An Offense-Severity Model for Stop-and-Frisks
abstract. This Note joins a growing chorus of scholarship criticizing the lack of proportionality analysis in the Supreme Court’s Fourth Amendment jurisprudence. Rather than simply bemoan the current state of legal doctrine, we offer a practical test that state and federal courts could use to determine the permissible scope of pedestrian stop-and-frisks. Specifically, we propose that courts adopt an offense-severity model that distinguishes minor offenses (like jaywalking, public alcohol consumption, and simple trespass) from more serious misdemeanors and felonies. Two state supreme courts—Massachusetts’s and Washington’s—have already adopted part of our approach, distinguishing noncriminal from criminal activity for the purposes of stop-and-frisks. That is, police in those states may not engage in stop-and-frisks based on mere suspicion of noncriminal offenses. Our Note further advocates for a rebuttable presumption against stop-and-frisks for petty misdemeanors. To overcome this presumption, prosecutors would bear the burden of demonstrating that an officer reasonably believed the suspected offense posed an immediate threat to public safety. In advocating such a model, our Note contributes to a broader debate about crime-severity’s usefulness as a rubric for assessing police conduct under the Fourth Amendment and its state law equivalents.
authors. David Keenan: Law Clerk, Hon. Paul G. Gardephe, Southern District of New York; J.D., Yale Law School, 2013. Tina M. Thomas: J.D. expected, Yale Law School, 2014. The idea for this Note arose from our experience representing indigent clients in the New Haven Superior Court. We would like to extend our gratitude to the director of Yale’s Criminal Justice Clinic, Professor Fiona Doherty; New Haven public defenders Thomas Ullmann, Jennifer Mellon, and Trey Bruce; and our student colleagues, particularly Alyssa Work, Kory DeClark, and Daniel Friedman. We also wish to thank Professor Jeffrey A. Fagan for his helpful feedback on an earlier draft, Professor James Forman, Jr., for his valuable comments at a workshop on our Note, and Renagh O’Leary and the Yale Law Journal staff for their insightful suggestions throughout the editorial process. Finally, we are especially indebted to our clients, particularly John Mitchell, whose experiences inspired us to pursue this topic.
In the mid-1990s, the New York City Police Department (NYPD) embraced a new strategy for crime suppression predicated on James Q. Wilson and George Kelling’s “broken windows” criminological theory.1 The basic thrust of that strategy has now been adopted in some form by hundreds of police departments across the country.2 Known as “order-maintenance policing,” the strategy calls for a zero tolerance policy towards so-called “quality of life” offenses whose occurrence is thought to reflect crime-generating social disorder.3 In departments that follow an order-maintenance approach, officers aggressively enforce city ordinances against activities like panhandling, public drunkenness, graffiti, prostitution, and loitering.4 The explicit aim of order-maintenance policing is to “reclaim” the streets in order to “undercut the ground on which more serious crimes seem possible and even permissible.”5 Practically speaking, it often means using aggressive enforcement of quality of life violations as a pretext to seize weapons or other contraband.6
One of the primary legal mechanisms for effectuating order-maintenance policing is the stop-and-frisk.7 A stop-and-frisk is a nonconsensual encounter between police and citizen that falls short of a full-blown arrest. The Supreme Court first recognized the procedure’s constitutional legitimacy in Terry v. Ohio.8 To make a Terry stop, an officer need only have reasonable grounds for believing that “criminal activity may be afoot.”9 A limited search of the suspect’s person (the “frisk”) is similarly permissible so long as the officer reasonably believes the suspect is armed and dangerous.10 Subsequent case law has clarified that pretextual motivations for executing a stop-and-frisk are irrelevant.11 Courts are directed to apply an objective standard in reviewing such encounters.12
Challenges to stop-and-frisk policies in recent years have proven successful. In 2011, Philadelphia chose to accept judicial monitoring of stops rather than contest an ACLU lawsuit.13 The following year, Seattle did the same in response to a Department of Justice investigation.14 And, most significantly, in August 2013, a federal district court judge granted a preliminary injunction against the NYPD’s stop-and-frisk program.15 In her controversial Floyd v. City of New York ruling, Judge Shira A. Scheindlin found the city had been deliberately indifferent to an unconstitutional policing policy that (1) permitted stop-and-frisks to be made on less than reasonable suspicion and (2) utilized racial classifications to determine whom to stop-and-frisk.16
The time is ripe, then, to reconsider the purported legal justification underlying aggressive stop-and-frisk practices. This Note asks one narrow, but exceedingly important, question about the stop-and-frisk: should officers be able to stop individuals on the basis of any suspected offense, no matter how minor? As the leading treatise on Fourth Amendment law notes, this question “has seldom been confronted head on by the lower courts.”17 Ordinarily, courts limit their inquiries to whether officers have reasonable suspicion that anoffense is being, has been, or is about to be committed, regardless of its severity.18 We argue that such an approach is both unfaithful to Terry’s reasoning and misguided as a matter of policy.
In so arguing, we join a growing chorus of academic voices criticizing the lack of proportionality in the Supreme Court’s Fourth Amendment jurisprudence.19 As these scholars have persuasively shown, the Court’s “transsubstantive” approach to search and seizure law ill suits the Fourth Amendment’s reasonableness requirement.20 After all, how can courts strike a proper “balance between the public interest and the individual’s right to personal security”21 without taking into account the seriousness of the offense the government seeks to investigate?
