Levitt v. Yelp: Why Twombly/Iqbal Are Regrettable

I’m just at the start of my pleadings journey with my Civil Procedure II students, which has got me reading lots of terrific scholarship on the regrettable impact of Twombly and Iqbal.  Federal district court Judge Chen’s recent dismissal of the third amended complaint in Levitt v. Yelp, 2011 WL 5079526, seems a perfect example of why Twombly and Iqbal’s insistence upon factual plausibility before discovery transforms FRCP 12(b)(6) motions into something far different from sussing out “fair notice” and more like summary judgment as Suja Thomas astutely suggests.  (Scott Dodson and Adam Steinman also have insightful pieces on Twombly and Iqbal — New Pleading, New Discovery” and “The Pleading Problem,” respectively).

Let me first tell you a bit about the Levitt suit against Yelp.  Plaintiffs, two subclasses of business owners, alleged that Yelp unlawfully manipulated its business review pages to induce them to pay for advertising in violation of California’s civil extortion statute.  In a previous ruling, the court dismissed the complaint with leave to amend because it lacked factual allegations from which a threat might be plausibly inferred.  The Third Amended Complaint (TAC) added allegations concerning Yelp’s dealing with plaintiff Wheel Techniques (WT) in late 2008 and early 2009.  According to the TAC, soon after WT noticed negative reviews on its Yelp page that did not correspond with its actual customers, it received calls from Yelp representatives seeking advertisements.  Plaintiffs alleged that Yelp created false reviews to induce WT to advertise with it and that “200 Yelp employees or individuals acting on their behalf have written reviews of businesses on Yelp.” Plaintiffs alleged that when Wheel Techniques contacted Yelp to ask why a competitor had a high rating on Yelp, the Yelp representative told him the competitor advertised and “we work with your reviews if you advertise with us.” Id. ¶ 78. On March 8, 2010, Wheel Techniques was again contacted to purchase advertising. Upon declining, Plaintiffs allege that a 1–star review was moved to the top of the business page “within minutes” as a threat to induce Wheel Techniques to advertise. Id. ¶¶ 79–81. Plaintiffs allege that Wheel Techniques owner John Mercurio was told several Yelp employees had been fired and computers had been frozen “as a result of scamming related to advertising.” Id. ¶ 82.

The district court ran through the Twombly-Iqbal standard, with a few tweaks.  It explained that a complaint may be dismissed for failure to state a claim if it lacks a cognizable legal theory or sufficient facts alleged under a cognizable legal theory.  Citing Twombly, the court noted that a motion to dismiss should be granted if a plaintiff fails to plead “enough facts to state a claim to relief that is plausible on its face.” “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 court stated that “allegations of material fact are taken as true and construed in the light most favorable to the non-moving party.”  (The “construed in the light most favorable” to the non-movant language was absent in Iqbal).  Citing Iqbal, the court said that: it “need not, however, accept as true pleadings that are no more than legal conclusions or the ‘formulaic recitation of the elements’ of a cause of action.   Determining whether a complaint states a plausible claim for relief … [is] a context-specific task that requires the reviewing court to draw on its judicial experience and common sense.”  The court dismissed the complaint for failure to state a claim.  Here’s the court’s explanation:

The TAC, like the SAC, fails to allege facts sufficient to support a conclusion that Yelp created any content. Plaintiffs have added an allegation that “approximately 200 Yelp employees or individuals acting on behalf of Yelp have written reviews of businesses on Yelp,” TAC 37, and that “Yelp has paid users to write reviews,” id. ¶ 38. Despite these allegations, however, it remains “entirely speculative that Yelp manufactures its own negative reviews or deliberately manipulates reviews to the detriment of businesses who refuse to purchase advertising,” and “[t]he [TAC] provides no basis from which to infer that Yelp authored or manipulated the content of the negative reviews complained of by plaintiffs,” Order Dismissing SAC at 17.  That Yelp employees have written reviews, even for pay, does not raise more than a mere possibility that Yelp has authored or manipulated content related to Plaintiffs in furtherance of an attempt to “extort” advertising revenues. See Iqbal, 129 S.Ct. at 1950 (“[W]here the well-pleaded facts do not permit the court to infer more than the mere possibility of misconduct, the complaint has alleged-but it has not ‘show[n]’-‘that the pleader is entitled to relief.’ ”) (quoting Fed.R.Civ.P. 8(a)(2)). Similarly, that Wheel Techniques noticed negative reviews of its business that did not match its customer records does not support the logical leap that Yelp created those reviews. See TAC ¶¶ 74–75. Nor does an allegation that Mr. Mercurio (Wheel Techniques’ owner) “was told” by an unnamed source “that a former Yelp employee stated that Yelp, upon information and belief, terminated a group of sales employees … as a result of scamming related to advertising” and that “the computers of sales employees were, at one point, frozen to prohibit employees from being able to change reviews” raise more than a speculative possibility that Yelp employees created or substantively manipulated the content of Plaintiffs’ reviews in this case. See TAC ¶ 82. “Scamming related to advertising” could have referred to a host of practices not involving manufacturing of false reviews.

