dissenting:
This case involves the patdown search of a passenger in, not the driver of, a vehicle. Recently, in Arizona v. Johnson, the Supreme Court held: “To justify a patdown of the driver or a passenger during a traffic stop, however, just as in the case of a pedestrian reasonably suspected of criminal activity, the police must harbor reasonable suspicion that the person subjected to the frisk is armed and dangerous.” 555 U.S. 328, 327, 129 S.Ct. 781, 172 L.Ed.2d 694 (2009).
The majority opinion agrees that the record in this matter shows conclusively that the district court made no findings regarding whether Defendant, the passenger in this case, was armed and dangerous. Ante at pp. 245^6. Even further, the majority opinion agrees that the district court erred in holding that the officers were justified in patting down the passenger based only on the “suspicion that the Defendant may have been involved in the theft of a ear.” J.A. 30.
However, whereas the majority sees the district court’s erroneous conclusion as a mere harmless failure to “dot the T and cross the ‘t,’ ” I consider the error to be both a misstatement and, more importantly, a misapplication of binding Supreme Court precedent. For a patdown search to be justified, Arizona v. Johnson explicitly requires a finding that the officer had a reasonable suspicion that this particular passenger was armed and dangerous. 555 U.S. at 327, 129 S.Ct. 781. Not only did the district court fail to make any such a finding, no evidence exists in this case that would be sufficient for this Court to assume the role of a trial court and make such finding in its stead. I must therefore, with great respect for the differing view of my colleagues, dissent.
It is important to note that the district court here did correctly observe that “merely being a passenger in a car that was stopped does not necessarily give the Government the right to frisk one.” Volume I, Transcript of Motions Hearing, at 120, United States v. Braxton, Case No. 1:08-cr-00444-WDQ-l (Mar. 16, 2009) [hereinafter referred to as “Transcript”]. Likewise, our own Court’s case law provides that “[b]ecause a frisk or ‘pat down’ is substantially more intrusive than an order to exit a vehicle or to open its doors, ... an officer must have justification for a frisk or a ‘pat down’ beyond the mere justification for the traffic stop.” United States v. Sakyi, 160 F.3d 164, 169 (4th Cir.1998) (emphasis added). Despite this clear precedent, under the majority’s holding, any passenger in a vehicle with bad license tags could be subjected to a pat-down search, even absent a finding of reasonable suspicion that particular individual is armed and dangerous, if the vehicle is stopped in a high-crime area.
To support this extension of the law, the majority states: “In accord with a variety of other courts, we merely hold that a factual finding that the defendant is a potential ear thief supports the legal conclusion that there is reasonable suspicion that *252he is armed and dangerous.” Ante p. 247. But the “variety” of cases relied upon by the majority generally involved the search of the driver, not the passenger, see United States v. Bullock, 510 F.3d 342 (D.C.Cir.2007), and United States v. Garcia-Rivera, 353 F.3d 788 (9th Cir.2003), or otherwise entailed additional extenuating circumstances, see United States v. Rowland, 341 F.3d 774, 784 (8th Cir.2003) (search of the vehicle’s interior); United States v. Shranklen, 315 F.3d 959, 963 (8th Cir.2003) (patdown of passenger was with consent); United, States v. Tuggle, 284 Fed.Appx. 218, 227 (5th Cir.2008) (search was not in context of a Temj stop but in course of ongoing investigation); and United States v. Williams, 7 Fed.Appx. 876, 885 (10th Cir.2001) (stop became detention, and passenger’s inability to answer questions suggested he might be involved in the theft of the vehicle).
In this case, the record shows that the officers stopped the vehicle because it had an incorrect license tag, had heavily tinted windows, and was being driven in a high-crime area. Those observations constituted the justification for the traffic stop. Thus, at the time the vehicle was stopped, Defendant was merely a passenger in a car stopped by police for traffic infractions. The police had no information that linked Defendant to the alleged theft of the tags or that the vehicle itself was even stolen.
To be sure, the record includes Officer Williams’s testimony regarding Defendant’s “nervousness.” But the district court made no findings with respect to that testimony, or concerning Officer Williams’s statement that he conducted the patdown search of Defendant for “officer safety.” In short, the district court made no findings regarding whether Officer Williams had a reasonable suspicion that Defendant was armed and dangerous. Indeed, beyond the findings about the traffic infractions that justified the stop of the vehicle, the district court made no reference whatsoever to Defendant or his conduct, and certainly not that he was a “potential car thief’ or even specifically that the suspected criminal activity was “auto theft,” rather than the theft of a license plate. Ante p. 247.
Under these circumstances, this Court should not seek to create a particularized “justification ... beyond the mere justification for the traffic stop,” Sakyi, 160 F.3d at 169, namely, Defendant’s “nervousness about the arrival of the police,” ante p. 246. Making such findings on a cold appellate record, particularly those that rest on the credibility determinations of witnesses, falls well outside the proper role of this Court:
Factfinders exist for definite purposes, one of which is to observe the demeanor of [witnesses]. Appellate courts are well-positioned to determine whether a factual finding is without support in the evidence; we are much less able simply to overturn a factfinder on a question on which two views of the evidence are possible.... Factfinders routinely resolve discrepancies between evidentiary sources, and by being able to observe testimony first-hand, they are in the best position to do so.