Where the scholarly literature has fallen short, however, is in proposing an offense-severity test that is workable both on the streets and in the courtroom. As Professor Eugene Volokh has noted: “[T]he devil is in the details. If courts can’t make the severity distinctions work in practice, then the distinctions’ merits in principle are of little consequence.”22 Indeed, in other Fourth Amendment contexts, the Supreme Court has cited administrability concerns as reason to avoid adopting an offense-severity model.23
Our Note seeks to remedy this shortcoming by proposing a model that uses preexisting legislative classifications to define offense-severity for Terry purposes. The model has two basic components. First, it distinguishes between civil infractions or violations, on the one hand, and criminal misdemeanors and felonies, on the other. A civil infraction or violation is a regulatory offense that is ordinarily punishable by fine only. Because Terry spoke in terms of proportionate government responses to suspected criminal wrongdoing, courts should clarify that suspicion of a civil offense does not justify the intrusiveness of a stop-and-frisk.
Second, our model deems Terry stops based on suspicion of petty offenses presumptively invalid. For constitutional purposes, petty offenses are criminal misdemeanors that carry a maximum possible sentence of six months in jail.24 Applying the petty offense distinction in the Terry context makes sense because the government’s law enforcement interest is least compelling, and the potential for harassment is greatest, when stop-and-frisks are premised on minor suspected crimes. Adopting a rebuttable presumption for petty offenses also helps mitigate the weightiest objections to our reliance on offense categorizations—namely, that those categorizations vary across jurisdictions and are susceptible to easy legislative manipulation.25
The Note proceeds in five Parts. In Part I, we familiarize readers with Terry and current stop-and-frisk doctrine and practice. As we show, courts have largely avoided asking whether the offense used to justify a pedestrian Terry stop should matter in determining its legality. In Part II, we describe offense-severity in greater detail before setting forth our proposed model in Part III. Part IV offers doctrinal and normative justifications for applying our model to pedestrian stops. Finally, in Part V, we consider several state and federal court cases that incorporate offense-severity into their Terry analyses. These cases suggest the feasibility and utility of distinguishing among suspected offenses when assessing the reasonableness of stop-and-frisks.26
I. stop-and-frisk doctrine and practice under terry
A. Stop-and-Frisk Doctrine
Ordinarily, a lawful warrantless search or seizure requires that officers have probable cause to believe that an offense has been, is being, or will be committed.27 Terry famously departed from this standard by recognizing the constitutionality of stop-and-frisks where officers possess merely a reasonable and particularized suspicion of criminal wrongdoing.28
Given the Court’s “agonized opinion,”29 there was ample reason to believe Terry would be limited to its operative facts, namely those situations where officers possess reasonable suspicion of (a) an ongoing or prospective offense30 of (b) a criminal nature31 that (c) threatens violence to persons or property.32 Indeed, in 1975, the American Law Institute adopted a similar standard in its Model Code of Pre-Arraignment Procedure.33 Courts have nonetheless gradually expanded the boundaries of permissible Terry stops to include stops for suspected past offenses,34 nonviolent drug crimes,35 and civil infractions.36 It is the last category that marks the least defensible expansion of Terry and the one most at odds with traditional notions of Fourth Amendment reasonableness. This Part offers a brief review of Terry in order to demonstrate that incorporating offense-severity considerations into their review of stop-and-frisks would enable courts to remain faithful to the Terry decision.
The facts of Terry are familiar. Officer Martin McFadden spotted two men pacing up and down a street, each pausing several times to look in a shop window.37 Suspicious the men were “casing a job,” McFadden followed them a short distance where the men met up with a third man.38 At that point, McFadden—who was alone—initiated a stop and began to pat down the outer clothing of one of the men, John Terry.39 This “frisk” revealed a .38-caliber revolver in the breast pocket of Terry’s overcoat.40
In upholding Terry’s conviction, the Court recognized the impracticality of subjecting “swift action predicated upon the on-the-spot observations of the officer on the beat” to pre-enforcement review under the Warrant Clause.41 It chose instead to analyze stop-and-frisks according to the Fourth Amendment’s general reasonableness requirement. This called for the adoption of a proportionality test balancing the individual’s liberty interest against the government’s generalized goal of “effective crime prevention and detection.”42 While acknowledging the potential for abuse,43 the Court ultimately recognized a “narrowly drawn authority” for warrantless stops based on an officer’s reasonable suspicion.44
