I'm afraid the new Supreme Court case on antitrust conspiracies, Twombly, is quite insignificant, notwithstanding the view of the estimable Baseball Crank that it "will almost certainly be the most practically significant case of this term." All the case holds is that bare allegations of a conspiracy and parallel conduct do not suffice without more specificity. That much was the widespread actual practice in the lower courts even before this decision. What we really needed guidance on was which "plus factors" would, coupled with parallel conduct, suffice to make out an antitrust conspiracy. On that the opinion is of no help. So all it really does is slap down one wayward decision without really settling anything important for the future.
Related Posts (on one page):
- Why All This Lateral Hiring By Elite Schools?
- The Exploding Laterals Law School Market.
- Doctrinalism and the Legal Academy:
- The Death of Doctrinalism and Its Implications for the Entry-Level Job Market at Law Schools.
- Sabermetrics and the Future of Legal Empirical Studies
- Twombly -- The New Supreme Court Antitrust Conspiracy Case
- Will Basic Legal Subjects Become Globalized?
- Einer Elhauge, Guest-Blogging:
The dissent places a lot of weight on the defendant's membership in a trade association to support the alleged antitrust claim. See Slip Op. at pg. 50 fn. 10.
Doesn't that raise potential First Amendment freedom-of-association problems? See Pfizer v. Giles, 46 F.3d 1284 (3d Cir. 1995)(In re School Asbestos Litigation). That case held that basing a conspiracy claim against a corporate defendant based largely on its membership in a trade association that allegedly engaged in wrongdoing violates the Supreme Court's freedom-of-association case law, which bars assuming that membership in a group (like a trade association) shows support for its illegal goals, as opposed to its lawful goals, unless the group has no lawful goals.
In that case, then-judge Alito found the First Amendment issue so troubling that he granted an extraordinary writ of mandamus to require that the lawsuit be thrown out on summary judgment based on the First Amendment, even though pre-trial refusals to throw out a lawsuit are usually not appealable.
Moments before reading your post I sent an email to the lawyers in my firm explaining the decision. It was titled "Conley v. Gibson is dead." For practicing lawyers, that is enormously important. I agree that the decision provides little affirmative guidance as to what is required, so it may not dramatically change law school antitrust course outlines, but the opinion sweeps away a whole set of rhetorical barriers to disposing of weak claims at the pleading stage. That matters a great deal.
http://www.fed-soc.org/doclib/20070321_EngageOct06.pdf
But as others have pointed out (including the dissent), in order to ratify that common practice in parallel conduct cases, the Court had to revolutionize Rule 8 doctrine in general. That is why the decision has such practical significance--by its terms it is not at all limited to parallel conduct cases, or antitrust cases in general.
By the way, I happen to agree with the dissent that this was the wrong way to go. My personal feeling is that a quick motion by the defense for summary judgment would have been appropriate, with the argument being that without "plus factors", there would be no genuine issue of material fact with respect to the plaintiffs' allegation of an agreement between the defendants, and thus the defendants would be entitled to a judgment as a matter of law. Note that Rule 56(e) would require the plaintiffs to respond to such a motion with specific facts, not just mere allegations.
But as an alternative to answering such as motion as required by Rule 56(e), the plaintiffs could move under Rule 56(f) for additional discovery. That sequence of motions would allow the trial judge to weigh the factors considered by the Court in this case (eg, the burdens of discovery) without revolutionizing the Rule 8 pleading standards.
Anyway, where do we go from here? My guess is that there will be no shortage of trial judges looking to use this decision to help clear "implausible" cases from their docket at the pleading stage. I also guess that the response of the Courts of Appeal will be a mixed bag, insofar as some may like looser pleading standards and vice-versa. Thus, I expect that in the nearish future the Supreme Court will get a chance to speak on these issues again outside of the antitrust context.
What happens then is beyond my powers to predict--the Court could either enthusiastically endorse and expand these new pleading doctrines, or find ways to limit this holding to this particular context. But it should be very interesting.
