Lead Opinion
SECOND AMENDED OPINION
On аppeal from his conviction for possessing cocaine with intent to distribute in violation of 21 U.S.C. § 841(a)(1), Andrew Sokolow challenges the district court’s ruling denying his motion to suppress evidence obtained during the retention of himself and his luggage at the Honolulu airport. The evidence he moved to suppress included 1,000 grams of cocaine. Exercising appellate jurisdiction under 28 U.S.C. § 1291, we initially reversed his conviction on the ground that agents of the Drug Enforcement Agency had violated the Fourth Amendment in detaining Sokolow and searching his luggage.
I.
On Sunday, July 22, 1984, Sokolow purchased two roundtrip tickets to Miami at the United Airlines counter at Honolulu Airport. Sokolow paid for the $2100 tickets in cash with approximately half of a large wad of $20 bills he was carrying, purchasing them under the names of Andrew Kray and Janet Norian. The ticket agent notified drug task force agеnt John McCarthy of the purchase. Agent McCarthy called the telephone number given to the ticket agent by Sokolow. The call was answered by a recorded message on an answering machine. Upon listening to a tape of this message, the ticket agent identified the voice as that of Sokolow. Agent McCarthy determined that the number was subscribed to by Karl Herman at 348-A Royal Hawaiian Avenue, Honolulu, Hawaii. What Agent McCarthy apparently did not know at this time was that both Herman and Sokolow lived at this address. On July 24, Agent McCarthy learned that Sokolow and Janet Norian were scheduled to return to Honolulu the following day on a flight with a layover in Los Angeles. On July 25, agents at the Los Angeles airport reported that during his layover Sokolow “appeared to be very nervous and was looking all around the waiting area”
Traveling with carry-on luggage only, Sokolow and Norian arrived at Honolulu airport and proceeded directly to the street to hail a taxi. They were at curbside waiting for a taxi when, at approximatеly 6:41 p.m., several Drug Enforcement Administration (DEA) agents approached them. As found by the district court, the agents grabbed Sokolow by the arm, pulled him onto the walkway, and sat him down. Agent Kempshall then asked Sokolow for his airline ticket and identification. Sokolow responded that he was not carrying any identification and did not have his airline ticket. Sokolow further stated that, although his name was Sokolow, he was using his mother’s maiden name of Kray, and that he had not made the reservations himself. Sokolow, Norian, and their lug
In the DEA office, the luggage was turned over to a Customs Service dog handler for examination by a narcotics detector dog. The narcotics detection dog alerted to a brown shoulder bag. Based on this information, the agents placed Sokolow under arrest and proceeded to secure a warrant to search the shoulder bag. Although the search uncovered no drugs, it did uncover certain papers that prompted the agents to have the narcotics detection dog reexamine the remaining three pieces of luggage. This time the dog alerted to a medium-sized carry-on bag. Ultimately, another narcotics dеtection dog confirmed this alert. The agents searched the medium-sized bag pursuant to a warrant and found 1,000 grams of cocaine. Sokolow was indicted for possession with intent to distribute cocaine in violation of 21 U.S.C. § 841(a)(1). The district court denied his motion to suppress all statements and evidence secured pursuant to his seizure, his arrest, and the search of his luggage. Sokolow entered a conditional guilty plea under Fed.R.Crim.P. 11(a)(2), thereby preserving his right to challenge the district judge’s ruling on his Fourth Amendment claims. Concluding that reversal was a possibility because the case was a “close one,” the district court granted Sokolow bail pending appeal.
II.
A.
Resolution of the Fourth Amendment issues presented by this appeal requires a close analysis of the DEA agents’ actions in detaining Sokolow and detaining and. searching his luggage. We begin with the initial contact between the agents and Sokolow at curbside. Without making any specific findings of fact, the district court originally ruled that the initial contact between the agents and Sokolow at curbside did not rise to the level of a seizure, citing Florida v. Royer,
We review de novo the question whether a seizure occurred. See LaDuke v. Nelson,
Although not all seizures require probable cause, “any curtailment of a person’s liberty by the police must be supported by at least a reasonable and articulable suspicion that the person seized is engaged in criminal activity.” Reid v. Georgia,
The agents knew the following facts matching their “drug courier profile” when they first approached Sokolow:
III.
