895 F.2d 1423
D.C. Cir.
1990

Lead Opinion

*1424Opinion for the Court filed by Circuit Judge STEPHEN F. WILLIAMS.

STEPHEN F. WILLIAMS, Circuit Judge:

Through a routine check of passenger reservation records, officer Cook of the Amtrak police identified Leo J. Tavolacci, travelling by train from Florida to Chicago via Washington, as a potential drug courier. He and detectives Beard and Hanson of the Washington police department then watched Tavolacci as he switched trains in Union Station, and approached him at the door of his sleeper compartment soon after he boarded the train to Chicago. Asking for his ticket and identification, the officers found that he was travelling under an assumed name, and directed him to get off the train with his bag so that they could have a dog check it out. While he and officer Cook waited on the platform for the dog to arrive, Tavolacci told Cook that he had a “personal stash” in his bag, “about a quarter [of a gram?].” He offered it to Cook if he would let him go. Not surprisingly, this offer led instead to an inspection of the bag, which proved to hold about thirteen kilograms of cocaine, and to Tavo-lacci’s arrest.

A grand jury returned an indictment charging Tavolacci with possession with intent to distribute cocaine, in violation of 21 U.S.C. § 841(a)(1) & (b)(l)(A)(ii). He moved to suppress the evidence seized at his arrest on the grounds that the seizure violated his rights under the Fourth Amendment. The district court denied the motion. See United States v. Tavolacci, 704 F.Supp. 246 (D.D.C.1988). Tavolacci then pled guilty to the charge, reserving his right to appeal the district court’s ruling under Fed. R.Crim.P. 11(a)(2). After sentencing, he filed a timely appeal.

We uphold the police conduct and resulting evidence. For legal purposes the episode divides into three phases: (1) the initial encounter (up to the officers’ discovery that Tavolacci was travelling under an alias), which we uphold as not being a search or seizure at all; (2) the limited seizure thereafter (up to Tavolacci’s offer of his “stash”), which we uphold as a “stop” permissible under Terry v. Ohio, 392 U.S. 1, 88 S.Ct. 1868, 20 L.Ed.2d 889 (1968); and (3) the search of the bag, which we uphold as a search incident to a lawful arrest based on probable cause.

The Initial Encounter

We assume without deciding that if the officers’ initial dealings with Tavolacci amounted to a seizure, the information gleaned from the reservation records would not qualify as the “reasonable suspicion” needed to justify a Terry stop. So long as the encounter was not a detention implicating the Fourth Amendment, however, we must sustain it, and the use of the resulting discovery that Tavolacci was travelling under an alias, regardless of the information on which the officers acted. See, e.g., Florida v. Royer, 460 U.S. 491, 497, 103 S.Ct. 1319, 1323, 75 L.Ed.2d 229 (1983); United States v. Carrasquillo, 877 F.2d 73, 76 (D.C.Cir.1989); Gomez v. Turner, 672 F.2d 134, 140-44 (D.C.Cir.1982).

A seizure occurs “[o]nly when the officer, by means of physical force or show of authority, has in some way restrained the liberty of a citizen.” Terry, 392 U.S. at 19 n. 16, 88 S.Ct. at 1879 n. 16. The Supreme Court has framed the test as whether “in view of all of the circumstances surrounding the incident, a reasonable person would have believed that he was not free to leave.” Michigan v. Chesternut, 486 U.S. 567, 108 S.Ct. 1975, 1979, 100 L.Ed.2d 565 (1988) (internal quotations omitted); see also United States v. Brady, 842 F.2d 1313, 1314 (D.C.Cir.1988). The test assumes a citizen not only reasonable but “innocent of any crime.” Gomez v. Turner, 672 F.2d at 140 (internal quotations omitted). It further assumes that the citizen is aware of police duties to keep the peace and prevent crime, and that that “awareness, coupled with feelings of civic duty, moral obligation, or simply proper etiquette, will often lead a reasonable person to cooperate_” Id. at 141-42. (Doubtless drug couriers often “cooperate” in the hopes of simulating the reasonable innocent person and thereby shaking the police.) Among the relevant circumstances *1425are the “visibility of weapons, physical intimidation, threats, or an unusual setting or time.” United States v. Lloyd, 868 F.2d 447, 450 (D.C.Cir.1989).