I guess the issue courts will have to decide, without much guidance from Twombley, is how much weight to give plaintiff's allegations in deciding a 12(b)(6) motion. It has long been the case that, for purposes of a MTD, the Court accepts plaintiff's allegations as true. If plaintiff makes a bare-bones allegation like the one I described, does the court have to accept it as true? If not, what's the new standard? Is it really "plausibility"? On what basis is the court supposed to determine plausibility? One searches Twombley in vain for the answer. I suspect that a lot of defendants in the coming years will argue plausibility (or lack thereof) in their motions to dismiss, and various district courts and circuits will fashion various tests to determine plausibility as instructed by Twombley. The Supremes will have to select the right test, or go back to Gibson, a few years down the road. They just made a mess of things, and we'll all have to lie in it until they clean it up.
That would seem to be an inadequately supported (ie sanctinable) statement if the only "information and belief" was the parallel conduct.
The Bell Atlantic v. Twombly decision resolves an ambiguity in the Supreme Court's 1957 Conley v. Gibson decision, which rejected a requirement that plaintiffs in general provide detailed factual allegations in support of their complaint at the start of a lawsuit, even if the plaintiff has not had the opportunity to seek evidence to prove his case through processes like discovery.
The Conley v. Gibson decision contradictorily said, on the one hand, that the plaintiff's complaint has to give the defendant "fair notice" of what is alleged, and, on the other hand, said that the defendant had the almost impossible burden of showing that "the complaint should not be dismissed for failure to state a claim unless it appears beyond doubt that the plaintiff can prove no set of facts in support of his claim which would entitle him to relief."
But aside from a few federal courts that harped on the "no set of facts" language in Conley at the expense of "fair notice," most federal appeals courts did not permit totally conclusory allegations, at least where the plaintiff is alleging the violation of a statute that specifies multiple elements that must be satisfied for the plaintiff to show liability.
Moreover, the decision seems consistent with most of the court of appeals' decisions interpreting Swierkiewicz v. Sorema, 534 U.S. 506 (2002), which rejected a requirement that a plaintiff include detailed supporting factual allegations in a lawsuit alleging intentional discrimination (i.e., disparate treatment).
The Stevens/Ginsburg dissent in Twombly also provides additional evidence that Justice Ginsburg is the most unfriendly justice to business (she opposes any limits on punitive damages, and is hostile to preemption), and that Justice Stevens is getting less sympathetic to business the longer he sits on the Court.
The Stevens/Ginsburg dissent argues that the Swierkowitz decision dispensed with any requirement of specific facts in a complaint. See Opinion at pg. 49 fn. 9. (Most courts of appeals did not so interpret Swierkiewicz, as I noted soon after the decision in one of my legal analyses at the Overlawyered law blog, in discussing discrimination litigation after Swierkiewicz. That post is reprinted below.)
Viewed in isolation, portions of the Swierkiewicz decision, which rejected a "heightened pleading" standard in discrimination cases (and said that plaintiffs have to specific allege each element of a prima-facie test used by judges, not for motions to dismiss, but for ultimate findings of liability) could support that claim.
But Rule 8(a)(2) of the Federal Rules of Civil Procedure does require some sort of a "showing that the pleader is entitled to relief," not just a conclusory allegation.
Moreover, as I noted in my legal analysis last year for Overlawyered, the Swierkewicz decision itself reiterated the requirement, found in the court's own past precedents, that the complaint has to give the defendant "fair notice" of what the lawsuit is about, presupposing at least a little bit of detail. See Conley v. Gibson, 355 U.S. 41, 47 (1957).
Moreover, the dissent's view that you can base an antitrust claim based merely on parallel business practices, plus membership in a trade association where you speculate that anticompetitive discussions occurred, raises deep First Amendment problems. See Opinion at pg. 50 fn. 10.
Then-judge Alito (the most business-friendly Justice) rejected corporate conspiracy liability on First Amendment grounds in the In re School Asbestos Litigation case also known as Pfizer v. Giles, 46 F.3d 1284 (3d Cir. 1995). That case held that basing a conspiracy claim against a corporate defendant based largely on its membership in a trade association that allegedly engaged in wrongdoing violates the Supreme Court's freedom-of-association case law, which bars assuming that membership in a group (like a trade association) shows support for its illegal goals, as opposed to its lawful goals, unless the group has no lawful goals.
