dissenting:
David Gray was convicted of murder in exchange for payment from a racketeering enterprise, 18 U.S.C. § 1959(a)(1), and of using a firearm in relation to a crime of violence, 18 U.S.C. § 924(c), largely on the basis of a confession made after his arrest. I believe that the arrest was not supported by probable cause, and the district court erred by denying Gray’s motion to suppress his confession. The government contends, disagreeing with the district judge’s finding, that the detention was supported by probable cause, and that, even if the arrest were unlawful, subsequent events sufficiently attenuated the taint so as to allow admission of the confession.
In its valiant efforts to find probable cause in this case, the en banc opinion is forced to spin and exaggerate' the underlying facts. As the majority describes the supposed “three facts,” which, “taken together, would readily lead a reasonable person to believe that it was probable that Gray was implicated in the murder,” majority op. at 770, its exaggerations stray farther and farther from the truth. The. truth, as the district judge realized, was that the arrest and interrogation were nothing more than a fishing expedir tion.
Because I believe that Gray’s arrest was not supported by probable cause and no circumstances removed the taint of the illegal arrest, I respectfully dissent.
I.
A.
To comply with the Fourth and Fourteenth Amendments, an arrest must be predicated on probable cause. See Papachristou v. City of Jacksonville, 405 U.S. 156, 169, 92 S.Ct. 839, 847, 31 L.Ed.2d 110 (1972). The probable cause requirement applies to custodial interrogations as well as formal arrests. See Dunaway v. New York, 442 U.S. 200, 216, 99 S.Ct. 2248, 2258-59, 60 L.Ed.2d 824 (1979). “Probable cause exists when the facts and circumstances known to the officer “would warrant the belief of a prudent person that the arrestee had committed or was committing an offense.’ ” Taylor v. Waters, 81 F.3d 429, 434 (4th Cir.1996) (quoting United States v. Garcia, 848 F.2d 58, 59-60 (4th Cir.1988)). “In assessing the existence of probable cause, courts examine the totality of the circumstances known to the officer at the time of the arrest.” Id. (citing United States v. Al-Talib, 55 F.3d 923, 931 (4th Cir.1995)). “Probable, cause must be supported by more than a mere suspicion, but evidence sufficient to convict is not required.” Id. (citing Wong Sun v. United States, 371 U.S. 471, 479, 83 S.Ct. 407, 413, 9 L.Ed.2d 441 (1963)).
*776In evaluating de novo whether the facts amounted to probable cause, we are to “give due weight to inferences drawn from those facts by resident judges and local law enforcement officers.” Ornelas v. . United States, 517 U.S. 690, -, 116 S.Ct. 1657, 1663, 134 L.Ed.2d 911 (1996). This does not mean that we are to defer to the judge’s or officers’ belief about whether they possessed probable cause or not. The probable cause inquiry is an objective oné, asking what reasonable officers would have believed: even if the actual officers did not believe they had probable cause, we will uphold the arrest if the “evidence was sufficient to support such a reasonable belief.” United States v. Han, 74 F.3d 537, 541 (4th Cir.), cert. denied, 517 U.S. 1239, 116 S.Ct. 1890, 135 L.Ed.2d 184 (1996). However, a “trial judge views the facts of a particular case in light of the distinctive features and events of the community____ The background facts provide a context for the historical facts, and when seen together yield inferences that deserve deference.” Ornelas, 517 U.S. at-, 116 S.Ct. at 1663.
B.
The en banc majority finds probable cause for Gray’s arrest from the wiretap “information,” ballistics evidence and a so-called identification. The district court found to the contrary. I believe that the three leads relied on by the en banc majority, even taken together, were no more than bases for suspicion. But for the en banc opinion’s exaggerating spin, it would be clear that none contributed enough to establish probable cause.
1.