Though subsequent cases have added flesh to Terry’s skeletal framework, doctrinal uncertainty remains over whether offense-severity distinctions matter in the Terry context. On the one hand, the Court has dispensed with a prime indicator of offense-severity—an offense’s potential to cause violence—by applying Terry to possessory drug offenses45 and completed felonies.46 On the other hand, Terry emphasizes the need to balance personal liberty interests against governmental objectives.47 And the Court has explicitly left open “whether Terry stops to investigate all past crimes, however serious, are permitted.”48 Furthermore, the Court has indicated in the automobile context, albeit obliquely, that a stop for a civil traffic infraction requires probable cause.49
In other Fourth Amendment contexts, the Court has generally declined to incorporate crime-severity in its reasonableness calculus. In Atwater v. City of Lago Vista, for instance, it made no difference that the seizure involved an arrest for a nonjailable petty misdemeanor.50 Likewise, in New Jersey v. T.L.O., the Court upheld as reasonable a school administrator’s search of a student’s purse based on nothing more than the administrator’s suspicion that the student was violating a school rule against smoking.51 And more recently, the Court held constitutionally permissible an invasive strip search of an individual detained for a minor offense involving the nonpayment of a fine.52
But at times, the Court has viewed offense-severity as highly relevant to its determination of reasonableness. In Welsh v. Wisconsin, for instance, the Court noted that “an important factor to be considered when determining whether any exigency exists” that would justify a warrantless home arrest “is the gravity of the underlying offense for which the arrest is being made.”53 Welsh involved a warrantless entry where officers had probable cause to believe the home’s occupant had recently engaged in drunk driving.54 The Court specifically rejected the state’s exigent circumstances argument, namely that its law enforcement needs necessitated immediate entry to the home to prevent the spoliation of blood alcohol evidence. Under Wisconsin law, the Court noted, driving under the influence was a “noncriminal, civil forfeiture offense for which no imprisonment [was] possible.”55 The Court explained that “[t]his is the best indication of the State’s interest in precipitating an arrest, and is one that can be easily identified both by the courts and by officers faced with a decision to arrest.”56 Ultimately, the Court concluded that Wisconsin’s minimal law enforcement interest did not trump the petitioner’s reasonable expectation of privacy in his own home.
In a development that underscores how perceptions of crime-severity can shift over time, this Term the Court heard argument in a case where California urged precisely the opposite outcome.57 The case involves a motor vehicle stop for suspicion of drunk driving based solely on an anonymous tip.58 California contends that because drunk driving poses such a serious threat to public safety, ordinary Fourth Amendment requirements for corroborating anonymous tips should be relaxed.59
There is still room, then, for arguing that crime-severity ought to matter for Fourth Amendment purposes, despite the clear trend away from such considerations in the Court’s recent jurisprudence. Our proposed model therefore has the potential to influence how courts, legislators, and litigants conceive of Terry’s boundaries going forward.
B. Stop-and-Frisk Practice
As currently practiced, stop-and-frisk knows few legal boundaries. So long as some law makes the suspected conduct illegal, most courts have deemed the stop and resulting frisk valid. Indeed, lower courts routinely uphold stop-and-frisks for even the most minor offenses so long as an officer can articulate a reasonable suspicion.60 Only rarely has a court paused to ask the obvious question—reasonable suspicion of what?61 Therefore, the law today “places only the lightest of limits on whether a suspect can be seized, and nearly no limits at all on how.”62
Take New York, for example. A 1999 report by New York’s Office of the Attorney General (OAG) found that more than ten percent of NYPD Terry stops were for “low-level” quality of life and misdemeanor offenses.63 As the OAG’s report noted, “because low-level ‘quality of life’ and misdemeanor offenses are more likely to be committed in the open, . . . the ‘reasonable suspicion’ standard may be more readily satisfied as to those sorts of crimes.”64
Several of the offenses cited in the OAG’s report are violation-level offenses.65 Under New York’s Penal Code, violations are considered civil rather than criminal.66 Because New York law explicitly limits an officer’s stop-and-frisk authority to instances where he or she “reasonably suspects [a] person is committing, has committed or is about to commit either (a) a felony or (b) a misdemeanor,” violations cannot serve as a lawful basis for a stop-and-frisk.67 Hence, it appears that NYPD officers regularly violated New York law during the 1990s by making such stops.
More recent circumstantial data suggest stops on suspicion of violation-level behavior remain commonplace. For every stop, NYPD officers must complete a form—the Unified Form 250 (UF-250)—that identifies the suspected crime.68 The form instructs officers to indicate the particular suspected misdemeanor or felony that serves as the basis for each stop. Between 2004 and 2009, the number of UF-250s that failed to state a suspected offense rose from 1% to 36%.69 Over the same period, the total number of stops increased 83%, from 314,000 to 576,000.70 Slightly more than 6% of those stops resulted only in the issuance of a summons.71 Although it is conceivable that in each and every one of those stops officers legitimately suspected criminal activity but found evidence of violations only, that hypothesis seems dubious, particularly in light of other evidence.