Then-judge Alito found the First Amendment issue so troubling that he granted an extraordinary writ of mandamus to require that the lawsuit be thrown out on summary judgment based on the First Amendment, even though pre-trial refusals to throw out a lawsuit are usually not appealable.
The majority's decision in Twombly doesn't reach the First Amendment issue, but it likewise dismisses the suit, and Alito joined in Justice Souter's opinion for the court dismissing the suit.
Below is my legal analysis of discrimination law after Swierkiewicz at the Overlawyered Blog:
OVERLAWYERED
JULY 1, 2006
http://www.overlawyered.com/2006/
07/court_quashes_suit_under_ada_r.html
Court Quashes Suit Under ADA Regulation
Can you be sued based on an obscure regulation drafted by bureaucrats that expands the reach of an already broad statute? The First Circuit Court of Appeals thought not in its ruling yesterday in Iverson v. City of Boston. Disagreeing with the Tenth Circuit, it held that lawsuits can't be brought under Justice Department regulations expanding the reach of the Americans with Disabilities Act (ADA) by requiring self-evaluation and transition plans, since having such plans is not always necessary to comply with the ADA's statutory requirement that the disabled receive reasonable accommodations.
It chided the Tenth Circuit for failing to follow the Supreme Court's 2001 decision in Alexander v. Sandoval, which held that regulations expanding the reach of Title VI's statutory ban on intentional racial discrimination to include unintentional discriminatory effects on minority groups were not enforceable through lawsuits, and thus rejected a challenge to Alabama's English-language requirement for drivers' licenses, which allegedly had the unintended effect of discriminating against Hispanics.
Like other circuits, the First Circuit also held that court complaints alleging disabilities-discrimination cannot simply make a "conclusory contention" of discrimination, but rather must contain some supporting allegations, such as that the plaintiff is a "qualified" person with a disability. This matters because the longer a meritless lawsuit stays in court, the more it costs. A suit that costs $250,000 to defeat at trial may cost only $75,000 if tossed out earlier on summary judgment after discovery, and may cost only $25,000 if tossed out prior to discovery on a motion to dismiss the complaint.
In its 2002 decision in Swierkiewicz v. Sorema, an age and national-origin discrimination case, the Supreme Court made it much harder to toss out meritless discrimination suits at an early stage, ruling that a typical discrimination case can survive a motion to dismiss and proceed to discovery even if the plaintiff does not allege specific facts supporting his discrimination claim, such as that he was qualified for the job. The plaintiff need only allege that he was denied a job because of his age, national-origin, etc., without giving his underlying reasons for believing he was the victim of discrimination.
However, the ADA is very different from the typical antidiscrimination statute. It is both broader (since it requires not simply that the disabled be treated as well as non-disabled workers, but also that they be given preferential "reasonable accommodations") and narrower (it expressly protects only "qualified" disabled people, unlike race, sex, and age discrimination statutes, which require that unqualified blacks, women, and elderly people be treated as well as their unqualified white, male, and younger colleagues), containing additional statutory elements that a plaintiff must satisfy.
Since the ADA, unlike other antidiscrimination statutes, requires more than a simple showing of discrimination, the First Circuit was right to require ADA plaintiffs to make more than a simple contention of discrimination in their complaint. As the Supreme Court observed in its Swierkiewicz decision, while a plaintiff's complaint need not provide unnecessary evidentiary details, it nevertheless must "give the defendant fair notice of what the plaintiff's claim is and the grounds upon which it rests."
Possibly so, but what would be the procedure to determine whether the allegation had support? A pre-motion-to-dismiss deposition of the plaintiff? I thought the whole point of Twombley was to eliminate discovery.