The en banc opinion begins by describing supposed wiretap “information” that “focused suspicion” and “directed the investigation” to Gray:
First, the FBI and state law enforcement officers had wiretap information that a “Fat David,” which the officers believed was the street name of David Gray, had been involved in Waller’s murder____ The first piece of information that the officers had, the wiretap information, focused suspicion on Fat David. While this information supported only a suspicion and was insufficient to support a conviction, it directed the investigation solely to David Gray as a potential murderer of Black Jessie.
Majority op. at 769-770.
In truth, the wiretap lead was so vague as to be insignificant in contributing to probable cause.1 • There are doubtless many Davids, even corpulent ones, in the Baltimore area other than David Gray. The most that can fairly be said about the wiretap “information” was that “the way we [the FBI agents] interpreted the [wiretapped] conversation” was that “the name Dave or David [was] used in reference to individuals related to the killing of Jesse [Waller].” (J.A. at 288 (direct examination of Agent Sheehy).) FBI Agent Hill suggested the name of David Gray “as being a possible name for the Fat David or David that we [the FBI agents] had had reference to in the conversations.” {Id. at 289.) The state law enforcement officer in charge of the murder investigation, Detective Sydnor, knew only that
[t]hey had—someone had a wire tap, I understood, and Jesse, Black Jesse’s name had come up, also David Gray’s name had come up, and whatever reason they had put those two together, at which time for whatever reason they notified [my superi- or] who in fact said, well, Marvin [Sydnor] has a ease with a Black Jesse that is involved, at which time I spoke with them and we started putting two and two together.
(J.A. at 139 (cross-examination of Detective Sydnor).) As Detective Sydnor explained, this wiretap lead suggested that Gray might have information about the Waller murder, but did not establish Gray as a suspect:
*777Q: Now, you said that it was in April of ’94 that you first developed Mr. Gray as a suspect; is that right?
A: No, I don’t think I said that. I said in late April is when it was brought to my attention that he might have information concerning, Mr. Gray may have information concerning this investigation.
(Id. at 186.) Detective Sydnor was adamant about this point:
Q: |Y]ou developed suspicions about Mr. Gray in April of ’94; correct?
A: Yes.
Q: All right [sic]. And that was in relationship— .
A: Wait a minute, wait a minute, wait a minute. No, no. Repeat that again, please?
Q: Okay. You learned of Mr. Gray’s identity as someone who might have information in April of ’94?
A: Yes, that is correct.
(Id. at 138.)
2.
The next “fact” relied upon by the en banc opinion is the ballistics evidence. About five months after the murder, Gray was arrested in possession of a handgun that ballistics tests demonstrated was fired during the murder. However, there was no evidence suggesting when or how Gray had obtained the gun. The district court found that:
The temporal nexus between the October shooting and the March 8th seizure of the gun from Mr. Gray is simply not sufficient, in the judgment of this Court, to support the inference that Mr. Gray was in continuous possession of that handgun for a period exceeding six months2 when it was seized from him on March 8,1994.
The real world simply tells us, and all you have to do is read the newspaper every day to know, that handguns', unfortunately, change hands quite rapidly, particularly in the inner city. They change hands because of burglaries, because of thefts, because of illegal sales and because of legal sales.
(J.A. at 501 (ruling on suppression motion).)
The trial judge’s rejection of the inference, that because Gray possessed the gun in March it was reasonably likely that he had used it in October, was informed by his knowledge of background facts about the Baltimore community in which he lives and practices. The Supreme Court has recognized that á trial judge views the facts through a lens of knowledge of the community, see Ornelas, 517 U.S. at-, 116 S.Ct. at 1663 (“The background facts, though rarely the subject of explicit findings, inform the judge’s assessment of the historical facts.”), and has explained to us that these “inferences ... deserve deference,” id. “[A] reviewing court should take care both to review findings of historical fact only for clear error and to give due weight to inferences drawn from those facts by resident judges and local law enforcement officers.” Id.