Consider, for example, an audio recording played during the Floyd trial of a conversation between a patrolman and his commanding officer. “We go out there and we summons people,” Deputy Inspector Christopher McCormack told Officer Pedro Serrano.72 McCormack further instructed Serrano that the way to prevent violent crime was to stop, question, and frisk “the right people at the right time, the right location.”73 When asked who the “right” people were, Inspector McCormack, in reference to a report of earlier criminal activity, responded that “[t]he problem was, what, male blacks . . . . I told you at roll call, and I have no problem telling you this, male blacks 14 to 20, 21.”74
In her opinion declaring the NYPD’s stop-and-frisk policy unconstitutional, Judge Scheindlin provided further examples of officers emphasizing the importance of issuing summonses for violation-level offenses. She noted that “[b]etween 2002 and 2011, the number of stops increased from roughly 97,000 to roughly 686,000 per year.”75 How was it possible, the judge wondered, that the NYPD was able to increase stops by “roughly 700%, despite the fact that crime continued to fall during this period?”76 In answering her own question, she noted the following statements of NYPD supervisors:
· “If they’re on a corner, make them move. They don’t want to move, you lock them up. Done deal. You can always articulate later.”77
· “Shake everybody up. Anybody moving, anybody coming out that building, 250, verticals, and give me a couple of community visits. C-summons as well.”78
· “[G]o crazy . . . . If we get every single summons in St. Mary’s [Park], I don’t care.”79
That officers appear to stop people on suspicion of violations is perhaps unsurprising when one considers the contradictory testimony that NYPD supervisors have given concerning the legality of such conduct. In 1998, one NYPD commissioner accurately testified before the New York City Council that “violation-level offenses cannot lawfully support a forcible ‘stop.’”80 But in the more recent Floyd litigation, an NYPD Inspector responded to a similar question by stating that it was his belief that officers could lawfully stop, question, and frisk based on their suspicion of “any violation of law,” not just misdemeanors or felonies.81
Such attitudes are not NYPD-specific. For example, in training their officers on proper stop-and-frisk tactics, Philadelphia police supervisors similarly “encourage officers to be clever and resourceful about using even minor infractions—something as routine as spitting, littering, loitering, or holding an open container of alcohol—as a rationale to stop a suspect person and conduct a legal frisk.”82
Because post hoc rationalizations matter a great deal in determining the legality of Terry stops, order-maintenance policing is premised on the idea that officers need only point to facts reasonably suggestive of some legal wrongdoing to satisfy the constitutional standard, and it is clear that low-level offense conduct provides the articulated justification for many stop-and-frisks. What has also become clear is that such justifications are just that—articulated. As Inspector McCormack’s comments lay bare, using low-level offenses to justify Terry stops tends to mask more invidious reasons for stop-and-frisks, such as racial profiling.83
C. The Path Not Taken
As Professor Sherry Colb reminds us, things need not have turned out this way.84 Justice Harlan penned a concurrence in Terry that, had it been followed by lower courts with greater regularity, might have averted many present-day problems.85 Instead of scrutinizing the reasonableness of the frisk, Harlan emphasized the need for lower courts to engage in a searching inquiry of the propriety of the initial stop. He explained that “if the frisk is justified in order to protect the officer during an encounter with a citizen, the officer must first have constitutional grounds to insist on an encounter, to make a forcible stop.”86 Of course, “[a]ny person, including a policeman, is at liberty to avoid a person he considers dangerous.”87 Justice Harlan would have therefore made clear that “the right to frisk . . . depends upon the reasonableness of a forcible stop to investigate a suspected crime.”88
Here, then, is an implicit justification for according offense-severity significant weight in Fourth Amendment balancing. Put slightly differently, the police should not be permitted to use suspicion of minor offenses to engage infishing expeditions aimed at ferreting out the armed and potentially dangerous. As Judge Friendly explained in unsuccessfully resisting Terry’sextension to suspected narcotics possession:
Terry v. Ohio was intended to free a police officer from the rigidity of a rule that would prevent his doing anything to a man reasonably suspected of being about to commit or having just committed a crime of violence, no matter how grave the problem or impelling the need for swift action, unless the officer had what a court would later determine to be probable cause for arrest. It was meant for the serious cases of imminent danger or of harm recently perpetrated to persons or property, not the conventional ones of possessory offenses.89
On the other hand, where the police have a reasonable basis for suspecting a serious crime is about to be committed, the law ought to permit them to act aggressively with a view towards officer safety. In Justice Harlan’s words, “[t]here is no reason why an officer, rightfully but forcibly confronting a person suspected of a serious crime, should have to ask one question and take the risk that the answer might be a bullet.”90
II. what’s in a label?: understanding offense-severity distinctions
Justice Harlan’s concurrence raises the obvious question: what is a “serious” offense? One possible way of approaching the question is to begin by defining what a serious offense is not. Accordingly, our task in this Part and Part III is, first, to describe how legislatures distinguish among offenses, and second, to explain why those distinctions are worthy of deference for Fourth Amendment purposes.