Also, I'm not so sure that such an allegation would violate Rule 11. If it did, then the plaintiff in Twombley would be subject to sanctions (beyond dismissal of the complaint), as they did allege an actual conspiracy. The point made by the Twombley plaintiff, as I understand it, is that there's enough fishy looking behavior (in the form of conscious parallelism) to entitle us to use discovery to determine whether an actual meeting of the minds occurred. Suppose that the CEO's of Verizon and Qwest exchanged emails in which they agreed to keep the ILECs off of their respective networks. How would such e-mails come to light, other than through discovery? Why isn't it reasonable to assume that such e-mails might exist (particularly given that, at least under Gibson, the plaintiff is entitled to the benefit of the doubt)? Why isn't it enough for plaintiff to say "this looks fishy, and on information and belief, I think they made an agreement."
The "information" is the fact of the consciously parallel behavior. The "belief" is that this consciously parallel behavior is the result of an actual agreement. So long as plaintiff actually believes this (or can reasonably believe this), why should such an allegation on information and belief be sanctionable? The SupCt seems to be saying that the District Court should substitute its own belief (in the goodness and altruism of business people) for the plaintiff's belief. On what basis should the District Court form its belief? Using what tools? What makes the District Court's belief more likely to be correct than the plaintiff's belief?
When business people enter into agreements in restraint of trade in violation of the Sherman Act, they don't put out press releases. Aggrieved plaintiffs (frequently sophisticated business people themselves) generally figure out that a Sherman Act violation has occurred by observing facts on the ground that themselves amount to mere circumstantial evidence of a violation. The system that has been set up allows (or allowed) such plaintiffs to bring a complaint setting forth such circumstantial evidence, which entitled them to take discovery to determine whether an actual violation has occurred. An allegation on "information and belief" like the one I posited has historically been sufficient to get the plaintiff past a MTD and allow him to take discovery. In the absence of such a procedure, civil enforcement of the antitrust laws by private lawsuits may well be gutted, leaving the government as the most important enforcer of these laws. (Twombley would not, for example, prevent the DOJ from empaneling a grand jury to investigate conscious parallelism standing alone.) It's certainly not clear that this shift from private to government enforcement is what Congress had in mind in crafting the antitrust laws.
I'm not sure what you mean by "conclusory." A bare allegation that "On information and belief, Able and Baker met and conspired to restrain trade by keeping ILECs off of their respective networks" is not a conclusory allegation. It is a specific allegation that two people met and formed a meeting of the minds. It lacks specifics such as date and location of the meeting, but those specifics are not really relevant to the inquiry. The question is whether they met and agreed, not when they met or where they met. What's conclusory about such an allegation?
I'll also just note again that it is not automatically the case that the plaintiffs get discovery if they can survive a 12(b)(6) motion. Rather, the defendants can move under Rule 56 at any time (indeed, with notice to the plaintiffs the trial court could just convert the defendants' 12(b)(6) motion into a Rule 56 motion), and it would be up to the plaintiffs to move for discovery pursuant to 56(f).
It's an interesting reading of the federal rules, but it wouldn't work in practice. Specifically, I am not aware of a court ever granting a motion for summary judgment prior to discovery on the ground that although the complaint stated a claim it was implausible. Rather, if there has been no undue delay, a court will grant plaintiff's 56(f) motion for further discovery.
In my experience, it is not too uncommon for defendants to actually offer a motion styled as a motion to dismiss or in the alternative for summary judgment, or for courts to convert a motion to dismiss into a motion for summary judgment. And I do think (in fact know) that courts have granted summary judgment at that stage without discovery.
Whether it would be proper to grant a motion for summary judgment in this instance without granting leave for discovery is a different matter. I believe the Supreme Court has said that leave under Rule 56(f) should be granted liberally. Nonetheless, Rule 56(f) is worded in a permissive fashion (the trial court "may" order discovery, but is not bound to).
And here I personally think this would be a proper resolution. The plaintiffs are basically alleging that because of evidentiary facts X, Y, and Z, they have concluded element A (that an agreement exists). To the extent they pled A, and indeed did not even need to mention evidentiary facts X, Y, and Z in their complaint, then I agree that they should survive a motion to dismiss.
