Defendant-Appellant Robert Earl Sanders appeals his conviction as a felon in possession of a firearm, alleging that the district court improperly denied his motion to suppress evidence seized following a warrantless stop and frisk of his person. Finding no reversible error, we affirm.

* FACTS AND PROCEEDINGS

While on patrol alone in his squad car one afternoon in December 1990, Waco Police Officer John Hambrick was dispatched by radio to Cruz's Grocery Store to investigate a complaint that a suspicious person with a gun was on the premises of that business. Cruz's Grocery sells alcoholic beverages and is located in a residential area known for its high rate of crimes involving both violence and weapons. Mr. Cruz, the grocery store owner, had called the police on previous occasions about disturbances and suspicious persons at his business.

The dispatcher advised Officer Hambrick that Mr. Cruz had called in the complaint, describing the armed individual as a black male wearing a blue baseball cap and a tan jacket. Officer Hambrick drove to Cruz's in approximately two to three minutes. Upon his arrival he observed eight to ten persons outside of the store. This group comprised at least two children as well as Sanders. He alone matched Mr. Cruz's description of the gunman. Sanders's tan jacket was long enough to cover the waistband of his pants. As Officer Hambrick neared the store, he could see that Sanders was holding a paper bag containing what appeared to be a bottle of some alcoholic beverage. The officer also noticed that Sanders turned and started walking away when he saw the police car approaching.

Officer Hambrick stopped his car, opened the door, drew his duty weapon and, taking cover behind the open door, told Sanders to stop and to get down on the ground. Sanders stopped but refused to lie down. Even though Officer Hambrick repeated his commands, Sanders never got down on the ground.

Within less than a minute Officer Roy Luna arrived to backup Officer Hambrick. Officer Hambrick kept his gun trained on Sanders while he was handcuffed by Officer Luna. Officer Hambrick then reholstered his gun and approached Sanders to frisk him for weapons. Officer Hambrick first removed a lock blade folding hunting knife from Sanders's back right pocket. Officer Hambrick had seen this knife visibly protruding from the top of that pocket when he approached to frisk Sanders, but had not been able to see it earlier.

Sanders asked Officer Hambrick why he (Sanders) was being arrested. Officer Hambrick told Sanders that he was not under arrest; that the police had received a call about Sanders being in Cruz's and carrying a gun, and that he (Officer Hambrick) was frisking Sanders for weapons. Sanders then volunteered that he had a gun, but had not done anything with it. Immediately after Sanders made this statement, Officer Hambrick found a fully loaded Raven .25 caliber pistol in the right pocket of Sanders's tan jacket and removed it. A Raven .25 caliber pistol is a small handgun that can easily be concealed without producing a tell-tale bulge; it is a type of gun commonly encountered by the police.

After completing this frisk and determining that Sanders did not possess any additional firearms or dangerous objects, Officer Hambrick unloaded the Raven pistol and secured it. He then formally placed Sanders under arrest for the state offense of unlawfully carrying a weapon in an establishment licensed for the sale of alcoholic beverages.1

When Sanders was determined to have three prior felony convictions, he was indicted for possession of a firearm by a felon under 18 U.S.C. §§ 922(g)(1) and 924(a). Following the district court's denial of Sanders's pretrial motion to suppress the gun from being introduced into evidence on the ground that it was the fruit of an illegal search and seizure, he was tried before a jury and convicted on the firearms charge. He timely appealed.

Sanders argues that the district court erroneously denied his motion to suppress the gun that had been removed from his pocket. He does not contest the validity of his initial stop and detention, but instead insists that the police exceeded the permissible scope of such a Terry stop by holding him at gunpoint and handcuffing him before frisking him. Sanders claims that this conduct transformed the stop into a de facto arrest that was illegal because it was not supported by probable cause. He concludes that, as the search of his person cannot be upheld as either a proper frisk or incident to a legal arrest, the gun should be suppressed.