In Terry v. Ohio,
In assessing whether a given set of facts constitutes reasonable suspicion, we must determine whether the facts collectively establish reasonable suspicion, not whether each particular fact establishes reasonable suspicion. “[T]he totality of the circumstances — the whole picture — must be taken into account.” United States v. Cortez,
Although the government may present a lengthy list of detailed observations, the courts are not relieved of their duty to review the list critically and decide whether each particular observation cited actually contributes something to the “whole picture” — that is, whether the particular observation bears any reasonable correlation to a suspicion that the person presently is engaged in criminal activity. See Erwin,
What, then, may constitute reasonable suspicion, to justify a seizure? We reaffirm our earlier conclusion that, even
The “mosaic” presented by the government on petition for rehearing, even with the addition of evidence of Sokolow’s nervousness in transit, fails to form an image of ongoing criminal activity. Instead, we see a vaguer shape resulting from the improper attempt to define not ongoing criminal activity but a class of people that is predominantly criminal. The Supreme Court, by contrast, requires that the reasonable suspicion supporting an investigative stop be reasonable suspicion of an ongoing crime, and thus forecloses the result requested by the government. We will support our conclusion with an analysis of Terry and its application to the use of criminal “profiles.” We then discuss the development of Terry principles in profile-stop cases. We finally reaffirm our earlier conclusion that the profile in this case failed to provide reasonable suspicion for a seizure of Sоkolow.
A.
Terry concerned a policeman’s stop of several persons the officer believed were “casing” a store for a robbery. The Court accepted that the officer had reasonable suspicion for a stop, and proceeded to discuss at greater length the reasonableness of his weapons frisk in light of his concern for his own safety from possible concealed weapons.
The factual basis for the Court’s brief conclusion is illustrative. The Court found that the defendants’ behavior, when viewed through the trained eye of a police officer, was “consistent with [the officer’s] hypothesis that these men were contemplating a daytime robbery.” Id. at 28,
The drug-courier profile, if used as a measure of reasonable suspicion, operates in a different manner than did the officer’s trained evaluation that warranted the stop in Terry. Profile elements include aspects of a suspect’s behavior that clearly are consistent with an ongoing crime, such as when a suspect uses an alias in travel or when a suspect takes an evasive or erratic path through an airport in a manner that demonstrates a desire to avoid detection. Traveling under an alias or evasive movements are part of the performance of the crime. These elements of the profile demonstrate behavior that, absent unusual circumstances,
Othеr elements of the profile, however, seek to identify personal characteristics shared by drug couriers and the public at large, but which, when present in sufficient number, arguably serve to identify drug couriers. Such elements as nervousness, a destination or arrival including a “drug source” or “drug reception” city, manner of attire, time of flight, and position among the disembarking passengers, among myriad others, describe cross-sections of the people who use planes, but without any indication that those cross-sections are predominantly, or even mainly, engaged in an ongoing crime. By themselves, they indicate no ongoing criminal enterprise, but attempt to identify an individual as the type of person who may engage in a criminal enterprise, based upon stereotypes of drug courier appearance or behavior.
An officer attempting to justify a seizure based solely upon these aspects of the profile necessarily must present a different type of testimony than the factors discussed earlier. The officer must demonstrate that the combination of behavior exhibited by the suspect, although not directly probative of оngoing criminal behavior, is unlikely to exist in innocent persons or, perhaps, in some “significant” percentage of innocent persons.
Comparative examples may help. An officer in the first type of case will testify that, in accord with the sharpened understanding of his or her profession, a suspect’s travel under an alias or meandering path through an airport demonstrate behavior indicative of an ongoing criminal activity. While probable cause for arrest does not yet exist, the most reasonable explanation of such activity may be that a criminal enterprise is in progress.
In this type of case, the traditional focus on criminal activity shifts to a focus on the personal characteristics of the individual under scrutiny. Not only is this transfer of focus impermissible, its accuracy is often uncorroborated. Here, an officer must testify that a pattern of behavior, otherwise explicable as innocent behavior, does not exist in a significant number of innocent people. The officer testifies not about his own trained observation of a criminal activity, but instead about the probability that drug couriers generally exhibit certain external characteristics. Unfortunately, the testimony seldom is constructed in that extended a fashion. Empirical documentation would be necessary for the assertion that, for example, the class of nervous, cash-paying travelers to Miami does not include significant numbers of innocent persons. The court is left to evaluate not the reasonableness of an officer’s assessment of facts demonstrating an ongoing criminal enterprise, but the probabilistic evidence (compiled from cases not before the court) that indicates that “innocent” behavior is not so innocent.