Despite the court’s claim that it would draw all inferences in plaintiffs’ favor, it seemed to do the opposite.  The plaintiffs provided a factually plausible account of extortion: that Yelp employees write reviews; that after WT notice negative reviews that did not accord with its customer records, a Yelp representative called seeking advertising; that a Yelp employee told WT that it works with customers that advertise with it; that as soon as WT refused Yelp’s advertising solicitation, a one-star rating appeared at the top of its Yelp page; and that WT’s owner learned that Yelp employees had been fired due to scamming related to advertising.  Nonetheless, the court reads the allegations seemingly in the light most favorable to defendant movant, reasoning that scamming related to advertising could have referred to a host of practices not involving false reviews and that even if Yelp hired people to write reviews it does not raise more than a mere possibility that it manipulated content regarding WT to extort advertising revenues.  The court seemingly ignores the allegations related to the solicitation calls and the subsequent negative review appearing on WT’s page after it declined Yelp’s second advertising solicitation.  It’s difficult to imagine what more plaintiffs could have alleged at this stage in the game.  Taking a cue from Scott Dodson’s article “New Pleading, New Discovery,” plaintiffs need discovery to figure out if Yelp employees did what the circumstances suggest–manipulate reviews to bully WT and other plaintiffs into buying advertising.  In the end, Suja Thomas has it right.  This whole endeavor sounds a lot more like a summary judgment motion than a motion to dismiss.

You may also like...

15 Responses

  1. TJ says:

    I can agree that, at least based on the presentation here, the district court seemed to have applied to high a standard. But it seems to have applied too high of a standard regardless of the pleading standard at issue. As you note, the district court seemed to have drawn all inferences against the plaintiff, which it is not supposed to do under either Twombly/Iqbal or even summary judgment. In other words, there will always be crazy district courts making bad decisions in individual cases. Laying the blame for every errant decision on the doorstep of Twombly/Iqbal is not a fair critique.

  2. Zack says:

    @TJ: I believe the point the article was trying to make was that Twombly/Iqbal opened the gate for “crazy district courts” to make bad decisions. Absent Twombly/Iqbal, it is without a doubt that Judge Chen would not have dismissed the complaint. However, Twombly has murkied the waters and made it possible for absurd interpretations.

  3. TJ says:

    Zack, that is selectively levying an impossible burden of proof. Judging a legal standard on how crazy district judges might wrongly implement it–and we can agree that Judge Chen wrongly implemented (at least as his decision has been described) Twombly and Iqbal–is an exceedingly onerous test, one that I am sure you do not impose on legal standards you favor for other reasons.

    In fact, please name one legal standard that crazy district judges cannot possibly get wrong. I will contest your assertion that Conley was such a legal standard (implicit in your assertion that Judge Chen would have not dismissed the complaint under Conley), by referring you to the the district court in Twombly dismissed that complaint, purportedly under the Conley standard.

  4. Zack says:

    TJ, I would like to apologize for my earlier words being unclear. Often, the Supreme Court will use an opinion to set forth a legal doctrine. Naturally, district courts will occassionally go awry when ruling over cases dealing with a “new” legal doctrine. In dealing with this phenomenon, the Supreme Court will usually offer up a “clarifying” opinion to clear up the murkiness. My point before was that the Supreme Court could have used Twombly/Iqbal to reaffirm their ruling in Conley and forever resolve the issue with pleading via a “clarifying” opinion. District courts would know that a “conceivable” standard (one that is arguably quite simplistic if it had been clarified by the court) was to be used for assessing complaints in all circumstances and, as previously mentioned, Judge Chen would have undoubtedly ruled in the plaintiff’s favor. Instead, the Court chose to not only modify the doctrine but increase its complexity by making it a “plausibility” standard and “context-specific”. Consequently, the doctrine has become cloudy and there are countless unanswered questions. When there is great uncertainty surrounding a legal doctrine, the risk for adverse decisions only increases. Judge Chen’s ruling is but one example of what happens when a doctrine, no matter its merits, lacks clarity.

