The District Court dismissed Roman's case after giving him two opportunities to state a claim upon which relief may be granted and after reconsidering its order of dismissal. Based on the record presented to it, the District Court's decision was correct and should be affirmed as to all but one of Roman's claims (the municipal liability claim for failure to train, supervise, or discipline).
The Majority vacates part of the District Court's judgment by reciting facts found nowhere in Roman's amended complaint and by adding facts of its own creation that were neither pleaded nor argued to the District Court with sufficient specificity. The Majority's deviation from standard civil practice and procedure compels this partial dissent.
I
This dissent results principally from a disagreement with my colleagues about which facts were properly before the District Court. First, the Majority proffers a narrative that Roman never gave the District Court and which has no relevance to the claims it revives. This Court need not (and should not) recite these "facts" and "background" as true. Second, Roman did not sufficiently plead a municipal liability claim based on Newark's alleged pattern or practice of Fourth Amendment violations. If the facts as pleaded (or subject to judicial notice) were as the Majority recites them, I would agree that Roman stated a claim for relief. But since the actual facts before the District Court were quite different from those enunciated by the Majority, the District Court did not err by dismissing this claim.
Despite these disagreements with my colleagues, I agree with them that Roman's amended complaint sufficiently stated a municipal liability claim for failure to train, supervise, or discipline. Yet I cannot agree with their reasoning in toto because we should not extrapolate-and the District Court did not err by declining to extrapolate-from extraneous documents (like the consent decree Roman never provided nor cited to the District Court) to reach that conclusion. This single claim should be resuscitated, but only based on the face of the amended complaint.
A
The Majority purports to recount the facts of this case "as set out in the amended complaint and the transcript of the [state court] suppression hearing." Maj. Op. 793 n.1. Yet precious few of those facts were actually pleaded, primarily because the state court transcript was not proffered to the District Court by Roman. Moreover, the Majority's narrative of Roman's alleged mistreatment has effectively no bearing on the municipal liability claims it revives.
The lion's share of the troubling facts recited by the Majority were taken from sources other than Roman's amended complaint. Those sources-including the state-court proceedings and subsequent briefs-paint a picture the District Court never observed while considering the motion to dismiss. In truth, the amended complaint says nothing about how the investigation began, or the surveillance of Roman's apartment, or the initial interaction between police and Melissa Isaksem, or the officers' use of Isaksem as a decoy to gain entry into the apartment, or the fact that drugs were seized from a common area, or the expletives and threats that specific officers yelled at both Roman and his father, or Officer Mendes's use of physical force. Unlike those graphic and specific facts the Majority extracts, the amended complaint is replete with conclusory and generalized assertions. See App. 261-63.
Here are some examples of the Majority's approach: Instead of averring that Officer Mendes flipped Roman on his stomach and put a knee in Roman's neck, Maj. Op. 794-95, the amended complaint merely states that "[t]he Defendant Officers and Defendant John Does 1-20 (fictitious names) illegally assaulted the Plaintiff, throwing him against a wall and handcuffing him," App. 262. And rather than recounting a detailed plan to initiate an illegal search that included using an unwitting friend as a decoy, Maj. Op. 794-95, the amended complaint states only that " 'Defendant Officers' ... and Defendant John Does 1-20 (fictitious names), after having the opportunity to observe that the Plaintiff was a person of Latino descent, initiated an illegal search and seizure of the Plaintiff's residence," App. 261.
Now on appeal, for the first time Roman cites facts establishing how the police gained entry into the apartment, the threatening words they spoke, and the actions of Officer Mendes. Roman Br. 10-12. We should not endorse this unpleaded narrative, nor suggest the District Court erred by failing to manufacture it in the first place.
B
The Majority concludes that Roman's amended complaint (supplemented by the consent decree, a news article, and a press release) contains enough facts to make plausible his claims that his injuries were proximately caused by Newark's: (1) pattern or practice of constitutional violations in the area of arrest practices; and (2) failure to adequately train, supervise, or discipline its officers. Maj. Op. 799-800. The first conclusion is unwarranted. And while the second conclusion is correct, the Majority still errs in its reliance on a document Roman never cited and inferential leaps that Roman's pleadings themselves do not admit.
1
On its face, the amended complaint contains very few facts related to Roman's arrest or Newark's alleged pattern or practice of rights violations, and what it does contain amount only to conclusory statements. See Ashcroft v. Iqbal , 556 U.S. 662 , 678, 129 S.Ct. 1937 , 173 L.Ed.2d 868 (2009) (noting that a court's duty to "accept as true all of the allegations contained in a complaint is inapplicable to legal conclusions," and that "[t]hreadbare recitals of the elements of a cause of action, supported by mere conclusory statements, do not suffice"). When we excise from the Majority's narrative all facts that were neither pleaded nor presented to the District Court and the complaints' legal conclusions, it becomes clear that the District Court did not err by twice deeming Roman's complaint deficient regarding a pattern or practice of rights violations.