Instead of obeying the Supreme Court’s directive, the en banc court merely repeats the other “facts” that it believes supported probable cause, and asserts that the “combination of information raises a far greater probability of Gray’s involvement than that derived from his mere possession of a handgun that had earlier been used in a crime.” Majority op. at 771. In order to create something out of nothing, however, the en banc opinion is forced further to recharacterize and spin the facts. The wiretap lead, which suggested only that someone named David might have useful information about the murder, is now described as “information from wiretaps that it was Gray who was involved in Waller’s murder.” Majority op. at 771. And the en banc opinion characterizes the uncertain results of a photo array selection as “additional corroborative information,” a claim no elosér to the truth. Id.
3.
The third “fact” relied upon by the en banc opinion to establish probable cause is the photo array “identification.” The two surviv*778ing. victims of the shooting were shown a photo array, which included a photograph of Gray, and were asked to try to identify their attackers. One of the victims, Antoine Little, could not identify anyone from the array. The other, Tracy Ward, “looked at it [the array] and reluctantly said yeah, this guy resembles the guy.” (J.A. at 228 (cross examination of Detective Sydnor).) “Resembled” was the term actually used by Ward. (Id. at 230-31.) Detective Sydnor did not believe that this was a positive identification. (Id. at 228.) In fact, he used the word “reluctantly” to distinguish it from a more certain identification:
Q: You say reluctantly. What do you mean by reluctantly?
A: Well, there’s some people you’ll give the photo card, that’s the person, that’s the person. He didn’t do that. He looked and he was—look in his eye, again was looking at Gray’s photo, and said yeah, that resembles the guy.
(Id.) After .this reluctant selection, a photograph was again shown to Little, but he still could not identify Gray. (J.A. at 116 (direct examination of Detective Sydnor).)
The district court properly characterized these events as an “exceedingly tentative identification.” (J.A. at 502 (ruling on suppression motion).) I agree that the identification “adds absolutely nothing to the probable cause equation.” Id. However, the en banc opinion exalts this identification to the level of “additional corroborative information,” combining with the other “facts” to support probable cause. Majority op. at 771.
4.
The police officers in this case began with mere speculation and guesswork about vague hints in a wiretapped conversation.. They gathered some ballistics evidence, but its timing undermined its usefulness. To this they added an identification so tentative as to be practically worthless. Although piecing together these three clues was a good first step,3 the officers should have followed up their hunches with more investigation. Because those hunches did not amount to probable cause, the warrantless arrest was illegal. “Arresting a person on suspicion, like arresting a person for investigation, is foreign to our system” and is not acceptable behavior. Papachristou, 405 U.S. at 169, 92 S.Ct. at 847.
The Supreme Court “has required that the existence of probable cause be decided by a neutral and detached magistrate whenever possible.” Gerstein v. Pugh, 420 U.S. 103, 112, 95 S.Ct. 854, 862, 43 L.Ed.2d 54 (1975). Here there were no exigent circumstances requiring a warrantless arrest. There was no suggestion that all magistrates were unavailable. The police had ample opportunity to apply for a warrant from a judicial officer. When the police officers decided not to seek a warrant, but “to go get [Gray], bring him in, interview him, see what he had to say,” (J.A. at 304 (direct examination of Agent Sheehy)), they violated his rights guaranteed by the Fourth and Fourteenth Amendments. Because the police did not have probable cause to believe Gray had been involved in the crime, his arrest for the purpose of interrogation was unlawful. See Davis v. Mississippi, 394 U.S. 721, 726-27, 89 S.Ct. 1394, 1397-98, 22 L.Ed.2d 676 (1969). In this regard I agree with the district judge’s conclusion.
II.
Having erroneously found probable cause for the arrest, the en banc opinion does not even address the next question—whether the confession is admissible despite the unlawful arrest because the taint of the arrest had been dissipated by the time Gray confessed. That is the point on which the district court relied to admit the confession. However, it is a legal determination subject to de novo review. I believe that the confession is inad*779missible because the taint was not sufficiently dissipated.