The test has been criticized as “artificial” and as based on a false assumption that ordinary citizens believe they are normally free to cut police inquiries short. Edwin J. Butterfoss, Bright Line Seizures: The Need For Clarity in Determining When Fourth Amendment Activity Begins, 79 J.Crim.L. & Criminol. 437, 439 (1988). However that may be, the test allows officers to make inquiries so long as they don’t throw their official weight around unduly. It thus appears to be a rather conventional application of the idea of reasonableness, the line actually drawn by the Fourth Amendment (“people to be secure ... against unreasonable searches and seizures”). It is, in any event, the law.

Here the encounter started when detective Beard approached the open door of Tavolacci’s roomette on the train to Chicago, identified himself, and asked his permission to ask some questions. Tavolacci answered “sure.” At Beard’s request he handed him his ticket, bearing the name Tom Marks. Beard then asked if he had a photo identification. Tavolacci asked what the questioning was about. Beard responded by giving, as he put it, his “spiel,” i.e., an explanation that he was with the drug interdiction unit investigating narcotics, and he again requested an ID. At this point Tavolacci “readily went in to his right hand pocket and removed [his driver’s license].” Transcript of Suppression Hearing, July 22, 1988, at 104-05. As Tavolacci produced his license in his own name, not Marks’s, his “facial expression changed and his mouth visibly dropped,” 704 F.Supp. at 248, and he hastily volunteered an explanation for having a ticket in another’s name. Beard returned the ticket as soon as Tavolacci handed over the license.

If the contact with Tavolacci had taken place on the street or in the station, it would be clear that it was not a seizure. None of the officers displayed weapons or handcuffs, and all were in plainclothes. The time was not unusual (about 5:30 PM). There is no suggestion that the interviewing detective used any but conversational tones. He did not touch the defendant. Although three officers were present, the defendant could see at most only the interviewing detective and the head of one other, Tr. 102; the presence of two officers does not by itself transform a contact into a seizure. See, e.g., Carrasquillo, 877 F.2d at 75; United States v. Palen, 793 F.2d 853 (7th Cir.1986); United States v. Viegas, 639 F.2d 42 (1st Cir.1981) (all involving two officers).

Nor does the mere presence of officers in the doorway of a train roomette (and the adjacent aisle) defeat the “free to leave” test. United States v. Savage, 889 F.2d 1113, 1116-17 (D.C.Cir.1989); cf. INS v. Delgado, 466 U.S. 210, 219, 104 S.Ct. 1758, 1764, 80 L.Ed.2d 247 (1984). Where that position is virtually compelled by the location of the interview, it does not inherently signify an intent to prevent the interviewee from leaving the scene (much less from ending the conversation and shutting the door). See Savage, at 889 F.2d at 1116; see also Brady, 842 F.2d at 1315 n. 5 (repudiating the notion that “the narrowness and confinement of a train compartment are inherently isolationist, hence coercive”). (Although the test is an objective one, police intentions are relevant “to the extent that [they have] been conveyed to the person confronted” since the inquiry turns on the perceptions of the reasonable person. See Chesternut, 108 S.Ct. at 1980 & n. 7.)