Hams v. Dir., Office of Workers’ Comp. Programs, U.S. Dep’t of Labor, 3 F.3d 103, 106-07 (4th Cir.1993). See also Anderson v. City of Bessemer City, 470 U.S. 564, 573, 105 S.Ct. 1504, 84 L.Ed.2d 518 (1985) (“In applying the clearly erroneous standard to the findings of a district court sitting without a jury, appellate courts must constantly have in mind that their function is not to decide factual issues de novo.” (quotation marks and citation omitted)).
It is worthwhile to remember that we have long resisted engaging in such fact-finding, even when it might be expedient *253for reasons of judicial economy, and even when the evidence is undisputed:
We are unwilling to ... substitute fact finding at the appellate level for fact-finding at the trial level. The fact-finding responsibility has long been recognized as one for the trial court and sound practice suggests strict observance of this division of responsibility between trial and appellate courts except in the most exceptional circumstances.
Wimmer v. Cook, 774 F.2d 68, 76 (4th Cir.1985). See also United States v. Stevenson, 396 F.3d 538, 543 (4th Cir.2005) (“If appellate courts were to begin the practice of making competitive findings with respect to undisputed or documentary evidence, they would usurp the trial function _”).
Here, while there was “testimony about the license plates that did not belong to the vehicle, heavily tinted windows on the car, [Defendant’s] nervousness about the arrival of the police, ... and the nature of the area in which the stop occurred,” ante p. 246 (emphasis added), there are no actual findings by the district court about these critical facts, particularly with respect to Defendant’s failure to make eye contact, or whether Officer Williams’s concerns for his safety were reasonable or particularized to Defendant.1 Hindsight cannot be the judge of such behavior. See United States v. Martinez-Fuerte, 428 U.S. 543, 565, 96 S.Ct. 3074, 49 L.Ed.2d 1116 (1976) (noting that a purpose of the Fourth Amendment is to “prevent hindsight from coloring the evaluation of the reasonableness of a search or seizure”).
On this record, it cannot be discerned why the district court failed to make the necessary findings. It is entirely possible that the district court heard Officer Williams’s testimony, the Government’s reasoning, and defense counsel’s arguments, and found the stated justifications for a suspicion of Defendant’s being armed and dangerous did not meet the standard of reasonable suspicion. Perhaps the district court instead only found it reasonable for Officer Williams to have suspected Defendant was engaged in criminal activity. Under clear Supreme Court precedent, that is not the required justification for a patdown search of a passenger in a vehicle subjected to a Terry stop.
Likewise, I can find no support in the record for the majority opinion’s statement that, “Braxton then elbowed Officer Williams in an attempt to escape, but he was subdued after a struggle with Officer Williams and another assisting officer.” Ante p. 244. The transcript shows that Officer Williams recalled that Defendant “attempted to elbow me to get me off of him.” Transcript, at-56. Nothing in the record indicates that Officer Williams stated that Defendant made “an attempt to escape.”
To the contrary, Officer Williams testified under cross-examination that the police report was incorrect if it reflected that Defendant “attempted to push [Officer Williams] back and run” and that in fact Defendant complied with Officer Williams’s request to get out of the vehicle *254and put his hands up, allowing Officer Williams to pat him down. Transcript, at 61. He further agreed that “any insinuation in the police report” that Defendant did not comply was incorrect. Id. at 62.
Moreover, because Defendant’s “nervousness” is the sole factor particularized to him and not related to the justification for the Terry stop, that conduct is the linchpin of whether the search in question was reasonable. See, e.g., Illinois v. Wardlow, 528 U.S. 119, 124, 120 S.Ct. 673, 145 L.Ed.2d 570 (2000) (“An individual’s presence in an area of expected criminal activity, standing alone, is not enough to support a reasonable, particularized suspicion that the person is committing a crime.”). See also Terry v. Ohio, 392 U.S. 1, 21 n. 18, 88 S.Ct. 1868, 20 L.Ed.2d 889 (1968) (“This demand for specificity in the information upon which police action is predicated is the central teaching of this Court’s Fourth Amendment jurisprudence.”). As such, the appellate fact-finding engaged in by the majority on this point is especially inappropriate.