But then I think it is perfectly fine for the defendants to test the plaintiffs' basis for alleging A with a Rule 56 motion. And if all the evidence the plaintiffs can offer in support of their allegation A is X, Y, and Z, then I think it is fine to grant that motion, because as a matter of law, we know that X, Y, and Z alone are not enough to show A.
Of course, I think it would be fine for the court to order discovery as well. But in my view that is a decision for the trial court, and in a case where the trial court does not believe that X, Y, and Z give rise to an inference that evidence of A is more likely than usual to exist (which is different from the question of whether X, Y, and Z are sufficient on their own to prove A), I think it would be fine for the trial court to deny discovery.
In short, if this is really a problem about people abusing the discovery process with unfounded allegations that nonetheless make out a legal claim (the "fishing expedition" issue), then I do in fact believe the proper mechanism for dealing with that problem is a Rule 56 motion and a consideration of whether discovery is warranted under Rule 56(f), not a modification of the traditional Rule 8 standards.
"Denial of a Rule 56(f) application is proper where it is clear that the evidence sought is almost certainly nonexistent or is the object of pure speculation." Terrell v. Brewer, 935 F.2d 1015, 1018 (9th Cir. 1991).
So, the idea of using the Rule 56(f) context to test whether the plaintiff's allegations are based on "speculation" does not appear to be unprecedented. Again, the mechanism would be the defendant using a Rule 56 motion to see why the plaintiffs alleged A in the complaint. If what evidence they have so far is insufficient--which will often be the case, of course--the trial court can determine if what evidence the plaintiffs have so far at least gives them adequate grounds for a Rule 56(f) motion. If the plaintiff's basis for the allegation is not even adequate enough for that purpose, then it would seem summary judgment is appropriate, even without discovery.
The quote you cite says that summary judgment will be granted when the evidenceis speculative. It doesn't say when the allegations are speculative.
I know summary judgment is sometimes granted without discovery when the plaintiff fails to bring a proper 56(f) motion or when the issues are sufficiently narrow that it is obvious further discovery would be irrelevant. But I am not aware of a case in which discovery was denied because the allegations in the complaint (as opposed to the discovery requests) are speculative. In particular, in any case in which testimonial evidence is relevant to the allegations in the complaint, it's hard to imagine a judge denying plaintiff the opportunity to take depositions.
Exactly--the Rule 56(f) issue would be whether the hypothetical claim by the plaintiff that discovery would reveal evidence of an agreement is purely speculative. But again, I think in this case that determination would be within the trial court's discretion.
Again, keep in mind the sequence. First, I agree with the dissent that the plaintiffs did allege an agreement, and therefore they should survive a 12(b)(6) motion. Moreover, the plaintiffs do have circumstantial evidence for that allegation: the defendants' parallel conduct. In that sense, their allegation was not purely speculative.
But the basis of the defendants' Rule 56 motion would be that such evidence is insufficient as a matter of law. So, the argument would be that without more, the plaintiffs could not meet their burden in opposition to the defendants' Rule 56 motion.
Thus to stave off summary judgment and get discovery pursuant to Rule 56(f), the plaintiffs would have to claim that discovery would reveal additional evidence of an agreement. It is precisely that claim which I believe could be deemed purely speculative, and that is pretty much exactly why I think a Rule 56(f) determination, and not a Rule 8 determination, is the correct approach to the issues raised in this case.
In any event, whether or not this approach would be in accordance with existing Rule 56(f) doctrine is not really determinative, because the Court's decision was not really in accordance with Rule 8 doctrine. And so one could frame my argument as suggesting that if the Court was going to change something to allow trial courts additional discretion to avoid discovery in cases like this, the Court should have changed Rule 56(f) doctrine, not Rule 8 doctrine. Indeed, that is the more obvious choice if for no other reason than that Rule 56(f) is actually about discovery.