The standard of review for a district court's action on a motion to suppress is well established in this circuit. Questions of law are reviewed de novo;2 questions of fact are treated more deferentially:

In reviewing a trial court's ruling on a motion to suppress based on live testimony at a suppression hearing, the trial court's purely factual findings must be accepted unless clearly erroneous, or influenced by an incorrect view of the law, and the evidence must be viewed in the light most favorable to the party prevailing below, except where such a view is either not consistent with the trial court's findings or is clearly erroneous considering the evidence as a whole.3

Twenty-five years after the Supreme Court's opinion in Terry v. Ohio,4 it is now axiomatic that the police are allowed to stop and briefly detain persons for investigative purposes if the police have a reasonable suspicion supported by articulable facts that criminal activity "may be afoot."5 This reasonable suspicion standard is less demanding than the probable cause standard required for an arrest.6

The importance of Terry, however, is not limited to that one holding. Of equal significance, Terry identified the circumstances in which the police are permitted to conduct a carefully limited search of such persons to discover weapons that might be used to assault the officer.7 A police officer may conduct such a limited search if "a reasonably prudent [person] in the circumstances would be warranted in the belief that his safety or that of others was in danger."8 Such a belief must be founded on specific and articulable facts rather than on a mere suspicion or "hunch."9 In so holding, the Court acknowledged:

We are now concerned with more than the governmental interest in investigating crime; in addition, there is the more immediate interest of the police officer in taking steps to assure himself that the person with whom he is dealing is not armed with a weapon that could unexpectedly and fatally be used against him. Certainly, it would be unreasonable to require that police officers take unnecessary risks in the performance of their duties.10

The Terry opinion further instructs how its namesake stop and frisk may be conducted and how they should be reviewed by the courts. If an officer develops--and is able to articulate--reasonable grounds to believe that a suspect is armed and presently dangerous to the officer, third parties, or himself, the officer may "take swift measures to discover the true facts and neutralize the threat of harm if it materialized."11 In determining whether the officer's response to the perceived threat was reasonable, a court must give due weight to the specific reasonable inferences that the officer is entitled to draw from the facts in light of his experience and training.12

1. Intrusiveness of Investigations Generally

Supreme Court opinions subsequent to Terry have amplified these basic lessons. In Florida v. Royer, a plurality of the Supreme Court stated that an investigative detention must be temporary and last no longer than is necessary to effectuate the purpose of the stop, and that the police should use the least intrusive means reasonably available to verify or dispel their suspicions in a short period of time.13 In turn, the proper interpretation of these statements was explained by the Court in United States v. Sharpe.14

In Sharpe, the court stated that, even though the length of an investigatory detention is an important factor in determining whether that detention is proper, "we have emphasized the need to consider the law enforcement purposes to be served by the stop as well as the time reasonably needed to effectuate those purposes."15 Further, although questioning whether the police diligently pursued a means of investigation that was likely to confirm or dispel their suspicions quickly is appropriate, "[a] court making this assessment should take care to consider whether the police are acting in a swiftly developing situation, and in such cases the court should not indulge in unrealistic second-guessing."16

The fact that the protection of the public might, in the abstract, have been accomplished by 'less intrusive' means does not, by itself render the search unreasonable. The question is not simply whether some other alternative was available, but whether the police acted unreasonably in failing to recognize it or to pursue it.17

The use of some force by a police officer does not necessarily cause an encounter to exceed the scope of Terry, which itself involved the use of force. There the officer physically grabbed the suspect and spun him around before frisking him for weapons.18 And, since Terry, the Supreme Court has sustained at least two investigatory detentions involving the display of firearms by the police, albeit that in neither case was the firearm aspect of the detention the primary thrust of the Court's discussion.

The police officer in United States v. Hensley19 stopped a vehicle being operated by a robbery suspect. The officer approached the suspect's car "with his service revolver drawn and pointed into the air."20 The Court held that this conduct was acceptable in connection with an investigatory stop under the relevant conditions:

When the Covington officers stopped Hensley, they were authorized to take such steps as were reasonably necessary to protect their personal safety and to maintain the status quo during the course of the stop. The Covington officers' conduct was well within the permissible range in the context of suspects who are reported to be armed and dangerous.21

In United States v. Sharpe, the Supreme Court examined the propriety of an investigatory stop that began with a police officer stopping a pickup truck. "After stopping the truck, [the officer] had approached it with his revolver drawn, ordered the driver, Savage, to get out and assume a 'spread eagled' position against the side of the truck, and patted him down."22 The Court found that this was a proper investigatory stop and not a de facto arrest.23

We too have previously addressed the propriety of an officer drawing his weapon or pointing it at a suspect in the course of an investigatory detention. In United States v. Maslanka,24 we approbated an officer's approach with service revolver drawn and pointed at the occupants of a vehicle that had just led the police on a high speed chase.25 In so doing, we stated: "To require an officer to risk his life in order to make an investigatory stop would run contrary to the intent of Terry v. Ohio."26 We likewise found in United States v. Worthington27 that the police acted properly when they blocked the path of an airplane that was suspected of involvement in drug smuggling and approached it with drawn weapons.28 And, in United States v. Holloway,29 we held that the police made a proper investigatory stop rather than a de facto arrest when they blocked the path of a suspect's car with their own car and exited it with guns drawn.30

Other circuits have held uniformly that, in and of itself, the mere act of drawing or pointing a weapon during an investigatory detention does not cause it to exceed the permissible bounds of a Terry stop or to become a de facto arrest.31 Other circuits also have held that a suspect may be detained in a police car32 or ordered to lie prone on the ground33 in the course of an otherwise proper investigatory stop.