When a confession is the fruit of an illegal arrest, the exclusionary rule prevents the prosecution from introducing it into evidence “un less intervening events break the causal connection between the illegal arrest and the confession so that the confession is sufficiently an act of free will to purge the primary taint.” Taylor v. Alabama, 457 U.S. 687, 690, 102 S.Ct. 2664, 2667, 78 L.Ed.2d 314 (1982) (internal quotation omitted); see also Dunaway v. New York, 442 U.S. 200, 217-18, 99 S.Ct. 2248, 2259-60, 60 L.Ed.2d 824 (1979); Brown v. Illinois, 422 U.S. 590, 602, 95 S.Ct. 2254, 2261, 45 L.Ed.2d 416 (1975). Although the question of whether the taint has been purged turns on the individual facts of each case, the Supreme Court has established an analytical framework to guide courts making that determination. See Brown, 422 U.S. at 603-04, 95 S.Ct. at 2261-62. As a threshold matter, in order to be admissible, the confession must be voluntary within the meaning of the Fifth Amendment, but voluntariness is, by itself, insufficient to attenuate the taint of an illegal arrest. See Taylor, 457 U.S. at 690, 102 S.Ct. at 2666-67; Dunaway, 442 U.S. at 217, 99 S.Ct. at 2259; Brown, 422 U.S. at 604, 95 S.Ct. at 2262. Miranda warnings may indicate that a confession is voluntary for the purpose of the Fifth Amendment, but such warnings do not remove the taint of an illegal arrest. See Dunaway, 442 U.S. at 217, 99 S.Ct. at 2259; Brown, 422 U.S. at 602-03, 95 S.Ct. at 2261-62.
If the confession was voluntary, the next step which must be addressed is to determine whether the causal connection between the arrest and the confession was broken, paying particular attention to the factors set forth in Brown: “the temporal proximity of the arrest and the confession, the presence of intervening circumstances, and, particularly, the purpose and flagrancy of the official misconduct.” Brown, 422 U.S. at 603-04, 95 S.Ct. at 2261-62. If these factors indicate that the confession was obtained by exploitation of the unlawful arrest, the confession must be suppressed. Id. at 600-04, 95 S.Ct. at 2260-62. The prosecution bears the burden of demonstrating that the confession is admissible. Id. at 604, 95 S.Ct. at 2262.
As the district court properly held, Gray’s confession meets the threshold requirement of voluntariness. He confessed after the officers explained his Miranda rights and after he knowingly and intelligently waived those rights. Although some of the officers talked with Gray before the Miranda warnings were administered, the district court found, as a factual matter, that they merely engaged in small talk and did not discuss the Waller murder. Also, the district court excluded all statements made before Gray waived the Miranda warnings.
Because the confession was given voluntarily, I move on to examine the factors set forth in Brown. This examination reveals that the confession was obtained by exploitation of the illegal arrest. First, the confession was made in close temporal proximity to the arrest, coming less than three hours after Gray was arrested. See, e.g., Taylor, 457 U.S. at 691, 102 S.Ct. at 2667-68 (excluding confession made six hours after illegal arrest); cf. Wong Sun v. United States, 371 U.S. 471, 491, 83 S.Ct. 407, 419, 9 L.Ed.2d 441 (1963) (admitting unsolicited confession made voluntarily several days after illegal custody ended). '
‘ Next, no intervening events of any significance broke the causal connection between the arrest and the confession. Before Gray made the confession, he was not released from illegal custody. See id. (finding release from custody to be an attenuating intervening circumstance). Nor did he consult with counsel. See Carnejo-Molina v. I.N.S., 649 F.2d 1145, 1149 (5th Cir.1981) (finding consultation with counsel to be an attenuating intervening circumstance). Instead, he was simply brought to the police station, informed of his rights,4 and interrogated.