A. Offense Distinctions
All states distinguish between serious and minor crimes in some way.91 The most common and recognizable distinction is between misdemeanors and felonies. Historically at common law, felonies were those crimes for which death and forfeiture were the prescribed punishments.92 All other crimes were considered misdemeanors. Today, jurisdictions distinguish between felonies and misdemeanors primarily by the length of the authorized punishment and the place of incarceration. Generally speaking, felonies are those offenses that are punishable by more than one year in the state penitentiary.93 Misdemeanors, by contrast, encompass less serious crimes that are punishable by fines only or jail sentences of less than one year.94
Federal law and many state penal codes further distinguish some misdemeanors as “petty.”95 While statutory codes differ in precisely where they draw the line between petty and serious offenses, for constitutional purposes the Supreme Court in Baldwin v. New York defined a “serious” crime as one where the authorized term of imprisonment exceeds six months.96 Prior to its Baldwin holding, the Court had struggled in its efforts to distinguish petty from serious crimes, variously looking to factors like the authorized term of punishment, the crime’s character as malum in se or malum prohibitum, and whether the offense was indictable at common law.97 In Baldwin, the Court rejected prior approaches and simply created a constitutional floor where the crime charged authorized a maximum term of imprisonment of more than six months.98 This followed from the Court’s recognition that every jurisdiction save New York City afforded defendants the right to a jury trial in such instances.99 While the Court has since declined to hold that an offense carrying a maximum term of six months or less “automatically qualifies” as petty, it has “presume[d] for purposes of the Sixth Amendment that society views such an offense as ‘petty.’”100
Many jurisdictions have adopted another category of offenses, commonly labeled violations or infractions.101 These are considered civil, rather than criminal, offenses.102 Significantly, civil offenses usually lack the social stigma and collateral consequences associated with misdemeanor and felony offenses.103 While in some states violations may be theoretically punishable by short jail terms,104 many states punish violations with fines only.105 In Connecticut, for example, littering, vandalism, simple trespass, and possessing an open container of alcohol in public are all fine-only violations.106 Connecticut has also joined several states in decriminalizing marijuana possession.107
Federal and state legislatures are not the only actors that define offense-severity. Municipalities also exercise considerable regulatory control by virtue of either their inherent police powers or special legislative grants of authority.108 Traditionally, courts treated municipal ordinances as creating quasi-criminal offenses, describing them as “public torts,” “public welfare,” “police,” or “regulatory” offenses.109 The modern trend has been to characterize an offense as civil or criminal based on the penalty the ordinance authorizes or whether the ordinance has a counterpart in the state criminal code.110 Indeed, some states have explicitly empowered municipalities to choose whether municipal offenses should be classified as misdemeanors or infractions.111
B. Legislative Choices
This Section considers the factors that motivate legislatures—including municipalities—in choosing among offense classifications. Classification choices affect more than just the punishment an offender receives. They also help determine the symbolic meaning that society attaches to the offense conduct, the procedural rights to which a suspect or defendant is entitled, and the collateral consequences that result from a finding of guilt.
Criminal laws, like all laws, reflect communal judgments about social norms and social utility.112 Those judgments are historically contingent and are shaped by prevailing cultural attitudes and technology.113 Laws prohibiting sodomy and alcohol consumption are representative of the former; the enactment of Internet-crimes legislation typifies the latter.
They also serve expressive and instrumental purposes.114 At a commonsense level, the maximum authorized punishment for an offense reflects its severity.115 But categorical labels add a further layer of meaning. Civil offenses carry less opprobrium than do criminal ones, and petty misdemeanors are likewise perceived as less serious than other misdemeanors and felonies. Civil sanctions are generally nonpunitive. Misdemeanors and felonies, on the other hand, have traditionally signaled society’s judgment that the offender has violated not only its social order, but its moral norms as well.116
Offense classifications also serve an instrumental purpose by defining the scope of procedural rights afforded to suspects or defendants. Distinctions between criminal and civil offenses, or between completed felonies and completed misdemeanors, can shape the police’s investigatory authority.117 Similarly, many jurisdictions continue to adhere to the common law division between warrantless arrests for felonies and misdemeanors.118 The former require only probable cause whereas the latter require that the offense actually be committed in the officer’s presence. And a host of post-arrest procedural rights are grounded on the nature of the offense charged, including rights to counsel,119 grand jury indictment,120 and jury trial.121
Because the rights of defendants are contingent upon offense classifications, legislative choices reflect decisions about resource allocation. For instance, legislatures have reaped significant cost savings by reclassifying certain misdemeanors as violations.122 Similarly, the six-month demarcation for petty offenses exercises a strong pull in favor of petty offense classifications because, by designating offenses as petty, legislatures can avoid the costs associated with jury trials and, potentially, court-appointed counsel.123
Finally, while it is true that Fourth Amendment law generally treats all offenses alike, courts have occasionally accorded crime-severity significant weight in determining the reasonableness of police and prosecutorial actions.124 In the next Part, we aim to encourage such efforts by proposing a model for stop-and-frisks that incorporates offense-severity.
III. an offense-severity model for stop-and-frisks
It is easy to forget that Terry itself embraced the principle of proportionality.125 That is to say, Terry was grounded on the idea that the government’s burden in justifying a search or seizure should be inversely proportional to the law enforcement interest in effectuating that search or seizure. Unfortunately, the Court has applied the proportionality principle in its Fourth Amendment reasonableness inquiries only infrequently since Terry.126
Our model calls for a return to Terry’s fundamental principle by encouraging courts to explicitly incorporate offense-severity into their analyses of stop-and-frisks. The model’s underlying premises are twofold. First, because legislative offense classifications represent considered democratic judgments about offense-severity, they are entitled to substantial deference. Second, the strength of the government’s law enforcement interest depends on the severity of the underlying crime that it is seeking to detect or prevent.
We think these premises represent a commonsense understanding of reasonableness capable of accommodating both the dignitary and privacy interests of individuals as well as the legitimate needs of law enforcement. But we also recognize that the utility of our proposal depends on its ease of application. To that end, in this Part we offer a brief description of how our model would work in practice.