We have already demonstrated, by reference to Judge Becton’s painstaking and insightful examination of the profile, see supra p. 1418, the unusual vicissitudes that the profile may undergo in the hands of different agents and when justifying different seizures. The very transmutability of the profile demonstrates that it fails to justify a Fourth Amendment seizure. When the focus is away from facts indicating ongoing criminal activity and instead upon innocent behavior in which criminals may engage, virtually anything may support “reasonable” suspicion.
We do not believe that our conclusion destroys the mosaic by rеjecting tile after tile. We merely recognize that the mosaic that may cause a trained officer of the law to investigate further is not the same mosaic that creates reasonable suspicion to allow a Fourth Amendment seizure. Some distinction between investigation and detention must remain.
Our formulation of the reasonable suspicion standard also allows for the continued use of the profile in justifying reasonable suspicion. Accumulating a certain number of positive responses to elements of the profile would not create reasonable suspicion. Profile elements that indicate ongoing criminal behavior, however, may be supplemented by other aspects of the profile that would not by themselves create reasonable suspicion. Hence, a traveler under an alias arriving from Miami may arouse a greater suspicion than a traveler arriving under an alias from Dubuque. All of these supporting factors, however, should be demonstrated by more statistical evidence than “common knowledge.”
We thus believe that reasonable suspicion must be founded upon evidence of ongoing criminal behavior. Probabilistic evidence, absent more, is insufficient to create reasonable suspicion. Such evidence, when sufficiently documented, may serve to confirm or deny reasonable suspicion based upon evidence of ongoing criminal behavior. We further believe that this interpretation of reasonable suspicion is not new, but has been announced by the Supreme Court in its examinations of the profile.
B.
We believe that our interpretation of the reasonable suspicion standard is in harmony with the Supreme Court’s pronouncements regarding the profile. The Court has consistently looked beyond the profile to determine whether a reasonable suspicion exists of a criminal enterprise. Searches based solely on the personal characteristics of a suspect have been rejected as unreasonable.
The Supreme Court first encountered the profile in United States v. Mendenhall,
Although Justice Stewart noted that the initial confrontation was based upon the “so-called ‘drug courier profile’ — an informally compiled abstract of characteristics thought typical of persons carrying illicit drugs,” id. at 547 n. 1,
The court again considered the drug-courier profile later in the 1979 Term. In Reid v. Georgia,
[o]f the evidence relied on, only the fact that petitioner preceded another person and occasionally looked backward at him as they proceeded through the concourse relates to their particular conduct. The other circumstances [drug source city departure, eаrly-morning arrival, no luggage other than shoulder bags] describe a very large category of presumably innocent travelers, who would be subject to virtually random seizures were the Court to conclude that as little foundation as there was in this case could justify a seizure.
Id. at 441,
Florida v. Royer,
The facts leading up to the detention of Royer are far different than those at hand here, and Royer quite simply cannot serve as a simple equation of the drug-courier profile with reasonable suspicion. Royer was not seized (as was Sokolow) when initially approached, and therefore the officers properly discovered that he was traveling under an alias, since the name they observed on his luggage was different that the name on his license and ticket. It was this indication of ongoing criminal activity, and not the factors of the profile noted by the Court, see id. at 493 n. 2,
when the officers discovered that Royer was traveling under an assumed name, this fact, and the facts already known to the officers — paying cash for a one-way ticket, the mode of checking the two bags, and Royer’s appearance and conduct in general — were adequate grounds for suspecting Royer of carrying drugs and for temporarily detaining him and his luggage while they attempted to verify or dispel their suspicions in a way that did not exceed the limits of an invеstigative detention.