The important point here is that the cost of dismising what would later be proved to be a meritorious anti-trust complaint are tremendously high. But any individual citizen — including the trial judge — will capture only a fracture of the value of a meritorious anti-trust complaint. But the cost to the *parties* and the *district judge* of going forward with an anti-trust complaint later shown to be non-meritorious is very high (large discovery costs, overseeing discovery, tying up the docket, boring, etc.). As a result, it will be human nature for district judges to reach out and dismiss "likely losers" even if doing so if net socially inefficient. That's why Twombly will end up being so dangerous.
I also think that judges will quickly use Twombly outside of the anti-trust context. In a race discrimination case, for example, it seems that a judge could dismiss a complaint that says the following: "(1) I was fired. (2) I am black. (3) The person who fired me knows I am black. (4) The person who fired me did so because I'm black." Allegation #4 is completely analogous to the allegation in Twombly that "defendants conspired."
The court acknowledged that plaintiff made the bare "on information and belief, defendants conspired" allegation, but essentially found that they didn't mean it. It's not clear on what basis they made this finding, or how lower courts can be guided. It seems that they think plaintiff didn't mean it because plaintiff also included the allegations of conscious parallelism. Hence, my point that, had plaintiff omitted the parallelism allegations, the Court might have thought they really meant the conspiracy allegation.
I find your conclusion doubtful. The majority suggested that the "they conspired" allegation was a mere "legal conclusion" that a district court could simply "redact" from the complaint when testing its sufficiency.
Is your view that the "they conspired" allegation was a mere legal conclusion *because* of the conscious parellelism allegation?
Remember, conscious parallelism, even with plus factors, does not constitute conspiracy. And you have to have conspiracy to win a Section 1 case. All that conscious parallelism (plus the plus factors) does is get you far enough so that you can get the case to the jury. The jury then decides if the CP plus the other factors are sufficient proof of an actual conspiracy (by a preponderance of the evidence).
Plaintiff's complaint alleged (a) parallelism and (b) actual conspiracy. I think the Court (incorrectly?) read this to mean "we allege that the defendants have behaved in a parallel fashion, that such parallelism constitutes a conspiracy, and we (Plaintiffs) believe that all we have to do is prove parallelism in order to get to a jury." If that was the sum and substance of Plaintiff's complaint, it should have been dismissed. But I don't think it was.
Also, if the Court does want to move to fact pleading, why not do it through the rule making process? Subject to Congressional disapproval, the Court can change Rules 8 and 9 and 12 if it wants to do so. There is also the question of a remedy -- usually prior to dismissal under Rule 12(b)(6) there is an opportunity to amend the pleadings, which amounts to a Rule 56 motion less the affidavits. Is dismissal really the proper remedy for a Complaint light on facts when one can certainly imagine facts that could have been pleaded and been sufficient?
I'm particularly struck by the fact that this ruling comes after the court argued the pleading requirements of the Private Securities Litigation Reform Act scienter pleading standard in another case. Is this an attempt to fend off perceived Congressional displeasure with notice pleading in big corporate cases?
I'm also struck by the fact that SCOTUS didn't leap to the relatively simple to understand and apply rule that a non-conclusory allegation of fact must be offered in support of every element of a cause of action. Certainly, this case would have survived Rule 12 if the Plaintiffs had alleged that "on May 24, 2005 from 5:00 p.m. that the CEOs of each of the defendants held a meeting at the Greenview Country Club in Washington County, Colorado, at which they reached an oral argeement to restrain trade by honoring each other's territories." Yet, it would appear from the ruling that Swierkowitz is not strictly overruled, so that there could even be some elements of the cause of action (e.g. an allegation of interstate commerce involvement or an express statement that the companies were in the same market) that could be omitted which would not allow a dismissal under Rule 12.
Also, suppose that it turned out in discovery that the meeting in Washington County never happened, but that the allegation met Rule 11 because the Plaintiff's informant had lied to them. Suppose further that initial disclosures in the case revealed that there was in fact, instead, an e-mail exchange that supported the conspiracy. Surely, at that point, dismissal under Rule 12 or Rule 56 would no longer be appropriate and the case could continue, even if the statute of limitations had run when the e-mails were discovered and the absence of a Washington County meeting came out.