  5. Bruce Boyden says:

    TJ, I think this case fits a pattern of post-Twombly/Iqbal dismissals that deviates from the pre-Twombly state of affairs (although it’s important not to overemphasize the distinction; I basically agree with Christopher Fairman’s “Myth of Notice Pleading” article). I was just reading about another one over at Rebecca Tushnet’s blog: http://tushnet.blogspot.com/2012/01/pleading-copying-or-substantial.html ; and yet another at the 7th Circuit: McCauley v. City of Chicago, 2011 WL 4975644 (from the dissent: “It is difficult to imagine what more McCauley might allege on the crucial question of intent without reciting a list of specific states of mind that Chicago police policymakers might have.”). I don’t think this is two crazy district courts and a crazy circuit court. I think these rulings are invited by Twombly and particularly Iqbal’s language indicating that judges should dismiss if the allegations do not seem plausible based on “common sense and judicial experience,” which can in part be determined by looking to see if there is an “obvious alternative explanation” for the facts pled by the plaintiff. Despite some denials in Twombly/Iqbal that plausibility is a mere probability test, courts seem to be applying the “obvious alternative explanation” language to dismiss cases where the defendant’s story seems, to the judge’s ears (using “common sense and judicial experience”), more likely. Iqbal itself I would argue is an example.

  6. TJ says:

    Bruce, my point in the first comment was not that Twombly/Iqbal cannot ever be criticized for reaching bad results. My point was that Levitt v. Yelp does not seem a good example. In fact, I am sure that Twombly/Iqbal produces some bad results, even when properly applied. I happen to think that the bad results (cases dismissed that, in an ideal world, would not be, and meritorious cases deterred by the prospect of dismissal) are outweighed by the good results (cases dismissed that in an ideal world would be, but which would not have been under Conley, and unmeritorious cases deterred); but that is an empirical argument that no one has good data on.

    Zack, three points. First, contrary to your assumption, a “we really meant it in Conley” opinion that would not have settled the issue or resulted in clarity. The Supreme Court in fact issued multiple such “we really mean it” opinions before Twombly, but courts of appeals kept coming up with new ways to tighten pleading standards, so I doubt that another such opinion would have had the clarifying effect you suppose. Second, you seem to be taking a position of putting predictability and clarity above all other considerations, but I seriously doubt you mean that. If clarity was the only virtue, a rule of “dismiss every complaint” is even more clear and predictable than the Conley rule of “never dismiss any complaint.” And if you want to say that the “dismiss every complaint” rule produces terrible outcomes, you’ll need to say why Conley’s “accept every complaint” rule produced better ones. Third, speculating on hypothetical Supreme Court opinions and their effects seems very far from the point that the main post was making.

  7. Zack says:

    TJ, it seems to me that we have multiple disagreements. I am a proponent of the Conley standard and you are a proponent of the Twombly standard. In addition, you feel Judge Chen ran afoul and that the decision should not lay at Twombly’s door step, while I feel Judge Chen’s decision, where the pleading standard has evolved to the level of summary judgment, is the natural result of Twombly/Iqbal. In the interest of time, I will not speak to our first disagreemnt. Twombly set forth the “plausibility” standard and Iqbal explained that using the standard would be a “context-specific” exercise that would rely on a judge’s “common-sense” and “judicial experience”. A learned judge, such as Judge Chen, could come to the conclusion that Twombly/Iqbal stands for the proposition that when discovery is expensive and there is an “obvious alternative explanation”, common sense dictates that the complaint be dismissed. We agree that Judge Chen’s decision is a bad result. The Supreme Court needs to step in to prevent similiar interpretations from happening or risk making complaints withstand a summary judgment-like standard.