Finally, the investigators’ misconduct in this case is indistinguishable from the misconduct at issue in Dunaway and Taylor. In *780this case, as in those cases, “the police effectuated an investigatory arrest without probable cause ... and involuntarily transported petitioner to the station for interrogation in the hope that something would turn up.” Taylor, 457 U.S. at 693, 102 S.Ct. at 2668-69. Arrests made for investigatory purposes without probable cause are precisely the type of police misconduct prohibited by Brown and its progeny. See Dunaway, 442 U.S. at 215, 99 S.Ct. at 2258; Brown, 422 U.S. at 605, 95 S.Ct. at 2262-63.
The government attempts to distinguish this case from Supreme Court precedent on three grounds: the police did not employ coercive interrogation methods, Gray’s sophistication makes it more likely that his confession was an act of free will, and the police believed in good faith that they had probable cause to detain Gray. None of these facts, whether considered separately or cumulatively, demonstrate a break in the causal connection between the arrest and the confession. Gray’s sophistication and the absence of coercive interrogation methods merely support the conclusion, which I have already reached, that the confession was made voluntarily. As previously stated, although voluntariness is a threshold requirement of admissibility, it is by itself insufficient to show a break in the causal connection between the arrest and the confession. Dunaway, 442 U.S. at 219, 99 S.Ct. at 2260. As the Court put it in Taylor, “[t]he fact that police did not physically abuse petitioner, or that the confession they obtained may have been ‘voluntary’ for purposes of the Fifth Amendment, does not cure the illegality of the initial arrest.” 457 U.S. at 693, 102 S.Ct. at 2668.
The government does not explain, and I do not see an explanation, how the officers’ good faith interrupts the causal connection between the illegal arrest and the confession. It is well established that the officers’ subjective-state of mind does not affect whether an arrest was lawful. Beck v. Ohio, 379 U.S. 89, 97, 85 S.Ct. 223, 228-29, 13 L.Ed.2d 142 (1964). In addition, the Supreme Court has expressly declined to adopt a good faith exception to the exclusionary rule in this context. Taylor, 457 U.S. at 693, 102 S.Ct. at 2668-69.
Because all three of the Brown factors weigh against the government, the causal connection between the illegal arrest and the confession was not sufficiently attenuated to permit the admission of the confession. In Taylor, the Court observed: “This ease is a virtual replica of both Brown and Dunaway. Petitioner was arrested without probable cause in the hope that something would turn up, and he confessed shortly thereafter without any meaningful intervening event.” 457 U.S. at 690-91, 102 S.Ct. at 2666-68. The same is true of Gray’s case.
III.
For the foregoing reasons, I respectfully dissent.
. It was not mere oversight nor coincidence that the supposed "wiretap information” was wholly ignored by the arresting officers in their explanation of their basis for believing that they had probable cause, (J.A. at 259 (redirect examination of Detective Sydnor)), by the district judge in his finding that there was no probable cause for • the arrest, (J.A. at 500-02 (ruling on suppression motion)), and by the government in its meager defense of the arrest, see Brief of Appellee at 12-13.
. In fact there were slightly less than five months between the October 19, 1993, shooting and the March 8, 1994, arrest.
. At least, in general, the police work was tenacious, although not as good as the majority seems to think. The majority claims that the police found Gray only "[ajfter looking for him unsuccessfully for weeks.” Majority op. at 768. However, Detective Sydnor testified that he "never, never went looking for him [Gray],” despite his supposed belief that there was probable cause to arrest Gray for murder. (J.A. at 262 (redirect examination of Detective Sydnor).) In fact, the district judge was baffled at this apparent dereliction of duty. (Id. at 260-62.)
. As explained above, the giving of Miranda warnings is not a significant intervening event. See Taylor, 457 U.S. at 691, 102 S.Ct. at 2667-68.