Under our model, a court reviewing a pedestrian stop would begin by identifying the suspected offense and how the relevant jurisdiction categorizes it. Significantly, it would do so prior to interrogating the objective reasonableness of the particularized facts cited in support of the officer’s actions. Where the offense in question is a civil violation or infraction, the court would deem the stop and any subsequent frisk unjustified, and its work would come to an end. To be sure, if there were debate as to whether the stop was based on reasonable suspicion or probable cause, the court would need to inquire into the facts surrounding the stop. But stops based on the mere suspicion of noncriminal behavior would be categorically invalid.
In contrast, if the offense in question met the constitutional definition of a petty misdemeanor, the stop or stop-and-frisk would be deemed presumptively invalid. To overcome this presumption, the prosecution would bear the burden of demonstrating that the officer reasonably believed the suspected offense posed an immediate threat to public safety.127 Accordingly, a court following our model would only proceed to consider the objective and particularized facts underlying the basis of a stop-and-frisk in three scenarios: (i) where there is a dispute as to whether the officer had reasonable suspicion or probable cause; (ii) where the officer claims that a suspected petty misdemeanor constituted an immediate threat to public safety; or (iii) where the suspected crime was a non-petty misdemeanor or felony. The penalties attached to the latter crimes are prima facie evidence of their severity and the substantial law enforcement interest in combating them.
Our offense-severity model would also apply to an officer’s decision to frisk a suspect. While the stop and the frisk involve analytically distinct inquiries, courts routinely consider the nature of the suspected offense as a factor in the frisk analysis—as they should.128 But they typically treat offense-severity as a one-way ratchet that permits officers to automatically frisk where the suspected crime is one in which the suspect is likely to be armed and dangerous.129 Our proposal encourages courts to treat frisk justifications with greater skepticism where the suspected offense is minor. Logically, the offense that gives rise to a stop ought to inform the reasonableness of an officer’s fear that a suspect is armed and dangerous.
Consistent with current case law, our model would primarily apply to stops that are initiated on the basis of reasonable suspicion rather than probable cause. A police officer who witnesses a civil infraction has probable cause to either temporarily detain the suspect for the purpose of issuing a citation or, where permitted, effectuate an arrest.130 Our proposal respects that power. However, in jurisdictions where the legislature has not authorized arrest for violations and infractions, even where committed in an officer’s presence, frisks would be subject to a rebuttable presumption of unreasonableness.
IV. doctrinal and normative justifications for our offense-severity model
In this Part we argue that our model offers a practical mechanism for incorporating offense-severity into Fourth Amendment jurisprudence, finds support in Terry and its progeny, and is buttressed by a number of public policy considerations. Our model assumes added importance in light of recent moves by state legislatures to decriminalize marijuana possession and other minor offenses.131
Post-Terry, courts and scholars have disagreed on whether and how to assess the severity of an offense when determining Fourth Amendment reasonableness. As described in Part I, the Supreme Court has generally refused to define the reasonableness of a search or seizure according to the severity of the offense at issue.132 In Atwater v. City of Lago Vista, for example, the Court held that the Fourth Amendment does not forbid warrantless arrests for even minor crimes, such as nonjailable seatbelt violations.133 In the rare instances where the Court has offered an explanation for its decision to forgo crime-severity analysis, it has focused on administrability problems. The Court’s concerns are threefold: first, courts have no principled way of dividing serious crimes from minor ones;134 second, officers on the street cannot be expected to distinguish categories of crime;135 and third, even if officers could differentiate between crimes in a broad sense, they cannot be expected to master the subtle distinctions of complex penalty schemes, such as whether “the weight of the marijuana [is] a gram above or a gram below the fine-only line.”136
Scholars insist, on the other hand, that it is problematic to treat all offenses identically.137 But they cannot agree on how or where to draw the lines. Professor Jeffrey Bellin, for example, has argued for the classification of crimes into “grave,” “serious,” and “minor.”138 Under Bellin’s formulation, courts “channeling the views of a hypothetical reasonable person” would make these determinations.139 Yet, relying on post hoc judicial determinations poses the very administrability problem the Supreme Court has warned about. Other commentators to consider the issue have similarly failed to adequately address the administrability concern.140
These failures are part of what makes our proposal attractive. Since our model piggybacks on preexisting legislative classifications and a settled bright-line constitutional rule, there is no need to worry about the inconsistencies of ad hoc judicial lawmaking, thereby addressing the first of the Court’s administrability concerns. By contrast, given the Supreme Court’s reluctance to adopt any kind of judicially created hierarchy of crimes, scholarly proposals that rely on judicial determinations alone are impractical.141
Likewise, under our proposal, police officers need not master a complex set of vaguely drawn distinctions like those between “grave,” “serious,” and “minor” crimes; this addresses the Court’s second administrability concern. Of course, in the real world, “officers can and regularly do make ex ante judgments that separate one class of crimes from another, because that is an important part of their job.”142 But if, as the Supreme Court suggests, such difficult line-drawing judgments are generally to be avoided, the distinctions between noncriminal, petty, and serious offenses are straightforward and “can be easily identified both by the courts and by officers faced with a decision to [stop or] arrest.”143 Indeed, in analyzing stops for completed offenses, some federal circuit courts already require police to distinguish misdemeanors from felonies.144 Other courts, while declining to embrace such a bright-line rule, have effectively adopted a presumption that officers should not stop people for completed misdemeanors unless those misdemeanors threaten public safety.145 If courts believe police can readily distinguish any completed misdemeanor from a felony, there is little reason to suspect they cannot similarly distinguish violations and petty offenses from more serious crimes.