There remains the possibility that Beard’s two-step process of first eliciting Tavolacci’s ticket and then his ID, and his retention of the ticket as he asked for the ID, could have transformed the interview into a detention requiring reasonable suspicion. A seizure is not established by a mere request for identification, see Royer, 460 U.S. at 501, 103 S.Ct. at 1326; Gomez, 672 F.2d at 142-44, nor by the initial holding and review of such documentation. On the other hand, the retention of papers under some circumstances may transform an interview into a seizure, where it is prolonged or is accompanied by some other *1426act compounding an impression of restraint. See United States v. Battista, 876 F.2d 201, 205 (D.C.Cir.1989) (court inf erred retention of license throughout extended interview including search of bag); United States v. Cordell, 723 F.2d 1283, 1285 (7th Cir.1983) (one officer handed defendant’s license and ticket to another); United States v. Waksal, 709 F.2d 653 (11th Cir.1983) (ticket and license retained at time that police asked consent to search); United States v. Robinson, 690 F.2d 869 (11th Cir.1982) (ticket and license retained at time police asked defendant if he would go to office for search); United States v. Elsoffer, 671 F.2d 1294 (11th Cir.1982) (ticket and license retained during lengthly interrogation); United States v. Elmore, 595 F.2d 1036, 1042 (5th Cir.1979) (officer took ticket out of defendant’s immediate presence); United States v. Thompson, 712 F.2d 1356 (11th Cir.1983) (while holding ticket, officer asked to look at object in defendant’s automobile).

In our case, the production of the ticket and the license could not have been separated by more than a few moments. The only act separating the requests for tickets and identification was detective Beard’s “spiel” as to his duties, delivered in response to Tavolacci’s question and without specific characterization of Tavolacci as a suspect. Such descriptions by themselves are neutral in the contact/detention inquiry. See United States v. Notorianni, 729 F.2d 520, 522-23 (7th Cir.1984) (noting that Joseph K., in Kafka’s The Trial, was not told the purpose of the investigation); see also Lloyd, 868 F.2d at 449.

Moreover, in United States v. Carrasquillo, 877 F.2d 73, 75-76 (D.C.Cir.1989), we found no seizure in a similar staggered request for documentation. There, in fact, the process was drawn out by questioning initiated by the police (whether the interviewee’s name was as it appeared on the ticket) rather than by the citizen. The former is surely more likely to create a feeling of restraint, and yet it is the latter of which Tavolacci complains. See also Savage, 889 F.2d at 1117 n. 3 (retention of ticket does not turn roomette interview into seizure); United States v. Collis, 766 F.2d 219 (6th Cir.1985) (per curiam) (no seizure found where police retained ticket and credit card while asking further questions and requesting further identification); but see Elsoffer, 671 F.2d at 1297 (finding staggered request for tickets and identification to result in a detention which the court upheld on the basis of reasonable suspicion). More generally, as police can ask simultaneously for both ticket and ID (and hold both until they have finished reading both), see Royer, 460 U.S. at 501, 103 S.Ct. at 1326 (“asking for and examining [defendant’s] ticket and his driver’s license were no doubt permissible in themselves”) (emphasis added), it would create an odd anomaly if retention of a ticket while asking for an ID created a seizure. Officers cannot be expected to script their interviews with the precision of lawyers drafting a corporate bond indenture.

Thus we cannot find a seizure merely because the officers asked for an ID after obtaining Tavolacci’s ticket, explained their official duties in response to his query, and held his ticket during the brief time needed for the follow-up request.

Off the Train and Onto the Platform

Once the officers discovered that Tavolacci’s ticket and identification were in different names, they clearly had reasonable suspicion strong enough to support a Terry stop. Besides the alias, the officers knew from the Amtrak reservation records that Tavolacci had paid $649 in cash for his ticket, that he came from a major drug source city, and that he made his reservation less than 24 hours before departure time and picked up his ticket only 32 minutes before train time. They could also draw inferences from “the look of shock” that came across Tavolacci’s face as he handed the identification to Beard. Tr. 91. Cf. United States v. Berry, 670 F.2d 583, 598 n. 17 (5th Cir.1982) (listing drug-courier profiles used in four airports, all of which include nervous behavior and three of which include the use of aliases as profile factors).