Id. at 502,
In this case, as opposed to Royer, there was no evidence of ongoing criminal activity, specifically the use of an alias. The fact that Sokolow’s voice was on an answering machine for a phone listed under the name of his roommate Karl Herman does not provide a basis for suspecting that Sokolow was using an alias. In contemporary society it is not unusual for persons with different last names to share a common residence and telephone. Nor is it unusual for a member of the same house
Judge Becton identifies a final Supreme Court drug-courier profile case in Florida v. Rodriguez,
In summary, the Court сonsistently has required that an officer’s suspicion be supported by evidence of ongoing criminal activity. Such behavior cannot be intuited from a hodgepodge assembly of “factors” about individual character rather than criminal acts. It must demonstrate the ongoing commission of a crime. The list of such actions would be extensive. Evasion and traveling under an alias are among the most common. Until such evidence of an ongoing crime is developed, however, other factors are irrelevant to reasonable suspicion. Once evidence of ongoing criminal behavior is produced, then other factors may confirm or contradict the reasonableness of suspicion, but other factors may not alone support a stop.
C.
We are left to apply our interpretation to the seizure of Sokolow at the curbside of the Honolulu airport. The six factors proffered by the government fail to indicate an ongoing criminal enterprise. The government suggests that Sokolow’s nervousness while awaiting a connecting flight in Los Angeles, when considered in the context of the other evidence, presented a reаsonable suspicion of ongoing criminal activity. There is no evidence on the record to indicate that Sokolow’s nervousness was indicative of an attempt to evade detection. Cf. Mendenhall,
The government assures us that “[t]he combination of facts in this case will rarely, if ever, describe an innocent traveler.” Gov’t Pet. for Reh’g at 13. The obvious lack of substantiation for this claim betrays its lack of merit. That courts have ultimately found no Fourth Amendment violation in cases that mention these factors can
CONCLUSION
Since we do not believe that the government has demonstrated that it had a reasonable suspicion that Sokolow was engaged in criminal activity, we reverse the district court and order suppression of the evidence illegally seized. We therefore do not reach questions related to the acts subsequent to Sokolow's curbside detention.
REVERSED and REMANDED.
Notes
. Transcript of Remand Hearing of September 22, 1986, at 20, 36, 67. The testimony that the Los Angeles agents made this report to the Honolulu agents was adduced during an evidentiary hearing conducted by the district judge in response to our limited remand for additional findings of fact on specified issues. The district judge’s supplemental findings of fact and conclusions of law were forwarded to this court, but the reporter’s transcript of the evidentiаry hearing was not made part of the appellate record until the government filed its petition for rehearing.
. After argument, we vacated submission and remanded to the district court for additional findings of fact on specified issues.
. We accept as not clearly erroneous the findings of fact upon which the district court based its conclusion that reasonable suspicion existed.
. The district court based its conclusion that the initial curbside stop was supported by a founded suspicion in part on the facts that Sokolow admitted he was not traveling under his real name and that he told the agents he did not have his ticket even though he had just gotten off the plane. We disagree. The initial seizure, which we hold occurred when the agents first grabbed Sokolow, must be based upon a reasonable and articulable suspicion that existed at that time. It cannot be based on information that is a fruit of the seizure itself. See United States v. Erwin,
. In our original opinion, we held that a report that simply characterized Sokolow’s behavior during his layover in Los Angeles as "suspicious” was too conclusory to constitute articulable information supporting a reasonable suspicion absent аny description of that "suspicious” behavior.
. The range of issues addressed by the Supreme Court via a Terry-type analysis has broadened greatly. See, e.g., United States v. Montoya de Hernandez,
. See, e.g., United States v. Mendenhall,
. In Judge Wiggins’s dissent to United States v. Erwin,
. Of course, the officer may approach the suspect prior to "reasonable suspicion,” so long as his approach does not constitute a seizure. See Florida v. Rodriguez,
. That the plurality in Royer mentioned nervousness as part of the profile,
Dissenting Opinion
dissenting:
If the present opinion becomes the law of this circuit, I conclude that many, and perhaps all, Terry stops that rely upon drug courier profile characteristics may fail on constitutional grounds. Because I believe such an unfortunate result is not mandated by the law, I respectfully dissent.
The fourth amendment prohibits unreasonable searches and seizures. Courts determine the reasonableness of a search or seizure by “balancing the need to search against the invasion which the search entails.” Camara v. Municipal Ct.,
I agree with the majority that a fourth amendment seizure occurred in this case when the DEA agents approached defendant Sokolow at the curbside. A person is seized when- his freedom of movement is restrained by means of physical force or a show of authority. See United States v. Mendenhall,
I believe that the majority’s approach seriously undermines the effectiveness of the drug courier profile as an investigative tool. The DEA developed the drug courier profile in the early 1970’s as part of an effort to combat escalating drug smuggling through this nation’s airways. The profile consists of a number of characteristics, all of which, both singly and collectively, are in themselves lawful. The DEA has determined, however, based upоn its collective experience, that certain characteristics are commonly associated with drug couriers.