This Court should not fault the trial judge for failing to take cognizance of facts or arguments never presented to her, especially here, where Roman chose not to include in his amended pleading facts that could have been gleaned from Defendants'
first motion to dismiss and the consent decree attached to it. See Snyder v. Pascack Valley Hosp. , 303 F.3d 271 , 276 (3d Cir. 2002) (noting that "[a]n amended complaint supercedes the original version in providing the blueprint for the future course of a lawsuit").
The Majority primarily (and incorrectly) relies on that consent decree to buttress Roman's pattern-or-practice claim. Although the District Court could take notice of the consent decree's existence, it's quite another matter to hold it accountable for not accepting as true everything its contents could possibly imply-especially when Roman neither pleaded nor relied upon the decree's contents.
What is crucial is whether Roman's complaint was "based" on the consent decree. 1 In re Burlington Coat Factory Sec. Litig. , 114 F.3d 1410 , 1426 (3d Cir. 1997). Only to the extent the Majority refashions Roman's pattern-or-practice claim such that it is now based on implausible inferences from the consent decree is it "based on" the consent decree. For Roman did not explicitly reference, quote, or rely on the document in his amended complaint-even after the City provided it. The amended complaint merely references the decree's announcement by the Department of Justice months after his arrest. 2 See App. 270 ¶ 68. The District Court, though it must draw all reasonable inferences in Roman's favor, had no obligation to abstract facts or inferences or claims Roman chose not to plead. 3 Instead, he was the master of his own complaint. See Judon v. Travelers Prop. Cas. Co. of Am. , 773 F.3d 495 , 505 (3d Cir. 2014).
And even if we accept as true all the consent decree contains, Roman's arrest was not plausibly part of the pattern or practice of rights violations the decree addressed, except perhaps at the highest level of generality. 4 For starters, the decree addressed police stops and arrests of pedestrians or the occupants of vehicles. See App. 79 (detailing the pattern or practice investigated by the DOJ that led to the consent decree's adoption). The consent decree says nothing about arrests or searches without consent that occur at residences, which is what Roman complains of in this case.
Another problem that distinguishes Roman's complaint from the problems that led to the consent decree is the fact that the decree addressed police practices that disparately impacted the black community. See App. 93-98 (detailing same). 5 But that racial disparity did not apply to Roman, who was Hispanic. See App. 261 ¶ 16. While the consent decree may have been "meant to protect all Newark residents," Maj. Op. 801, the point remains that the pattern or practice giving rise to it was not one that plausibly caused Roman's injuries.
Finally, the consent decree addressed erroneous narcotics arrest reports where "individuals often were purportedly seated in cars holding clear plastic baggies in front of them or on their laps and officers could 'immediately' see the contraband, even though the report indicated that the subject's back was to an officer, or that the officer had not yet approached the car." App. 92 (detailing the pattern or practice investigated by the DOJ that led to the consent decree's adoption). Wholly unrelated to those erroneous reports, Roman alleges that officers exhaustively searched the apartment without a warrant, and he is silent as to where and when the officers found the drugs. See App. 30-31, 262-63.
In sum, Roman's arrest was too dissimilar from the pattern or practice addressed by the consent decree to plausibly allege proximate causation for his injuries. While clear that the DOJ did not enter into the consent decree "because it was impressed with Newark's policing practices," Maj. Op. 801, it was not the District Court's duty to imagine all possible inferences from the document. It was Roman's duty to plead them. See Judon , 773 F.3d at 505 . For the Majority to conclude otherwise, it must derive that pattern or practice from sources not before the District Court and define it at the highest level of generality: Fourth Amendment violations writ large. In other words, my colleagues conclude that because Newark police allegedly engaged in a pattern or practice of Fourth Amendment violations of type x , it follows that they plausibly committed this violation of type y -all based on a document they cannot claim the District Court must have considered. The District Court did not err in failing to perform the Majority's inferential leaps to reach that conclusion based on a document it need not have considered in the first place. It properly dismissed this claim rather than indulge such speculation. See Iqbal , 556 U.S. at 679-80 , 129 S.Ct. 1937 .
With or without these sources, the amended complaint's bare legal conclusions need not be accepted as true. Id. at 678 , 129 S.Ct. 1937 . So Roman failed to state a pattern-or-practice claim on which relief could be granted.
2
Roman's failure-to-train, failure-to-supervise, and failure-to-discipline claim was sufficiently pleaded. But the Majority's method for arriving at this conclusion suffers from similar deficiencies to its pattern-or-practice reasoning. The Majority's reliance on the consent decree is again misplaced for the reasons discussed above. 6 And even if such reliance were appropriate, the consent decree does not make Roman's claim plausible.