After returning the ticket and license, detective Beard asked Tavolacci if he was carrying drugs. Defendant replied that he *1427was not. Beard then asked if he could search his suitcase; Tavolacci demurred, “I’d rather not.” He likewise refused consent for a dog sniff, at which point Beard ordered him onto the platform to conduct one. 704 F.Supp. at 248. First Beard and then Hanson went off to find the dog. Left with Cook, as we have noted, Tavolac-ci asked if he could turn over his personal “stash” and leave. He made the proposal about 10 or 15 minutes after the group had moved to the platform. The train to Chicago had not yet left.

Defendant claims that he was effectively under arrest at the time he offered Cook the stash, and that until that moment the officers lacked probable cause. Assuming the offer did not constitute a new crime — a bribe attempt — which would render the statement admissible even if made during an illegal arrest, see Wayne R. LaFave & Jerold H. Israel, Criminal Procedure § 9.4(f) (1984), we still reject the claim: the detention of Tavolacci up to the moment of the offer did not exceed the bounds of a Terry stop.

The duration of a seizure is a key factor in determining whether it can be sustained on reasonable suspicion. United States v. Place, 462 U.S. 696, 709, 103 S.Ct. 2637, 2645, 77 L.Ed.2d 110 (1983). Assuming that a detention began at or about the moment that Beard ordered Tavolacci to step out onto the platform, at most fifteen minutes had passed by the time Tavolacci offered his stash and afforded the police probable cause. See 704 F.Supp. at 248 n. 5. But in United States v. Sharpe, 470 U.S. 675, 683, 105 S.Ct. 1568, 1573, 84 L.Ed.2d 605 (1985), the Supreme Court found a twenty-minute Terry stop to “clearly meet[] the Fourth Amendment’s standard of reasonableness.” See also United States v. Willis, 759 F.2d 1486, 1497-98 (11th Cir.1985) (25-minute detention found acceptable under Sharpe).

In considering the duration of a limited detention, Sharpe also found “it appropriate to examine whether the police diligently pursued a means of investigation that was likely to confirm or dispel their suspicions quickly, during which time it was necessary to detain the defendant.” Sharpe, 470 U.S. at 686, 105 S.Ct. at 1575. We think the police here demonstrated enough diligence in trying to find their canine help. Although the dog would have been waiting on the platform in a world where the time of dogs and their handlers was cost-free, we cannot say that the 10-to-15 minute delay was unreasonable, even where the police had several hours notice of defendant’s arrival. See United States v. Rutherford, 824 F.2d 831, 834 (10th Cir.1987) (upholding 25-minute stop where delay was caused by police computer problem); United States v. Quinn, 815 F.2d 153, 157-58 (1st Cir.1987) (upholding 20-25 minute stop occupied by interrogation but culminating in alert by narcotics dog that had been on scene from point of initial seizure). Significantly, the District officers went off to find the dog instead of merely waiting for it and its handler. See Sharpe, 470 U.S. at 686, 105 S.Ct. at 1575 (police conduct found diligent where time was spent locating additional officer).

Nor does the change of location from the train to the platform entail a full-fledged arrest. See Florida v. Rodriguez, 469 U.S. 1, 4-6, 105 S.Ct. 308, 309-11, 83 L.Ed.2d 165 (1984) (no seizure found where officer asked interviewee to continue questioning fifteen feet from the point of initial contact). If Tavolacci had been taken directly to an office within the station, he would have a far stronger argument. See Royer, 460 U.S. at 502-03, 103 S.Ct. at 1326-27 (officers’ seclusion of defendant in “a large closet ... a police interrogation room” key in Court’s finding a seizure requiring probable cause); United States v. Berry, 670 F.2d 583, 602 (5th Cir.1982) (en banc) (finding compelled office interview “tantamount to an arrest”); but see United States v. Sokolow, 109 S.Ct. 1581, 1584 (1989) (detention occurring in Drug Enforcement Administration’s airport office found not to require probable cause). But detention on a railroad platform, a “spacious public concourse with other travelers present,” Black, 675 F.2d at 135, clearly does not constrain the citizen anywhere near as severely as in a police office, where the citizen is cut off from the outside world and *1428deprived of its potential support (including its inhibitions on the police). And that the sniff could have been conducted on the train does not mean that the move to the platform required probable cause. See Sokolow, 109 S.Ct. at 1587 (“[t]he reasonableness of the officer’s decision to stop a suspect does not turn on the availability of less intrusive investigatory techniques”); see also Carrasquillo, 877 F.2d at 75, 77 (not a stop where police “told” defendant that they would remove his bag to the platform for a dog sniff).