The Court’s third administrability concern—that the boundary between offense categories is too thin—does not pose an intractable problem for our model. For every crime, officers must know the elements of the crime in order to initiate a proper stop. For instance, statutes decriminalizing simple possession of marijuana specify an ounce limit; above that limit, possession remains a crime. The Atwater Court made much hay of this distinction by noting that officers executing arrests cannot be expected to know whether “the weight of the marijuana [is] a gram above or a gram below the fine-only line.”146 That may be true in the arrest context, but in terms of Terry stops based on reasonable suspicion such a concern is misplaced. Quantity is an obvious element of the offense. In order to have reasonable suspicion of criminal (or non-petty) activity, an officer must necessarily possess some indication that the quantity is of a criminal (or non-petty) amount. Importantly, that does not mean that all well-founded suspicions will turn out to be correct.
A recent case in Massachusetts, where possession of small amounts of marijuana has been decriminalized, deftly handled this issue.147 We discuss the case more fully in Part V below. Here, it is sufficient to note the court held that where police officers have reasonable suspicion that someone possesses marijuana, but have no indication that the amount exceeds the criminal limit, reasonable suspicion of a crime is lacking.148 Observation of small amounts of marijuana or detection of marijuana odor does not adequately support the inference that an individual is engaged in criminal activity. The court’s decision implies that officers must be able to point to some indicia of quantity—an informer’s tip, visual observation by the officers, drug paraphernalia in plain view, or even a bulge in the suspect’s pants pocket—to justify reasonable suspicion.149
Our proposed model admittedly represents a tradeoff between crime suppression and individual rights. After all, one of the central tenets of modern order-maintenance policing is that minor acts of legal wrongdoing may be indicative of more serious illegality. Notwithstanding powerful criticisms of the efficacy of order-maintenance policing,150 we acknowledge the potential for our model to reduce crime suppression at the margins. Such tradeoffs, however, are generally an unavoidable aspect of criminal procedure law. As Judge Scheindlin stated in Floyd: “Many police practices may be useful for fighting crime—preventive detention or coerced confessions, for example—but because they are unconstitutional they cannot be used, no matter how effective.”151
We also think this particular tradeoff is justified for doctrinal and policy reasons. While perhaps less intrusive than other kinds of searches and seizures, the dignitary harms to individuals subjected to a stop-and-frisk are real and substantial. “[I]t is simply fantastic,” the Terry Court wrote, “to urge that such a procedure performed in public by a policeman while the citizen stands helpless, perhaps facing a wall with his hands raised, is a ‘petty indignity.’ It is a serious intrusion upon the sanctity of the person, which may inflict great indignity and arouse strong resentment, and it is not to be undertaken lightly.”152 That is why Terry’s focus, as one state’s highest court recognized, was on “preventing crimes, and promoting the interests of justice in arresting felons,” which suggests the government’s interest in preventing and investigating lesser offenses should not be accorded the same weight.153
Beyond its doctrinal consistency, an offense-severity model is also worth adopting for at least three public policy reasons. First, our society should aim to close the gap between stop-and-frisk doctrine and practice that has emerged in the decades following Terry. As one court has held, “In light of the lower risk to society involved with civil infractions, the common law principle . . . suggests that a less intrusive procedure would be more acceptable than with the commission of a felony or even a misdemeanor.”154 Our offense-severity model aims to close the Terry gap by linking practice more faithfully with the public and officer safety concerns that permeated the Terry opinion. Since minor public order offenses implicate safety to a lesser degree than do criminal offenses, public policy interests counsel against upholding such stops.
Second, embracing offense distinctions provides courts with an objective basis for preventing law enforcement from using stop-and-frisks pretextually.155 As Judge Friendly noted in reference to a stop for narcotics possession, “[t]here is too much danger that, instead of the stop being the object and the protective frisk an incident thereto, the reverse will be true.”156 Indeed, as described in Part I, critics of Terry have complained that officers routinely engage in pretextual stops based on impermissible factors such as race. Employing our model would help divorce stop-and-frisk procedures from these problematic practices.
Finally, making use of offense-severity distinctions respects the considered judgments of legislatures and voters described in Part II. Those judgments are entitled to deference because crime control is a quintessentially local matter. As one scholar has noted, “[i]t . . . makes intuitive sense that the substantive definition of crimes should emanate from locals, who at once can give expression to specific social and geographic conditions, and, as the criminal law does more generally, single out particular behaviors for sanction.”157
There is a danger, of course, that deferring to legislative classifications will lead to a one-way ratchet whereby legislatures transform every civil infraction into a criminal offense, and every petty misdemeanor into a major one. Scholars have, in fact, repeatedly accused legislatures of engaging in such overcriminalization.158 But the opposite has actually occurred in recent years.159 For a variety of reasons, legislatures have increasingly embraced decriminalization.160 Hence, one of the primary criticisms of legislatures—that they always favor expansive substantive crime definitions to the detriment of defendants—is empirically false. Moreover, while we are not oblivious to this danger, we do think it overstated. As the Atwater Court pointed out in refusing to create a constitutionalrule against warrantless misdemeanor arrests, it is “only natural that States should resort to this sort of legislative regulation, for . . . it is in the interest of the police to limit petty-offense arrests, which carry costs that are simply too great to incur without good reason.”161 And while the leeway Terry has traditionally afforded officers may tip the scales slightly, it is unlikely to be a legislature’s predominant concern when considering offense classifications.