Of particular importance to the instant case, several circuits have specifically considered whether the police may handcuff a suspect during an investigatory detention without exceeding the limits of such a detention. Recognizing that handcuffing substantially aggravates the intrusiveness and is not typically part of a Terry stop, the Ninth Circuit nevertheless held that handcuffing may be proper during an investigatory detention.34 "The purpose of the Terry frisk is 'to allow the officer to pursue his investigation without fear of violence.' "35

In United States v. Bautista, the Ninth Circuit examined the act of a police officer who handcuffed two robbery suspects after they had been frisked for weapons and found to possess none. The officer testified that he handcuffed the suspects because he had to leave his partner alone with both suspects while he (the testifying officer) investigated the suspects' alibi. The level of concern for his partner's safety was elevated by the presence of needle marks on both suspects' arms and the suspects' behavior, which suggested that they might attempt to flee. In approving of that handcuffing, the court held that even a total restriction of liberty, if brief and not excessive under the circumstances, is permissible during a Terry stop and does not automatically convert it into an arrest.36

In United States v. Taylor, the Ninth Circuit found it permissible for officers to order two suspects out of their car at gunpoint, order one of the suspects to lie on the ground, then handcuff that suspect before frisking him.37 Requiring the suspect to lie down and handcuffing that suspect were found particularly appropriate because he had twice disobeyed orders to raise his hands and had made furtive movements inside the vehicle where his hands could not be seen.38 Likewise, in United States v. Thompson, the Ninth Circuit found that the suspect's repeated attempts to reach into a coat pocket justified the police's handcuffing him before frisking him.39

In United States v. Glenna,40 the Seventh Circuit recognized that handcuffing can be consistent with an investigative stop even though it is normally associated with an arrest.41 There, an officer stopped a suspicious person who, when asked for identification, reached into his pants pocket, which contained a large lump that the officer thought might be a weapon. The officer grabbed the suspect's hand and pulled a loaded handgun magazine from the pocket. The officer then handcuffed the suspect before continuing his frisk. The court announced that it would not substitute its judgment for that of the handcuffing officer when common sense and ordinary human experience convince the court that the officer reasonably believed that handcuffing was required to conduct the investigative stop safely.42

The D.C. Circuit upheld the handcuffing of a suspect during a Terry stop in United States v. Purry43 despite the apparent lack of any concern about officer safety. There, a police officer stopped a person he suspected of fleeing after committing a bank robbery minutes earlier. After being informed that he would have to go to the bank for a showup, the suspect attempted to flee. The officer responded by handcuffing the suspect and transporting him back to the bank in a police car. The court held that when the suspect attempted to frustrate further inquiry, the officer "was not required to shrug his shoulders and allow the suspected criminal to walk away. The handcuffing was an appropriate method of maintaining the status quo while further inquiry was made."44

Clearly, using some force on a suspect, pointing a weapon at a suspect, ordering a suspect to lie on the ground, and handcuffing a suspect--whether singly or in combination--do not automatically convert an investigatory detention into an arrest requiring probable cause. We caution, however, that this holding is not to be interpreted as meaning that the police are automatically authorized to employ these procedures in every investigatory detention. The relevant inquiry is always one of reasonableness under the circumstances. The court must determine case by case whether the police were unreasonable in failing to use less intrusive procedures to conduct their investigation safely.45

When Officer Hambrick arrived on the scene at Cruz's Grocery, he had only a matter of seconds to assess the situation, formulate a plan of action, and implement it. In so doing, he had to balance several competing priorities: to investigate the alleged crime and make any appropriate arrests; to prevent the commission of any additional crimes; not to infringe on the rights of Sanders or any other persons who might be affected by the officer's actions or inactions; to ensure the safety of others of the general population present or nearby; and to go home in one piece at the end of his shift.