Defendant argues more heatedly that because the officers had finished their questioning, “there was no investigative purpose to be fulfilled by ordering [him] off the train” along with his suitcase. Reply Brief of Appellant at 10. Thus, he claims, the order amounted to an arrest requiring probable cause. But the concept of investigatory purpose under Terry is broader than defendant assumes. Police may hold a person under Terry in order to pursue “the legitimate law enforcement interest in preventing flight in the event that incriminating evidence is found.” Michigan v. Summers, 452 U.S. 692, 702, 101 S.Ct. 2587, 2594, 69 L.Ed.2d 340 (1981).

Defendant goes on to suggest that the police could simply have let him depart on the train, alerting the Chicago authorities to his scheduled arrival. Evidently police often allow a suspect to continue an airplane journey while they inspect his baggage at an airport. See, e.g., United States v. Place, 462 U.S. 696, 698, 103 S.Ct. 2637, 2639, 77 L.Ed.2d 110 (1983); United States v. West, 731 F.2d 90 (1st Cir.1984). But even if this practice be constitutionally necessary (which we do not decide), train travel quite obviously presents a different problem: a suspect can drop off at any stop — or at a non-stop, as in the movies. We also note that Tavolacci’s train had not in fact departed at the time he offered Cook his stash and thus supplied probable cause. See 704 F.Supp. at 248 n. 5. If the police had continued to detain him without probable cause as his train rolled on to Chicago, this would be a substantially closer case; the Supreme Court has understandably recognized a “liberty interest in proceeding with [an] itinerary.” United States v. Place, 462 U.S. at 708, 103 S.Ct. at 2645. Here the detention did not interfere with defendant’s travel plans. We find it as a whole within the bounds established by Terry.

The Search

After Tavolacci offered him his stash, agent Cook directed him to accompany him to his office. Cook testified that at this point he considered defendant to be under arrest, although he refrained from any formal arrest out of concern for his safety and that of others in the station. 704 F.Supp. at 248. On the platform, they encountered detective Hanson, who had located the dog. The dog was allowed to work the bag but did not alert. Cook directed defendant to open the locked bag. After doing so, Tavolacci tried to focus the officers on one side of it, where he said the “stash” was located. Cook told him three times to back away from the bag, and defendant finally did so. The dog had meanwhile alerted to the other side of the bag, where the police then discovered over 26 pounds of cocaine. On finding the contraband, the police put defendant under formal arrest.

The district court found the search to fall within the arrest exception to the warrant requirement. We agree and cannot improve on its cogent analysis of the issue. See 704 F.Supp. at 251-53. That Tavolacci was not formally arrested until after the search is immaterial; the sequence makes no difference since police did not need the fruits of the search to establish probable cause. See Rawlings v. Kentucky, 448 U.S. 98, 111 & n. 6, 100 S.Ct. 2556, 2564 & n. 6, 65 L.Ed.2d 633 (1980). Nor is it significant that the bag had been locked and was opened by defendant only at the direction of the officers, even though it is plain that immediate search was unnecessary to assure their safety. We have held that a search qualifies as incident to an arrest so long as it is an “integral part” of the arrest. See United States v. Brown, 671 F.2d 585 (D.C.Cir.1982) (per curiam) (interpreting Supreme Court’s decision in New York v. Belton, 453 U.S. 454, 101 *1429S.Ct. 2860, 69 L.Ed.2d 768 (1981)). Thus we have upheld the search of a pouch even though it had been moved beyond the reach of the arrestee by the time of the search; it is enough, we said, if the search is confined to “containers in hand or within reach when the arrest occurs.” Id. at 587. United States v. Wright, 577 F.2d 378 (6th Cir.1978), is not to the contrary. There defendant had checked his locked suitcases hours earlier, and the only reason they were anywhere near him was that officers brought them back into his presence (but not his possession) just before his arrest. Here, defendant had control of the suitcase until moments before the search, as has been true of myriad cases upholding war-rantless searches. See 704 F.Supp. at 253 (test is whether container was in grabbing distance when subject was seized). This is one area of criminal procedure in which the courts have achieved some degree of clarity, refraining from any slippery test of actual necessity. We see no need to introduce new confusion.