Because conformance with some aspects of the profile could “describe a very large category of presumably innocent travelers,” Reid v. Georgia,
The majority contends that its mechanical view of the reasonable suspicion standard is “in harmony” with Supreme Court cases discussing the drug courier profile. This reading of the Supreme Court cases is unjustified. Although the Court has not delineated any explicit guidelines, it has approved the use of drug courier profile characteristics to support investigative stops. The Supreme Court first discussed the profile in United States v. Mendenhall,
Justice Powеll's concurring opinion, joined by Chief Justice Burger and Justice Blackmun, assumed that the initial encounter constituted a seizure, but found that it was justified by reasonable suspicion. This finding was based on the fact that Mendenhall met the following profile characteristics: 1) she arrived in Detroit from Los Angeles, a source city, 2) she appeared nervous, “engaged in behavior that the agents believed was designed to evade detection,” 3) she was the last to deplane, 4) she claimed no luggage, and 5) she changed airlines for her flight out of Detroit. Id. at 564,
The Supreme Court again considered the constitutionality of airport drug stops in Reid v. Georgia,
Subsequently, in Florida v. Royer,
The plurality relied on the following facts to support its finding of reasonable suspicion: 1) Royer was traveling under an assumed name, 2) he paid cash for a one-way ticket, 3) he wrote only a name and destination on his baggage identification tags, and 4) his general appearance and conduct were unusual. Id. at 502,
The majority in the present case contends that Royer’s use of an alias was the requisite evidence of ongoing criminal activity that justified the stop; the other factors were then only relevant once the alias was discovered. This gloss on Justice White’s opinion is wholly unsupportable. Justice White referred to the use of an alias in the reasonable suspicion equation, but he in no way suggested that the alias is the central or absolutely necessary factor that makes all the other factors relevant. Moreover, Justice White expressly rejected the view of the Florida District Court of Appeals that “a mere similarity with the contents of the drug courier profile is insufficient even to constitute ... articulable suspicion.” Id. at 495 n. 7,
I conclude, therefore, that a fair reading of Supreme Court precedent is that conformity with drug courier profile characteristics does not automatically provide reasonable suspicion, but can justify an investigative stop in a particular case. The fourth amendment requires a case-by-case determination, and on the facts of the case before us the DEA agents possessed reasonable and articulable suspicion that Sokolоw was engaged in narcotics trafficking. The agents knew the following facts when they initially approached Sokolow at the airport curbside: 1) Sokolow had just returned to Hawaii from a three-day trip to Miami, a well known source city for drugs, 2) Sokolow paid for his ticket with cash from a large roll of twenty dollar bills, 3) neither Sokolow nor his companion checked any luggage, 4) Sokolow appeared very nervous and looked all around the waiting area during his layover, and 5) Sokolow used the name Andrew Kray on his airline tickets, but his voice was on an answering machine at a phone subscribed to by Karl Herman.
Viewed collectively, these articulable facts were sufficiently suspicious to justify a brief and minimally intrusive investigative detention. Of particular importance is the fact that Sokolow paid $2100 for his tickets from a stack of twenty dollar bills approximately double that amount. Innocent persons do not characteristically carry thousands of dollars in twenty dollar bills on their persons. Certainly, this cash pay
Further, Sokolow’s travel pattern belies-any assumption that he traveled to Miami for a pleasure trip. He flew from Honolulu to Miami, a well known drug source city, on July 22 and returned on July 25, a very short time considering that it takes a minimum of ten hours to travel each way. Also, neither Sokolow nor his companion checked any luggage. These relatively anomalous characteristics enhance the suspicious circumstances of the large cash purchase of the tickets.