As he pulled up, Officer Hambrick observed that Sanders was wearing, among other articles of clothing, a blue baseball cap and a tan jacket which concealed the area around his waistband. A firearm or other weapon could have been concealed under such a jacket with ease. Sanders was also holding a paper bag that appeared to contain an alcoholic beverage. If Sanders had been consuming alcoholic beverages, his judgment might have been impaired and his inhibitions reduced. In such a state Sanders might be more likely to resort to violence as a perceived solution to the apparent impending police investigation. Additionally, the paper bag could contain a weapon or some object that might be used as an improvised weapon (the ubiquitous broken bottle). Once Sanders saw the squad car pulling up, he turned and started to walk away. Such an action can be used by a criminal to prepare for a violent confrontation by surreptitiously retrieving a concealed weapon then spinning back around to face the officer and use the weapon against him. Officer Hambrick also observed several other people standing nearby, including children. Further, he knew that the surrounding housing area was filled with additional innocent persons and that, as it was around five o'clock in the afternoon, those persons were likely to be out and about.

We conclude that, under these circumstances, Officer Hambrick did not act unreasonably in immediately drawing his weapon when he confronted Sanders. He had reason to believe Sanders was armed, given Cruz's report. Officer Hambrick had to act swiftly to preserve the status quo and avoid exacerbating the situation. Any less forceful or decisive action on his part could have invited a suspect with aggressive tendencies to escalate that situation, thereby increasing the danger to himself, to the other officer, and to the bystanders.

Neither was it unreasonable for Officer Hambrick to order Sanders to lie face down on the ground. Admittedly, being held at gunpoint by the police would be a powerful incentive for most persons neither to flee nor resist an impending investigation. Such a situation is not, however, equally inspirational for everyone. For example, some individuals will be sufficiently familiar with police procedure to believe that the officer will not actually use deadly force against them--particularly in close proximity to innocent bystanders--or will be so hesitant to use force that the suspect can "get the jump" on the officer. Others, such as persons who have been consuming drugs or alcoholic beverages, might misjudge the gravity of the situation or might be confused as to which responses on their part are or are not appropriate. A third group will be so desperate to avoid apprehension that they will knowingly take any risk to evade the police.

If a suspect complies with a police order to lie face down on the ground, his ability to fight or flee is significantly reduced, thereby helping to preserve the status quo. Thus ordering a person whom the police reasonably believe to be armed to lie down may well be within the scope of an investigative detention. We conclude that it clearly was within those limits in the instant case.

But Sanders did not comply with the order to lie down. Instead, he remained standing, facing Officer Hambrick. This placed the police in the unenviable position of deciding whether to continue the investigation, relying solely on the gunpoint intimidation of Sanders to ensure that he would neither flee nor resist the investigation, or to restrain Sanders in some other way so as to reduce further the chances of flight or resistance. We conclude that the officers were not unreasonable in choosing the latter course and handcuffing Sanders.

Sanders nevertheless insists that the police should have continued detaining him at gunpoint, without resorting to handcuffs, while they brought Mr. Cruz out of his store to confirm or deny that Sanders was indeed the man with the gun whom Cruz had reported. Sanders further contends that he should not have been searched until and unless Mr. Cruz positively identified him as the gunman. He reiterates the truism that the police should employ the least intrusive means available to investigate the complaint. And finally he argues that even if reasons for conducting a Terry frisk had existed earlier, those reasons evaporated once he was handcuffed because he was no longer dangerous to the officers or the bystanders. We find Sanders's arguments unpersuasive.

As previously discussed, a gunpoint detention does not ensure that a suspect will not fight or flee. Neither is such a detention without risk to either the police officers,46 the person detained,47 or to innocent bystanders.48 And, even though handcuffing a suspect long enough to frisk him more safely is intrusive, so is a prolonged, less safe detention at gunpoint. We are aware of no formula by which to test the relative intrusiveness of different but similar police procedures with such a degree of precision. Even if we were to assume arguendo that a prolonged gunpoint detention were less intrusive than handcuffing, the issue is not whether "the protection of the public might, in the abstract, have been accomplished by 'less intrusive' means."49 The question instead is whether "the police acted unreasonably in failing to recognize [a less intrusive method] or to pursue it."50 Given the risks of the method actually used by the police vis-a-vis the risks attendant on the method proposed by Sanders on appeal, we cannot say that the police acted unreasonably in handcuffing Sanders.

At oral argument, Sanders questioned whether the police were justified in frisking him after he was handcuffed. As noted, he claims that once he was handcuffed he could not flee or present any risk to the police officers or others, so that the Terry justification for frisking him disappeared ipso facto. Sanders again fails to convince us of the merit of his position.