Conclusion

We note that by slight modifications of their conduct the police can avert the risk that their efforts will misfire. Most important, they can make initial approaches in public areas, and tell the interviewee that he is free to call a halt. Nonetheless, while “the police procedures in this case may have been less than ideal,” 704 F.Supp. at 250, we find no infringement of Tavolacci’s constitutional rights. The judgment is affirmed.





Dissenting Opinion

*1430HARRY T. EDWARDS, Circuit Judge,

dissenting:

In Terry v. Ohio, 392 U.S. 1, 88 S.Ct. 1868, 20 L.Ed.2d 889 (1968), the Supreme Court created an exception to the probable cause and warrant requirements of the Fourth Amendment, holding that a person could be stopped and detained by a police officer for investigative purposes so long as the officer was acting pursuant to a reasonable, articulable suspicion. In Florida v. Royer, 460 U.S. 491, 103 S.Ct. 1319, 75 L.Ed.2d 229 (1983), the Court added new meaning to the investigative detention exception, when it held that a person is not “stopped” within the meaning of Terry if, in view of all the circumstances surrounding the incident, a reasonable person would have believed that he was free to leave an encounter with the police. See id. at 502, 103 S.Ct. at 1326 (quoting United States v. Mendenhall, 446 U.S. 544, 554, 100 S.Ct. 1870, 1877, 64 L.Ed.2d 497 (1980) (Opinion of Stewart, J.)). We now hold that a person who is blocked in a small train roomette by three police officers is not detained within the meaning of Terry because there was no “intent to prevent the [defendant] from leaving the scene.” If this judgment is correct, then I respectfully suggest that the so-called “free to leave” test is virtually meaningless. In my view, the majority comes perilously close to holding that armed agents of the law can effectively block the free movement of a citizen without a warrant, without probable cause, and even without articulable suspicion.

In this case, the Government concedes, and we agree, that there was no articulable suspicion to justify the police action in question. Therefore, the only question before us is whether the defendant reasonably believed that he was free to leave his encounter with the police. It seems to me that the answer is clear. Here three police officers arrived uninvited at and crowded around a traveler’s small train compartment, effectively blocking him therein, rendering him unable even to close the compartment door without having to push an officer out of the door jamb. In these circumstances no reasonable person would feel free to leave the encounter with the officers. To hold otherwise is to cast the “free to leave” test in a fashion that is wholly divorced from reality, for it defies belief to suggest that a person in the defendant’s situation could or would “slam the door” on three uninvited police officers.

“Free to leave” means complete freedom of movement, without any police obstruction, something that the defendant lost once he was blocked in his roomette by the officers. The majority advances the proposition, quoting Michigan v. Chesternut, 486 U.S. 567, 108 S.Ct. 1975, 1980 n. 7, 100 L.Ed.2d 565 (1988), that “police intentions are relevant ‘to the extent that [they have] been conveyed to the person confronted.’ ” However, the majority also acknowledges that the test for seizure is an objective one; i.e., the intent that is relevant is the one perceived by the defendant, not the one locked up in the officers’ minds. In this case the officers never stated their intentions one way or the other. If their intent was relevant at all, it must be inferred from the circumstances of the encounter. A reasonable person confronted by three uninvited officers, one of whom blocked the doorway of his small train compartment, could infer only that they did not intend to allow him to leave. In short, Chesternut does not support the majority’s conclusion that the police did not seize the defendant.