Finally, Sokolow was nervous during his layover in Los Angeles and glanced all around the waiting area, hardly a sign of an innocent traveler. The Supreme Court has recognized that nervousness can contribute to the reasonable suspicion calculation. See Royer,
In Erwin, this court concluded that arrival from a drug source city after a one day stay with only carry on luggage required further particularized evidence to establish founded suspicion. The Erwin court concluded that defendant’s nervous behavior, circuitous route through the airport, and possible effort to conceal the truth fulfilled the requirement of particularized evidence. Id. at 1511. I submit that Sokolow’s payment for his airline ticket in thousands of dollars in twenty dollar bills, coupled with his unusual travel pattern and nervous behavior are more objectively suspicious than the facts on which the Erwin court relied.
My conclusion that Sokolow’s seizure at curbside was supported by reasonable suspicion requires me to address Sokolow’s additional arguments. Sokolow contends first, that investigative detention ripened into an arrest once he was moved to the DEA office, and second, that the retention of the three pieces of luggage the narcotics dog did not alert to exceeded the permissible limits of an investigative stop. I find that the detention of Sokolow and his luggage in the DEA office for the purpose of subjecting his luggage to a dog sniff was a permissible continuation of the investigative stop and did not constitute a de faсto arrest.
The scope of a stop is reasonable if the officers’ action is “reasonably related in scope to the circumstances which justified the interference in the first place.” Terry,
Even a valid detention can become excessive if it lasts longer or is more intrusive than necessary. See Sharpe,
Contrary to Sokolow’s position, the decision in Royer does not dictate a different result. The Court found in Royer that the officers’ conduct in removing Royer to a small, windowless room for interrogation was “more intrusive than necessary to effectuate an investigative detention ...”
Once the narcotics dog alerted to the brown shoulder bag, the DEA agents had probable cause to arrest Sokolow. See Royer,
This case, unlike Place, did not involve an investigative detention, but rather the detention of a person’s luggage as incident to his lawful arrest. Thus, the luggage detention did not implicate any of the fourth amendment interests discussed in Place. First, Sokolow’s privacy interest in the contents of his luggage was not impaired by the agents simply holding his luggage. Cf. id. at 706-07,
After the search of the shoulder bag yielded no apparent narcotics, the probable cause to hold Sokolow and his remaining luggage evaporated. The subsequent detention of the luggage for a second dog sniff, however, was justified by new grounds for reasonable suspicion that arose after the first dog sniff. The government contends that this suspicion existed on the basis of the following: 1) Sokolow had been identified as a cocaine customer, 2) Sokolow made a statement that he was in big trouble, 3) he possessed other used airline tickets to Miami and ho
In sum, the majority’s approach effectively throttles the efforts of drug enforcement agents to combat escalating narcotics trafficking. The fourth amendment protects against unreasonable searches and seizures. In my view, it is entirely reasonable for agents to detain and question a suspected drug courier briefly based upon a rational profile. Therefore, the district court did not err in refusing to suppress the over 1000 grams of cocaine seized from defendant Sokolow.
. The specific elements of the profile vary among airports, in accordance with the experience of DEA agents at a particular location. The original profile, developed in Detroit by DEA Special Agent Paul Markonni consists of seven "primary characteristics” and four "secondary characteristics.” The primary characteristics are: 1) arrival from or departure to an identified source city; 2) carrying little or no luggage; 3) traveling by an unusual itinerary; 4) use of an alias; 5) carrying unusually large amounts of currency; 6) buying airline tickets with a lаrge amount of small denomination currency; and 7) unusual nervousness. The secondary characteristics are: 1) the almost exclusive use of public transportation; 2) immediately making a telephone call after deplaning; 3) leaving a false call back number with the airline; and 4) excessive travel to source or distribution cities. See Cloud, Search and Seizure by the Numbers: The Drug Courier Profile and Judicial Review of Investigative Formulas, 65 B.U.L.Rev. 843, 871 n. 120 (1985) (citing United States v. Elmore,
. The Erwin court relied heavily on defendant's use of an allegedly circuitous route through the airport to distinguish itself from Reid where defendant's arrival in the early morning from a drug source city without checked luggage was held to be too generalized to constitutionally warrant a Terry stop. I dissented in Erwin because I disagreed with the majority that a circuitous route and possibly implausible explanation were sufficient facts to distinguish Erwin from Reid, and because I believed the record did not support the court’s conclusion that the defendant was "nervous'' before he was detained. I take the opposite view here because I believe the agents' suspicions were reasonably backed by specific facts that would not apply wholesale to innocent persons. Erwin,