Roman's arrest was not plausibly caused by the failures to train, supervise, or discipline Newark officers the Majority cites in the consent decree because no such failures appear in the document. The Majority claims "the consent decree indicates Newark police officers were not trained on 'the requirements of [the] Fourth Amendment and related law.' " Maj. Op. 799. And "per the decree" the City's "training did not cover the basics of the Fourth Amendment." Id. at 800 . It does no such thing. Rather, it indicates that Newark agreed to implement additional training on "the requirements of [the] Fourth Amendment and related law" without any detail as to the status quo it addressed. App. 160. While it's safe to say that the DOJ did not endorse that status quo, the difference between agreeing to train more and agreeing that prior training was constitutionally inadequate regarding the Fourth Amendment writ large should be clear. And, as discussed above, the consent decree arose from a host of policing practices unlike those Roman alleged (except at the highest level of generality). Newark could not plausibly have agreed to the extraordinary liability that would come from admitting that its police training violated the Fourth Amendment in every instance, or in every instance possibly connected to Roman's arrest. Indeed, the decree says no such thing about any instance.
The consent decree is even thinner as it relates to supervisory and disciplinary issues. From the City's agreement to adhere to certain review processes and disciplinary measures regarding unlawful searches and false arrests, the Majority perceives a "deliberate indifference to Roman's Fourth Amendment rights." Maj. Op. 800 (internal quotations and citation omitted). That does not follow. The decree does not describe or admit any processes or measures already in place or any existing pattern of unlawful searches or false arrests.
Instead, Roman's amended complaint directly alleged training, supervision, and discipline problems with adequate specificity to survive a motion to dismiss. See, e.g. , App. 270-72 ¶¶ 68, 70, 71, 78, 80, 82; see also Doe v. Luzerne Cty. , 660 F.3d 169 , 179-80 (3d Cir. 2011) (quoting Carter v. City of Phila. , 181 F.3d 339 , 357 (3d Cir.1999) ) (detailing standard at summary judgment for failure-to-train claim). The Majority's improper reliance on the consent decree and inferential leaps from that and other sources outside the amended complaint are, in my view, erroneous and unnecessary.
* * *
As we have noted many times before, we are a court of review, not a court of first view. See, e.g. , In Re: J & S Props., LLC , 872 F.3d 138 , 148 (3d Cir. 2017). Our review is based on the record as presented by counsel in our adversary system. We should not fault the District Court for failing to manufacture facts and craft arguments that Roman neglected to plead. By conjuring its own facts repackaged as if pleaded in the amended complaint, the Majority imposes a new duty upon district judges within the Third Circuit. It does so without citing precedent for the proposition that a district court must consider facts and arguments never pleaded or argued by the plaintiff. I cannot subscribe to this new rule.
This appeal implicates a fundamental legal principle: the plaintiff is the master of his complaint. Because of that time-honored principle, Roman's failure to state a policy-or-practice claim upon which relief may be granted requires the harsh sanction of dismissal. After his initial complaint was found inadequate, Roman failed to file an amended complaint that cured the deficiencies identified by the District Court. Even assuming Roman might have had a legitimate claim, it would have been improper for the District Court to try to make Roman's case for him. And it's especially inappropriate for us to overrule the decision of a district judge because of a failure to apprehend facts and arguments never presented to her. I respectfully dissent.
His appellate briefs' references to the document are not determinative, no matter how many times they cite the decree.
This timing further complicates the Majority's reliance on the consent decree. The decree's announcement months after Roman's arrest requires yet another "infer[ence] that the problems that led to it were occurring during the time of his allegations and for some time before that." Maj. Op. 799. It also requires an inference that all of the City's "corrective measures postdated the arrest." Id. at 802 n.8.
The document's undisputed authenticity as a government document says nothing about the reasonableness of the inferences the Majority abstracts from the consent decree. Nor does it speak to the contents' relevance to Roman's case. Such authenticity merely provides one reason for judicially noticing the decree's existence and eliminates one potential reason for not relying on it. It does not follow that it is "especially important" for district courts to rely on and extrapolate from such documents. Maj. Op. 797-98.
The consent decree itself admits no specific pattern or practice of rights violations. Although it followed a DOJ report that "revealed a pattern or practice of constitutional violations in areas including stop and arrest practices, use of force, and theft by officers," that report was never provided to the District Court. App. 137; see Maj. Op. 797. Instead, the consent decree only outlines measures Newark agreed to take-not any pattern or practice of rights violations, let alone one that plausibly caused Roman's injuries. In fact, that report actually demonstrates that even the pattern or practice that led to the consent decree could not plausibly have caused Roman's injuries.
In his motion for reconsideration, Roman claimed his municipal liability argument was based on "the City's widespread and systemic misuse of police powers to treat members of a protected racial class different from those of white citizens." ECF 43-3 at 8. Unlike the pattern or practice of Fourth Amendment violations the Majority now remands, he argued "racial profiling, racial discrimination, or other widespread discrimination of minorities" gave rise to his municipal liability cause of action. Id.
The Majority's reliance on the news article and press release hyperlinked in Roman's complaints is likewise inappropriate. The news article's identification of one officer-who may or may not have been involved in Roman's arrest-who told a reporter he " think [ s ]" he did not receive training for 20 years is not enough to subject the City to liability for failure to adequately train its entire police force. App. 134 (emphasis added). This demonstrates no custom; nor does it plausibly demonstrate the police academy training all officers receive was constitutionally inadequate without more follow-up. Nor does the article address supervision or discipline. Similarly, the press release addresses none of the three.