  8. Bruce Boyden says:

    TJ, if you are saying that Levitt is atypical of post-Iqbal cases in that the judge failed to interpret the plaintiff’s factual allegations in the light most favorable to the plaintiff, I specifically disagree that that is atypical.

  9. TJ says:

    Bruce, I’m not saying that Levitt is typical or atypical in a judge failing to interpret a plaintiff’s factual allegations in the light most favorable to plaintiff. I have seen it happen both before and after Twombly, and I’m not sure of the rate either before or after. My point, as iterated to Zack, is that judging a legal standard by the ways in which a district court could misapply it is an extraordinarily onerous test, and that suggests the real objection is based on something else (e.g. that one really thinks that the standard is wrong even if properly implemented, but don’t have the data to argue that proposition). If you want to argue that the standard is wrong even when properly applied, fine, but find an example of that (as your cite to McCauley may or may not be–I haven’t checked it).

    To be sure, if legal standard A was particularly easy to misapply and legal standard B was not, all else being equal one would prefer legal standard B. But it seems that this is rarely the true objection of anyone to Twombly, in the sense that Conley–as actually implemented with all sorts of informal exceptions–was not a particularly predictable standard, all else is not equal, and people are really upset about the other things like the fact that Twombly raised the substantive bar and people just want a lower bar.

    Zack, let us clarify what it means to hold complaints to a “summary judgment-like” standard. Practically speaking, there are two somewhat distinct differences between 12(b) motions and summary judgement, one formal and one informal. The formal difference is that summary judgment requires evidence, whereas pleading does not. The informal–but in practical terms much more important–difference is that district judges apply the “draw all reasonable inferences in favor of P” formulation quite differently: in summary judgment they tend to only pay lip service to this, while in pleading they (pre-Twombly) tended to carry this to unjustifiable extremes beyond reasonable inferences and to patently unreasonable ones, basically anything just short of little green men.

    I think when you and Danielle say that Judge Chen applied a “summary judgment-like” standard in Levitt, you mean he not only failed to draw inferences in favor of P, but actively drew them against P. What you certainly do not mean is that he required actual evidence rather than pleading.

    I think Judge Chen erred here because (at least based on the description) he actively drew inferences against the plaintiff, which is erroneous even under the formal summary judgment standard (but which trial judges in practice do all the time). When properly applied, I in fact quite favor a consistent “reasonable inferences in favor of P” standard between pleading and summary judgment, where the only difference between the two is that pleading draws inferences from the facts pled, whereas summary judgment draws inferences from facts that have evidentiary support. In fact, I would argue that this consistent application of “reasonable inferences”–which is what the whole disagreement with Twombly/Iqbal really comes down to–is a much better conceptually fit with the Anderson/Celotex framework, and it is Conley that is the odd one out.

    And that brings me to one last point. As we have just discussed, the summary judgment standard is also capable of being, and regularly is, misapplied. Trial judges frequently fail to draw evidentiary inferences in favor of P, or even draw them against P. Yet I don’t see people complaining about how Anderson/Celotex are terrible, horrible, no good, very bad; and I don’t see blog posts with cases where a trial judge misapplied the standard and erroneously SJed a case–which of course happens all the time–with headings like “Why Anderson/Celotex Are Regrettable”.

  10. Bruce Boyden says:

    TJ, I find your response to me puzzling. Surely you are aware that one of the biggest criticisms of Twombly and Iqbal is that it is extremely difficult to figure out what the decisions mean and how they are to be applied consistently with all of the other authority (Swierkewicz, Erickson v. Pardus, Rule 9(b), Rule 8(e) and 8(f), Form 11 and the other forms) that the court left standing — far, far more than in your average Supreme Court opinion. See Judge Hamilton’s excellent dissent in McCauley for more. And I hope you admit that it is a fair basis for criticism of a Supreme Court opinion that it fails to actually clarify the law, or perhaps even makes it more unclear, and consequentially brings some (arguably) harmful results within the bounds of plausible interpretation.

    Given all that, I find your proposed restriction on critics of Twombly and Iqbal, that they must limit their criticisms of lower court decisions to *only those* that come out the way they *most* like, bizarre. That’s certainly a convenient restriction for defenders of Twombly and Iqbal. But it seems a bit ad hoc. It would seem to me fair game to criticize Twombly and Iqbal for leading to a substantial line of lower court cases that were either unintended (shame on the Supreme Court for not making the standard more clear) or intended (shame on the Supreme Court for being surreptitious!).