We initially note that Sanders's argument parallels one made by the defendant in New York v. Belton,51 in which the Supreme Court upheld the validity of a search of a vehicle as incident to arrest of the occupants of that car. Just as a Terry frisk is justified principally on the basis of officer safety, a search incident to arrest is justified on the similar basis of preventing an arrestee from reaching a weapon, as well as preventing the destruction of evidence.52 In Belton, the Court found that the search of the defendant's car was a contemporaneous incident of arrest even though the suspects had been removed from the car and placed under arrest before the officers searched the car.53 As the suspects had been in the vehicle immediately prior to their arrest, the contents of the passenger compartment were still " 'within the arrestee's immediate control' within the meaning of the Chimel case."54

If anything, the timing and sequence of the stop-aim-handcuff-search-arrest in the instant case militate against Sanders's argument more strongly than do the sequence and timing of the stop-arrest-search in Belton. As the suspects in Belton had been placed under arrest formally prior to the search of their vehicle, it was quite unlikely that thereafter they would have been released at the scene regardless of the outcome of any additional investigation at the scene. Consequently, it was also unlikely that they would have been able to retrieve either weapons or contraband from their vehicle after their arrest. In contrast, Sanders had only been detained up to the time that the gun was discovered in his pocket. Officer Hambrick testified unequivocally that if he had not developed probable cause to arrest Sanders,55 he would have removed the handcuffs and released him. Sanders then could have easily retrieved his pistol.

But assume arguendo that we were now to hold that, under Terry, the police could not frisk Sanders once he was handcuffed; and further assume that the police could not otherwise develop probable cause sufficient to arrest him, but that they still had reasonable grounds to believe that he was armed and presently dangerous. What options would remain? Clearly, the police could not keep Sanders handcuffed indefinitely. But neither could they simply remove the handcuffs, for then the danger would return in full force, again justifying the officers' frisking of Sanders for weapons.56 Continuing, if the officers were again to handcuff Sanders so that they could accomplish the frisk in relative safety, the danger justifying a frisk would again cease to exist. The circular nature of this premise is obvious and need not be pursued ad absurdum, for it is built on a flawed foundation.

Sander's argument is entirely dependent on the assumption that, by handcuffing a suspect, the police instantly and completely eliminate all risks that the suspect will flee or do them harm. As is sadly borne out in the statistics for police officers killed and assaulted in the line of duty each year, however, this assumption has no basis in fact.

Handcuffs are a temporary restraining device; they limit but do not eliminate a person's ability to perform various acts.57 They obviously do not impair a person's ability to use his legs and feet, whether to walk, run, or kick. Handcuffs do limit a person's ability to use his hands and arms, but the degree of the effectiveness of handcuffs in this role depends on a variety of factors, including the handcuffed person's size, strength, bone and joint structure, flexibility, and tolerance of pain. Albeit difficult, it is by no means impossible for a handcuffed person to obtain and use a weapon concealed on his person or within lunge reach, and in so doing to cause injury to his intended victim, to a bystander, or even to himself. Finally, like any mechanical device, handcuffs can and do fail on occasion.

The limitations of handcuffs' effectiveness are widely known to law enforcement personnel,58 and Officer Hambrick was likely aware of them.59 Despite this widespread knowledge, in 1991 alone (the most recent year for which statistics are available), at least four police officers were killed by persons who had already been handcuffed.60 Our belabored point is that, in and of itself, handcuffing Sanders would not, ipso facto, eliminate entirely the risk of harm and thereby prevent a reasonable officer from concluding that Sanders should still be frisked for weapons to ensure that he was not armed and presently dangerous. We are convinced that for the officers to frisk Sanders after he had already been handcuffed was not unreasonable under the circumstances; we are satisfied instead that their actions were well within the permissible bounds of an investigatory detention under Terry and its progeny.

The Supreme Court has instructed that "the policeman making a reasonable investigatory stop should not be denied the opportunity to protect himself from attack by a hostile suspect."61 Officer Hambrick's investigatory detention of Sanders was reasonable, as was his concern for his own safety and the safety of others. Giving due weight to the specific, reasonable inferences that the officers were entitled to draw from the facts in light of their experience and training,62 and declining to "indulge in unrealistic second-guessing,"63 we cannot say that the police acted unreasonably in their choice of methods used to conduct the subject investigation or in the manner and extent of their application of those methods. As we find that the officers' conduct did not exceed the proper scope of an investigatory detention, the judgment of the district court is

AFFIRMED.