A class action suit that has been consolidated for pretrial proceedings in the district court in Chicago charges the defendants with conspiring to fix prices of text messaging services in violation of federal antitrust law. The district court allowed the plaintiffs to file a second amended complaint despite the defendants’ objection, based on
Bell Atlantic Corp. v. Twombly,
Section 1292(b) requires our permission to appeal as well as the district court’s. The defendants have asked our permission and the plaintiffs urge us to turn them down. They argue that the proposed appeal does not present a “controlling question of law,” as the statute requires. The question presented is whether the second amended complaint states a claim under the standard for pleading set forth in Twombly. It is a controlling question, because if the second amended complaint does not state a claim, the case is likely (though, as the district judge said, not certain) to be over; the plaintiffs are unlikely without discovery to be able to allege additional facts that would persuade the district court to allow them to file a third amended complaint if *625 we held that the second should have been dismissed.
But is it a controlling question of law? It is not an abstract legal question such as whether the Sherman Act forbids price fixing; it is a question whether a particular complaint satisfies the pleading standard of Twombly. Yet the question’s narrowness should not disqualify it, at least in the rather special circumstances presented by the appeal. Suppose the defendants had been overheard to say to each other “let’s fix our ice machines,” and the controlling question in the case was whether “ice machines” was a euphemism for prices of text messaging services. That would be a question of fact — and it would be pointless to allow an interlocutory appeal from its resolution. Disputed facts are resolved at trial — by the verdict if it’s a jury trial and if it’s a bench trial by the judge’s findings of fact — and thus resolution comes at the end of the trial, which ordinarily is too late for an interlocutory appeal. Though there are cases in which a protracted hearing on relief follows the determination of liability, one can understand why Congress didn’t think it necessary to authorize interlocutory appeals to decide whether a finding of fact by a district court was clearly erroneous.
The interlocutory appeal that we are asked to authorize in this case does not seek to overturn any findings of fact. The defendants are arguing rather that even if all the factual allegations of the complaint are true, the complaint is insufficiently plausible to satisfy Twombly. They are asking us to apply a legal standard — the pleading standard set forth in Twombly— to a set of factual allegations taken as true for purposes of the appeal.
A challenge to a trial court’s application of a legal standard to a set of facts is often described as presenting a “mixed question of fact and law” or an “ultimate question of fact,” but these are not helpful labels. The appellate court’s task in such a case is to determine the legal significance of a set of facts. In
Pullman-Standard v. Swint,
Furthermore, when the question presented by an appeal is whether
Twombly
requires dismissal of a complaint, the concerns underlying that decision argue for empowering the district court and the court of appeals to authorize an interlocutory appeal.
Twombly,
even more clearly than its successor,
Ashcroft v. Iqbal,
— U.S.-,
Now it is true that we ruled in
Ahrenholz v. Board of Trustees of University of Illinois,
But in this case there is no question of hunting through a record or immersing ourselves in a complicated contract, and moreover we
do
have a question of the meaning of a common law doctrine — namely the federal common law doctrine of pleading in complex eases, announced in
Twombly.
Decisions holding that the application of a legal standard is a controlling question of law within the meaning of section 1292(b) are numerous. See
Portis v. City of Chicago,
Not that routine applications of well-settled legal standards to facts alleged in a complaint are appropriate for interlocutory appeal. But
Twombly
is a recent decision, and its scope unsettled (especially in light of its successor,
Iqbal
— from which the author of the majority opinion in
Twombly
dissented; and two of the Justices who participated in those cases have since retired). This court has only twice
*627
discussed the application of
Twombly
to antitrust violations, and in both cases only in passing.
Tamburo v. Dworkin,
So we grant the application for interlocutory appeal, and, since the merits of the appeal have been fully briefed in the parties’ submissions and would not, we think, be illuminated by oral argument, we proceed to the merits.
The complaint in
Twombly
alleged that the regional telephone companies that were the successors to the Bell Operating Companies which AT & T had been forced to divest in settlement of the government’s antitrust suit against it were engaged in “parallel behavior.” Bluntly, they were not competing. But section 1 of the Sherman Act, under which the suit had been brought, does not require sellers to compete; it just forbids their agreeing or conspiring not to compete. So as the Court pointed out, a complaint that merely alleges parallel behavior alleges facts that are equally consistent with an inference that the defendants are conspiring and an inference that the conditions of their market have enabled them to avoid competing without having to agree not to compete. The core allegations of the complaint in
Tivombly
were simply that “In the absence of any meaningful competition between the [defendants] in one another’s markets, and in light of the parallel course of conduct that each engaged in to prevent competition from [other carriers] within their respective local telephone and/or high speed internet services markets and the other facts and market circumstances alleged above, Plaintiffs allege upon information and belief that [the defendants] have entered into a contract, combination or conspiracy to prevent competitive entry in their respective local telephone and/or high speed internet services markets and have agreed not to compete with one another and otherwise allocated customers and markets to one another.”