    As an aside, I see lots of criticism of Anderson and Celotex, and I see lots of criticism of judges granting summary judgement when they shouldn’t have. However, I don’t see how either issue is relevant to this discussion.

  11. TJ says:

    Bruce, the “bizarreness” is entirely a matter of your own creation. I do not see how you get from what I said to requiring critics to “limit their criticism of lower court decisions to only those that come out the way they most like.” I asked for an example where the Twombly standard was properly applied but where the decision produced a bad outcome. This is presumably not one that comes out the way you most like. And my question does not seem an unreasonable question nor a particularly self-serving one.

    Giving examples that clearly fit within a rule but produces a bad outcome is the typical way of criticizing a rule. For example, people criticize the exclusionary rule by giving cases where evidence is properly suppressed (i.e. fits the properly applied rule) but allows a clearly guilty criminal to go free (bad outcome). I don’t see how that is a “convenient” restriction for me unless there are no such cases, which would speak well of the rule. Using cases that are clearly outside the rule (we all agree that Judge Chen misapplied Twombly) as a way to demonstrate terribleness of the rule is much rarer, because it is bad logic.

    Look, put the shoe on the other foot. Take your favorite rule, in default of which I’ll insert Conley. Imagine that I find an example where a crazy judge improperly applies the rule, such as by allowing a completely blank complaint to go forward to trial. Further imagine that I then write a blog post about why this bad outcome demonstrates the Conley rule is terrible, and that we need to replace Conley with a rule that automatically dismisses every complaint. And when you make the point that my particular case involves a crazy judge mangling the rule and doesn’t support my argument, I make these replies: (1) surely you know that other people, based on other cases, criticize Conley as unclear, (2) it is fair game to use individual crazy decisions (that clearly fall outside the rule to boot) to demonstrate the unclarity of a rule, (3) your argument is bizarre and self-serving.

  12. TJ says:

    And to respond to your aside, yes, we see “lots of criticism of Anderson and Celotex,” and we see “lots of criticism of judges granting summary judgement when they shouldn’t have.” But we don’t see criticism of Anderson and Celotex based on those isolated examples of judges improperly granting summary judgment.

    My criticism throughout this (now long) thread has been on the tying together of these unrelated arguments. You can criticize Judge Chen’s decision. You can criticize Twombly and Iqbal. But it is unfair to criticize Twombly/Iqbal based on Judge Chen’s decision, just as it would be unfair (and therefore is not done) to criticize Anderson and Celotex by pointing to individual judges granting summary judgment when they shouldn’t have. And that is why the issue is relevant.

  13. Thanks for this terrific discussion. Judge Chen seemed to get it wrong on a few scores but what matters most, to me, is that the opinion seemed to insist upon actual proof that when Yelp employees wrote reviews, it did so to bully WT and other plaintiffs into buying advertisements. In important ways, that is precisely what Plaintiffs did, as best it could without discovery, with allegations about the circumstances suggesting Yelp’s reasons for its actions–the employee’s answer about doing right by our advertisers when WT asked why a competitor had so many positive reviews, the moving of a one star review to the top of WT’s review profile after WT again refused to advertise. At the pleading stage, courts say they accept all of plaintiff’s allegations as true, but this court didn’t, in large part because the allegations did not seem a plausible explanation for extortion based on judicial experience. My point about Twombly and Iqbal more broadly is that some evidence suggests that we are seeing far more civil rights and other claims dismissed because of an insistence of proof that is impossible at the pleading stage, that is, before litigants have a chance for discovery. I did not mean to use this case as final proof of what’s wrong with Twombly, just as an example of a trend that the Court, without rulemaking process, has fundamentally changed Rule 8(a)(2)’s notice pleading into a requisite for plausible proof before plaintiffs have had a chance for discovery.

  14. DL Kimzey says:

    Examine Kimzey v. Yelp Inc.

  15. DL Kimzey says:

    FTC July 2013 had 865 complaints against Yelp from business owners most for extortion and false one star ratings because they refused to pay Yelp $300.00 per month.
    May 2014 – FTC complaints totaled 2,046.
    Gaining at a rate of 120 complaints a month,
    How long will it take for Judge Chens 10,000 ducks bites
    To become a reality? That’s how long Yelp’s business model will