices, so there could be a price effect without a quantity effect “because the price each consumer is willing to pay is a function of the price of the competing product and the relative value of the competing product and the negotiated product.”  But that would only be true if Reed and McGraw-Hill had a bigger range of prices they’d accept than prices that consumers would offer to pay.  But there was no evidence to support that, and no reason to believe that the CPI market had these “unusual economic characteristics.”

 

McGraw-Hill also convinced the court that construction volume was an important omitted variable in the analysis. National firms were hit harder by the 2008 recession than state and local firms, and price indices were in fact highly negatively correlated with construction volume data. The omission of a major variable was fatal to one of Warren-Boulton’s models.

 

When he added construction volume data, “a new problem arose: multicollinearity.” This happens when the independent variable is correlated with one of the control variables, making it impossible to isolate the effect of the independent variable on the dependent variable. “Because of the correlation between the explanatory variables, there is insufficient variation in the data set to produce statistically significant results.”  As it turns out, construction volume was highly correlated with both the independent and dependent variables. This made the independent variable (here, the misconduct) appear not to have statistical significance.

 

Warren-Boulton defended his choices by showing that using construction volume alone didn’t explain the prices and in fact had weird results (increasing prices for one party but decreasing them for another, and vice versa in different markets), but the court was unconvinced.  Among other things, Warren-Boulton was unable to explain his decision to pool local and national data in light of his expertise, and running the numbers without pooling produced opposite results (no price effect). Although there was no reason to believe that his judgment was “anything other than perfectly sensible,” he had no methodological explanation for his judgment, and a different judgment would also be reasonable and totally change the outcome.

 

Finally, and relatedly, the methodology he used was too manipulable to qualify as “scientific.”  The choice of end dates for measuring when the effect of McGraw-Hill’s misconduct fully dissipated was more or less arbitrary. That’s not fatal on its own; any statistical model requires some judgment. As long as the model is “robust with respect to different choices of arbitrary points, there is no pressing issue.” But here the choice of the end-date had an outcome-determinative effect; changing the end dates within a “very conservative” range produced a result of no price effect.  Generally, choice of a reasonable timeframe is an issue of credibility for the jury.  “But where, as here, very minor changes in arbitrarily selected model parameters can entirely alter the model’s conclusions, that model is insufficiently robust to withstand the scrutiny of Rule 702.”

 

Thus, Reed failed to meet its burden of showing that Warren-Boulton’s testimony was sufficiently reliable to be admissible.

 

Turning to the false advertising, Reed identified a number of false or misleading statements:

 

First, the court had to identify what was “advertising and promotion”; McGraw-Hill argued that only some of the misrepresentations were sufficiently disseminated to count. Should the ad hoc statements be considered together or separately? The court decided to take McGraw-Hill’s promotional efforts as a whole.  Unlike individual conversations that aren’t advertising or promotion, “the ad hoc comparisons at issue in this case were an undisputed part of a broader campaign to compete with Reed and to tout the supposed advantages of the Dodge Network over Reed Connect.”  There was also evidence that McGraw–Hill management directed individual salespeople to disseminate several of the allegedly false or misleading statements. “There is little difference between this and a traditional advertising campaign in either purpose or effect. … [T]he mere fact that the promotional campaign took the form of individual conversations does not mean that it is not advertising when taken as a whole.”

 

Turning to falsity, the court began with Roper’s involvement and the representations that Roper, an “independent” firm, “oversaw the entire comparison process [and] ensured that comparable categories were used” to evaluate the competing services. McGraw–Hill similarly represented that the reports were “independent,” “objective,” “audited,” and “unbiased.”  Roper’s “project director” testified that he made sure that the searches conducted were “worded similarly,” but he also told a colleague that McGraw–Hill paid Roper “just to say we oversaw the whole process.” He testified that he “did not know if [the searches were conducted] using Network or Dataline, another McGraw–Hill service.” Though the McGraw-Hill employee who conducted the comparisons testified that Roper “verified the numbers,” “made sure that they were not being misrecorded,” and “ensured that the comparisons were run in similar ways and that one search mirrored another search,” that didn’t make the truth of the claims uncontroverted. A reasonable jury could find literal falsity in the claims that the reports were “independent,” “objective,” and “overs[een]” by Roper.

 

Next set of statements: The Roper Reports and the ad hoc comparisons allegedly overstated the number of projects in McGraw Hill’s database as compared to Reed’s database by using the wrong database; exluding some Reed projects (including some utilities projects and the ASAP projects it counted for itself); double-counting some McGraw-Hill projects; and selecting search criteria designed to highlight its relative strengths. Reed alleged literal falsity in the use of a different database, Dataline, for at least one Roper Report, the double-counting of some of McGraw-Hill’s projects, and the imbalanced treatment of ASAP projects.  Reed failed to provide evidence that the Dataline listings weren’t in fact included in “Dodge electronic listings,” so its first literal falsity claim failed.  Likewise, the Dodge network listed some projects on dual tracks as multiple projects, but this is a perfectly sensible way to count: a school might seek asbestos removal while simultaneously planning a new wing.  Reed didn’t provide evidence that “projects” couldn’t have this meaning, so that single institutions could have multiple “projects.”