Id.
at 551,
Our defendants contend that in this case too the complaint alleges merely that they are not competing. But we agree with the district judge that the complaint alleges a conspiracy with sufficient plausibility to satisfy the pleading standard of Twombly. It is true as the defendants contend that the differences between the first amended complaint, which the judge dismissed, and the second, which he refused to dismiss, are slight; but if his refusal to dismiss the second complaint is properly described as a reconsideration of his ruling on the first, so what? Judges are permitted to reconsider their rulings in the course of a litigation.
The second amended complaint alleges a mixture of parallel behaviors, details of industry structure, and industry practices, that facilitate collusion. There is nothing incongruous about such a mixture. If parties agree to fix prices, one expects that as a result they will not compete in price — that’s the purpose of price fixing. Parallel behavior of a sort anomalous in a competitive market is thus a symptom of price fixing, though standing alone it is not proof of it; and an industry *628 structure that facilitates collusion constitutes supporting evidence of collusion. An accusation that the thousands of children who set up makeshift lemonade stands all over the country on hot summer days were fixing prices would be laughed out of court because the retail sale of lemonade from lemonade stands constitutes so dispersed and heterogeneous and uncommercial a market as to make a nationwide conspiracy of the sellers utterly implausible. But the complaint in this case alleges that the four defendants sell 90 percent of U.S. text messaging services, and it would not be difficult for such a small group to agree on prices and to be able to detect “cheating” (underselling the agreed price by a member of the group) without having to create elaborate mechanisms, such as an exclusive sales agency, that could not escape discovery by the antitrust authorities.
Of note is the allegation in the complaint that the defendants belonged to a trade association and exchanged price information directly at association meetings. This allegation identifies a practice, not illegal in itself, that facilitates price fixing that would be difficult for the authorities to detect. The complaint further alleges that the defendants, along with two other large sellers of text messaging services, constituted and met with each other in an elite “leadership council” within the association — and the leadership council’s stated mission was to urge its members to substitute “eo-opetition” for competition.
The complaint also alleges that in the face of steeply falling costs, the defendants increased their prices. This is anomalous behavior because falling costs increase a seller’s profit margin at the existing price, motivating him, in the absence of agreement, to reduce his price slightly in order to take business from his competitors, and certainly not to increase his price. And there is more: there is an allegation that all at once the defendants changed their pricing structures, which were heterogeneous and complex, to a uniform pricing structure, and then simultaneously jacked up their prices by a third. The change in the industry’s pricing structure was so rapid, the complaint suggests, that it could not have been accomplished without agreement on the details of the new structure, the timing of its adoption, and the specific uniform price increase that would ensue on its adoption.
A footnote in
Twombly
had described the type of evidence that enables parallel conduct to be interpreted as collusive: “Commentators have offered several examples of parallel conduct allegations that would state a [Sherman Act] § 1 claim under this standard ... [namely,] ‘parallel behavior that would probably not result from chance, coincidence, independent responses to common stimuli, or mere interdependence unaided by an advance understanding among the parties’ ...[;] ‘conduct [that] indicates the sort of restricted freedom of action and sense of obligation that one generally associates with agreement.’ The parties in this case agree that ‘complex and historically unprecedented changes in pricing structure made at the very same time by multiple competitors, and made for no other discernible reason’ would support a plausible inference of conspiracy.”
Bell Atlantic Corp. v. Twombly, supra,
What is missing, as the defendants point out, is the smoking gun in a price-fixing case: direct evidence, which would usually take the form of an admission by an employee of one of the conspirators, that officials of the defendants had met and agreed explicitly on the terms of a conspiracy to raise price. The second
*629
amended complaint does allege that the defendants “agreed to uniformly charge an unprecedented common per-unit price of ten cents for text messaging services,” but does not allege direct evidence of such an agreement; the allegation is an inference from circumstantial evidence. Direct evidence of conspiracy is not a sine qua non, however. Circumstantial evidence can establish an antitrust conspiracy.
Monsanto Co. v. Spray-Rite Service Corp.,
The Court said in
Iqbal
that the “plausibility standard is not akin to a ‘probability requirement,’ but it asks for more than a sheer possibility that a defendant has acted unlawfully.”
The plaintiffs have conducted no discovery. Discovery may reveal the smoking gun or bring to light additional circumstantial evidence that further tilts the balance in favor of liability. All that we conclude at this early stage in the litigation is that the district judge was right to rule that the second amended complaint provides a sufficiently plausible case of price fixing to warrant allowing the plaintiffs to proceed to discovery.