I note as well that we have previously recognized that mere nervousness is not necessarily suspicious behavior. See United States v. Massenburg, 654 F.3d 480, 489 (4th Cir.2011) (“Given the complex reality of citizen-police relationships in many cities, a young man’s keeping his eyes down during a police encounter seems just as likely to be a show of respect and an attempt to avoid confrontation.”). The panel in Massenburg further quoted the following persuasive language from other courts:
[I]t is common for most people to exhibit signs of nervousness when confronted by a law enforcement officer whether or not the person is currently engaged in criminal activity. Thus, absent signs of nervousness beyond the norm, we will discount the detaining officer’s reliance on the detainee’s nervousness as a basis for reasonable suspicion. [United States v. Salzano,] 158 F.3d 1107, 1113 (10th Cir.1998) (internal quotation marks and citations omitted). See also State v. Lee, 265 Neb. 663, 658 N.W.2d 669, 678-79 (2003) (“[Nervousness is of limited value” to reasonable suspicion analyses as “it is common knowledge that most citizens whether innocent or guilty, when confronted by a law enforcement officer who asks them potentially incriminating questions are likely to exhibit some signs of nervousness.”).
Massenburg, 654 F.3d at 490. Even Officer Williams conceded that “different people react different ways when the police show up.” Defendant’s supposed nervousness, or failure to make eye contact, is hardly a definitive indicator of dangerousness.
Had the front-seat passenger in this case been the driver’s grandmother, whom the driver had perhaps just picked up to go shopping, rather than Defendant, then her nervousness may well have been deemed typical, rather than suspicious, behavior. And yet, I can find no evidence in this record to suggest that this particular Defendant had any more knowledge that the car in which he was a passenger was stolen than the driver’s grandmother might have.
Nothing indicates that the vehicle to which the license plate actually belonged was stolen just prior to the traffic stop, or that the vehicle appeared to be attempting to flee the police, or that there was any information that more than one individual was responsible for the theft. Moreover, the radio call mentioned only that the tags did not match the vehicle, not that the vehicle itself was stolen. See United States v. Foster, 634 F.3d 243, 248 (4th Cir.2011) (“[A]n officer and the Govern*255ment must do more than simply label a behavior as ‘suspicious’ to make it so. The Government must also be able to either articulate why a particular behavior is suspicious or logically demonstrate, given the surrounding circumstances, that the behavior is likely to be indicative of some more sinister activity than may appear at first glance.”); United States v. Powell, 666 F.3d 180, 185 (4th Cir.2011) (holding that the district court should have suppressed evidence seized during a patdown search of a passenger during a traffic stop despite the passenger’s prior criminal record for armed robbery and his “purported deliberate misrepresentation concerning the validity of his driver’s license”).2 See also Terry, 392 U.S. at 21-22, 88 S.Ct. 1868 (noting that police officers must have “specific and articulable facts” to justify a search, as “[ajnything less would invite intrusions upon constitutionally guaranteed rights based on nothing more substantial than inarticulate hunches”).
In sum, the district court’s error was not some harmless conflation of the two standards under Terry. Rather, the district court misstated and applied the wrong standard to determine the constitutionality of a search, a possible violation of one of the rights we hold most dear. The majority nonetheless makes its own findings to uphold the search in this matter on the basis that the reasonable suspicion that Defendant may have been involved in the theft of a car is sufficient to pass constitutional muster, when combined with an appellate court’s finding of “nervousness,” the sole factor particularized to Defendant. Such a conclusion, and the legal gymnastics it entails, is plainly contrary to the Supreme Court’s holdings in Terry and Johnson, as well as our own Court’s precedent concerning the propriety of appellate fact-finding. Accordingly, I must respectfully dissent.
. In fact, some of Officer Williams’s testimony suggests that his safety concerns were generalized and not specific to Defendant or his behavior. When he first heard the radio call, before he had even seen Defendant or gotten to the location of the traffic stop, he recalled that "four occupants in a vehicle in that area, bad tags, and, based on working in that area for ... six years prior to that incident, that usually means some criminal activity possibly involved.” J.A. 17. After noting that Defendant seemed nervous during the stop, he testified that he told Defendant, " ‘Sir, I've got to pat you down for weapons for safety,’ given the totality of the whole situation and my dealings with bad tags and individuals in that area." J.A. 19.
. Unlike the district court in this case, the district court in Powell explicitly held that “the officers had reasonable suspicion that [Defendant] was armed and dangerous and were thus entitled to frisk him.” 666 F.3d at 182, 2011 WL 5517347, at *1. This Court's analysis in Powell thus rightly focused instead on the question of whether the evidence supported such a conclusion.
In so focusing, this Court also recognized the danger of "cobbling together a set of facts that falls far short of establishing reasonable suspicion.” Id. at 183, 2011 WL 5517347, at *1. Quoting Foster, Massenburg, and United States v. Digiovanni, 650 F.3d 498 (4th Cir. 2011), as examples of cases in which "the Government failed to meet its minimal burden of articulating facts sufficient to support a finding of reasonable suspicion,” this Court vacated the judgment against Powell on the same grounds. Id. at 183, 2011 WL 5517347, at *1.
Thus, the most critical distinction from Powell is that in this case it is this Court, rather than the district court, that is “cobbling together a set of facts that falls far short of establishing reasonable suspicion.” Id. at 183, 2011 WL 5517347, *1.