 

It was undisputed that a search for projects whose bid date was ASAP would yield more results in McGraw-Hill’s database, because Reed listed ASAP projects by simply leaving the bid-date field blank. McGraw–Hill characterized that as an error in Reed’s search algorithm. Reed didn’t offer evidence that the statement that both comparisons were based on searches for projects whose bid-date was listed as “ASAP” was false.

 

What about stale “Executive Briefs” citing data from a “recent” comparison from 2007 when there were more recent comparisons?  The briefs didn’t claim to use the most recent comparison, and words like “recent” are subject to a range of reasonable interpretations, so even in 2012 that wasn’t literally false. “[T]he Lanham Act does not require that comparisons listed as recent be based on the most current available data.”

 

False claims of exclusivity: Reed offered some circumstantial evidence that projects that McGraw-Hill claimed were exclusive to it were also in Reed’s database. On at least one occasion, Reed searched its database the day after McGraw–Hill told a customer that seven projects were exclusive to its service and found six out of the seven purportedly exclusive projects. A reasonable juror could find literal falsity.

 

Claimed project ratios of 5:1 in exclusive projects and 3:1 in all projects: Reed’s expert came up with substantially smaller ratios, but McGraw-Hill argued that she just used different means of calculation. Reed presented “plenty” of evidence that McGraw–Hill’s employees did not know how the ratios were calculated when they distributed them. So, the evidence was that other calculations, of contested accuracy, showed significantly lower advantages for McGraw–Hill than the ratios it touted, but there was no evidence on how it calculated those ratios. A reasonable juror could find literal falsity.

 

For the literally false statements, consumer deception would be presumed. For the rest, evidence of deliberate deception or consumer confusion would be required.  Reed first tried to show deliberate deception.  (1) McGraw-Hill spent a lot of money getting access to Reed Connect and generating the Roper Reports. (2) McGraw–Hill conducted its comparisons when they would be most advantageous to McGraw–Hill and “crafted search queries designed to maximize the McGraw–Hill projects counted while minimizing the projects counted for Reed.” (3) McGraw–Hill convinced consumers that the Roper Reports were independent. The court found this evidence insufficient to allow a reasonable jury to find deliberate deception, only recklessness.

 

As for consumer confusion, Reed submitted one declaration to show confusion.  But McGraw-Hill’s evidence of lack of confusion was “overwhelming” and one declaration was not enough for a reasonable jury to find that a substantial number of consumers were misled by the challenged statements.  Reed identified one customer “out of a national market that both parties concede contains at least 70,000 customers,” and the declarant might not actually have made the purchasing decisions at his company.

 

The court then analyzed the materiality of the remaining, possibly literally false, statements: (1) the statements about Roper’s involvement, (2) the statements touting exclusives to certain individual customers, and (3) the statements about the 5:1 and 3:1 project ratios. No reasonable juror could conclude that any of these statements was material.  Interpreting the Second Circuit’s adherence to older language about misrepresenting “an inherent quality or characteristic of a product,” the court concluded that this phrase meant “likely to influence purchasing decisions.”

 

Reed’s evidence failed for the same reason its evidence of deception failed: at worst, one customer relied on the misrepresentations.  “Every other customer testified that the Roper Reports and ad hoc comparisons were immaterial.” Summary judgment on the Lanham Act claims was granted.

 

McGraw-Hill also sought to get rid of claims that its disparaging ads constituted monopolization and attempted monopolization in violation of Section 2 of the Sherman Act. It is very hard to show an antitrust violation through misleading advertisements, because the test has a bunch of weird presumptions that aren’t really consistent with how false advertising works. You’re better off with the Lanham Act.

 

In the Second Circuit, “a plaintiff asserting a monopolization claim based on misleading advertising must overcome a presumption that the effect on competition of such a practice was de minimis” and therefore insufficient to sustain an antitrust action. To rebut that presumption, a plaintiff must show that the challenged statements were “[1] clearly false, [2] clearly material, [3] clearly likely to induce reasonable reliance, [4] made to buyers without knowledge of the subject matter, [5] continued for prolonged periods, and [6] not readily susceptible of neutralization or other offset by rivals.”  Reed’s arguments that the use of Roper as a third party guarantor triggered special rules, and that an exception should exist for two-competitor markets, were unavailing.

 

Plaintiffs don’t need to win on every factor to rebut the presumption. The inquiry is simply “whether a disparaging advertisement is so deceptive as to constitute anticompetitive exclusionary conduct.” The presumption formalized the rule that “[i]solated business torts, such as falsely disparaging another’s product, do not typically rise to the level of a Section 2 violation unless there is a harm to competition itself.”