It may be that the pressures of the docket are driving our decisions in this area, especially in cases that reveal the threats of the illegal drug trade. But we cannot permit rules designed to assist police officers in responding to these threats to take on a life independent of constitutional measure. Constitutional caution must rise above fear and above even the legitimate desire to defend against societal dangers. Like other provisions in the Bill of Rights, the Fourth Amendment guarantee that police personnel cannot block and burden the movement of citizens without good reason for doing so is perhaps most urgently to be protected when it is least popular.1 Such *1431protection is the proper province of the judicial branch of government. By discharging this obligation, the judiciary helps to preserve the enjoyment of individual liberty, which this country has from its earliest days held up as its ideal.

To some there may seem to be a comfortable distance between the pressing everyday efforts of law enforcement officials to thwart drug trafficking and any threat those officials might pose to vital Fourth Amendment freedoms. Yet, it is when we ignore the proximity of the two that we come closest to destroying the balance of order and liberty that our Constitution demands. If we continue on our current course, we may soon embrace a doctrinal conclusion that armed agents of the law can block the free movement of citizens for any reason or for no reason, with or without reasonable articulable suspicion. Surely the Constitution does not permit such a result.

Notes

. Cf. Abrams v. United States, 250 U.S. 616, 630-31, 40 S.Ct. 17, 22, 63 L.Ed. 1173 (1919) *1431(Holmes, J., dissenting); Whitney v. California, 274 U.S. 357, 372-80, 47 S.Ct. 641, 647-50, 71 L.Ed. 1095 (1927) (Brandeis, J., concurring).



Concurrence Opinion

MIKVA, Circuit Judge,

concurring in part and concurring in the judgment:

I agree with my colleagues that if the officers’ initial encounter with Tavolacci amounted to a seizure, it would not be supported by the reasonable suspicion necessary to justify a Terry stop. While I concur in the judgment that the initial encounter did not constitute such a seizure, I write separately to articulate my reasons for so finding. In my dissenting opinion in United States v. Maragh, 894 F.2d 415 (D.C.Cir.1990), issued this day, I discussed at length why the rationales informing the Mendenhall test for when a seizure has occurred would best be served by adopting a clearly erroneous standard of review of trial court determinations under that test. There, I explained that this circuit should follow the course of several of our sister circuits in recognizing that deference must be paid to the trial judge, who had the opportunity to observe the testimony of the witnesses, assess credibility, interpret demeanor, and draw inferences as to the impact of the total factual circumstances of the encounter. Where subtle factors of demeanor have the potential to be critical to the outcome, appellate courts should not substitute their judgment for that of the trial court.

In this case we are faced with subtle questions, particularly what inference to draw from the fact that Mr. Tavolacci’s only means of exit, should he have desired to use it, was physically blocked by three policemen. Even if Mr. Tavolacci had no desire to leave, the question remains as to what impact this particular physical blockage coupled with the narrow surroundings would have on a reasonable person’s feelings of freedom to end the interview. The demeanor of the detective who conducted the interview in this ease is also important. As a trial judge, I might well have summed all of these facts to the conclusion of our dissenting colleague. But that call is not ours to make. The trial court found that the circumstances surrounding the initial questioning of appellant Tavolacci did not constitute a seizure and that the encounter was consensual. 704 F.Supp. 246, 249 (D.D.C.1988). Under the clearly erroneous standard, the district court’s decision must be affirmed unless upon considering the entire record we are left with a definite and firm conviction that a mistake has been made. United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 541, 92 L.Ed. 746 (1948). Because the trial court is best suited to assess the subtle factors at issue in this case, and because nothing in the record suggests that the trial court clearly erred in its conclusions, I concur in affirming its decision.