 

There was, as noted above, sufficient evidence of literal falsity for some statements.  But literal falsity is not clear falsity—otherwise the word “clear” would be meaningless. (This seems to me an example of courts seizing on terms that were basically accidental. The literally false/misleading distinction in Lanham Act jurisprudence is relatively new; and anyway there is no reason to think that courts deciding antitrust cases were thinking about the Lanham Act in when they were formulating the antitrust test.)  So what does “clearly false” mean?

 

Epistemologically speaking, falsity is an absolute: a statement is either false or it is not. But the level of justification of one’s belief in a statement’s falsity can vary by degree. Thus, while a statement is either false or it is not, it can be more or less “clearly” false, as measured by how much thought or effort one has to put into determining its veracity or how confident one is in its falsity—or, put another way, how obvious or apparent its falsity is in light of the statement itself and its relationship to the state of the world.

 

A reasonable person could believe that Roper’s involvement in its reports was not a sham, given that a Roper employee was present during the challenged comparisons and made sure that the individual search terms used were comparable.  A reasonable person could likewise believe from the evidence that, “upon learning that McGraw–Hill was touting exclusive projects that Reed did not have in its database, Reed scurried to add them, and, therefore, the claim of exclusivity was true when made.” And there was still no evidence in the record about how the claims about the 5:1 and 3:1 ratios were calculated.  So the evidence was insufficient to show that the challenged statements were clearly false.

 

Obviously, the evidence also didn’t show that the statements were clearly material or likely to induce reasonable reliance.  As for customers’ knowledge, Reed argued that, because its customers lacked knowledge of complex data and statistical analysis, they were unable to discern the accuracy of McGraw–Hill’s claims.  The court disagreed—“buyers do not need a degree in statistics to count how many projects of a given type, value, and location appear in either service,” and there was evidence that “plenty of buyers conducted their own analyses when deciding which service to purchase.”

 

Exposure to the claims was prolonged, but that didn’t help.  Reed argued that McGraw-Hill’s statements weren’t susceptible to neutralization because they couldn’t easily be disproven and because McGraw-Hill tried to keep some of the comparisons from Reed.  But the challenged statements were simple sums of how many projects were in each database, and Reed definitely knew about them. As a result of the combination of the factors, the presumption of de minimis effect on competition held and McGraw-Hill got summary judgment.

 

Only state law claims remained:  (1) fraud, (2) misappropriation of trade secrets, (3) misappropriation of confidential information, (4) unfair competition, (5) tortious interference with contractual relations, and (6) unjust enrichment. Only Reed’s unfair competition claim survived.

 

Fraud: Reed alleged that McGraw–Hill defrauded it by falsely representing that the “consultants” McGraw–Hill hired to access Reed Connect were not McGraw–Hill employees. New York law, which applied because the fraud was carried out in New York, requires that the alleged losses stemming from a fraud “be the direct, immediate, and proximate result of the misrepresentation,” and that those losses be independent of other causes. But Reed alleged lost profits due to lost customers stemming from McGraw–Hill’s misleading ads, based on information gathered from the fraud.  That wasn’t sufficiently proximate.

 

Trade secrets and misappropriation of confidential information: information in the database was not secret. “Reed’s CPI lost its trade-secrets status—if it ever had any—when Reed gave out free trial subscriptions unaccompanied by any contractual restrictions on their use.” Tortious interference: Reed couldn’t prove injury to its business relationship with customers, because of the lack of harm evidence detailed above. Unjust enrichment:  Again, the undisputed evidence suggested that the only customer Reed allegedly “lost” because of McGraw–Hill’s misconduct didn’t make any purchasing decisions.

 

Unfair competition: McGraw-Hill conceded that on “two or three isolated” occasions, McGraw-Hill employees used project leads that they acquired through their illicit access to Reed Connect in their own database. Reed argues this constituted misappropriation. Applying New York law again as the principal locus of the defendant’s conduct, this claim survived. INS v. AP provided the framework, though large portions of New York’s unfair competition jurisprudence are preempted by the Copyright Act. Still, New York protects business people from “all forms of commercial immorality, the confines of which are marked only by the ‘conscience, justice and equity of common-law judges.’” The defendant must have taken something in which the plaintiff had a property right, and that constituted free riding on the plaintiff’s efforts.

 

McGraw-Hill argued that there was no property interest in project counts, but there could be in the underlying data.  “McGraw–Hill used phony entities to surreptitiously subscribe to Reed’s database service, then took the projects it found there and added them to its own database. The project listings are the parties’ stock in trade. Reed has a property interest—or at least a “quasi” property interest—in its project leads.” When McGraw–Hill put those leads into its own database, it “free r[ode]” on the significant effort Reed expended to collect projects. Lack of significant damage or broad scope wasn’t dispositive at this stage.