Critics might also object that offense distinctions vary considerably among jurisdictions. Indeed, as noted above, the Supreme Court in Moore ridiculed the idea that “Fourth Amendment protections” could “be made to turn upon such trivialities.”162 Our model incorporates the constitutional petty offense line in part to address this criticism. A general presumption against Terry stops for petty crimes reduces offense variability while also recognizing the comparably lesser law enforcement interest in detecting and prosecuting such crimes. Moreover, while the lack of uniformity among jurisdictions is regrettable, such distinctions are not altogether meaningless. Federalism embraces the idea of localities as laboratories for experimentation.163 That localities choose to define offenses as noncriminal rather than criminal, for instance, is indicative of such experimentation. And as the accelerating trend towards marijuana decriminalization suggests, those who favor shrinking criminal liability often exercise considerable influence, notwithstanding frequent scholarly claims to the contrary.
V. applying the offense-severity model
Judicial opinions are skewed towards discussing citizen-police encounters that result in the seizure of incriminating evidence. Relying on those opinions to analyze the constitutionality of Terry stops is therefore problematic because it tends to highlight the tired debate of whether it is ever desirable to let the criminal go free because the constable blundered.164 What often gets omitted from the surrounding discussion are the countless searches and seizures that do not result in the discovery of contraband. Consider that fewer than one in nine recorded stop-and-frisks by the NYPD leads to an arrest or summons.165 By any measure, that is an alarming rate of false positives for encounters based on reasonable suspicion. And citizens victimized by such false positives suffer a real invasion of privacy, the collective costs of which are nearly impossible to quantify.
Because the law is nevertheless shaped by judicial opinions, our focus in this Part is on state and federal court cases that adopt elements of our offense-severity model. These cases demonstrate that our model provides a practical and doctrinally sound mechanism for applying proportionality analysis in reviewing stop-and-frisks.
A. The Noncriminal/Criminal Distinction
This Section examines two state supreme court cases that embrace a noncriminal/criminal distinction for Terry stops. Each case stands for the proposition that police may not engage in Terry stop-and-frisks where the offense suspected is noncriminal in nature. As such, they represent an important limitation on order-maintenance policing tactics.
In State v. Duncan, a stop initiated on suspicion of an open container infraction resulted in the discovery of a gun and stolen goods.166 Seattle police officers approached three men at a bus stop to question them about a nearby half-empty beer bottle.167 After smelling alcohol on one of the men, Demetrius Duncan, the officers cited him for possession of an open container, a civil infraction under Washington law.168 Subsequently, based on the officers’ knowledge of Duncan’s record and his bulky jacket, they frisked him, finding a handgun, a stolen purse, and stolen credit cards.169 Duncan was charged with unlawful possession of a firearm, possession of a stolen firearm, and possession of stolen property.170 The trial court granted his motion to suppress, but the Washington Court of Appeals reversed.171
In reinstating the trial court’s order, the Supreme Court of Washington rejected the state’s invitation to “extend” the Terry exception to non-traffic civil infractions.172 Instead, it clarified that Terry applies only to criminal behavior. The court reasoned that because noncriminal offenses involve a lower safety risk than criminal ones, their detection warrants a less intrusive procedure than that sanctioned by Terry.173 Accordingly, the court found the stop unconstitutional.174
More recently, in Commonwealth v. Cruz, the Massachusetts Supreme Judicial Court reached a conclusion similar to Duncan in a case involving a suspected marijuana offense.175 The court held that “the lesser standard of reasonable suspicion is tied, by its very definition, to the suspicion of criminal, as opposed to merely infractionary, conduct.”176 The primary question in Cruz was whether marijuana odor provided justification for officers to order the defendant, a passenger in an illegally parked vehicle, to exit the vehicle and submit to a search of his person. The officers in Cruz were driving down a Boston street when they spotted a car parked in front of a fire hydrant.177 The officers recognized the car’s passenger, Benjamin Cruz, from his previous encounters with law enforcement.178 Suspicious, the two officers pulled up beside the driver for the ostensible purpose of investigating the civil offense.179 As they approached the vehicle, an officer observed Cruz smoking a cigar and smelled a “‘faint odor’ of burnt marijuana.”180 Based on that odor, together with the driver’s nervous behavior and statement that he had smoked marijuana earlier that day, officers ordered both the driver and Cruz out of the car.181 Prompted by officer questioning, Cruz then acknowledged having crack cocaine on his person.182
Much like the Supreme Court of Washington in Duncan,the Cruz court stated that “to order a passenger in a stopped vehicle to exit based merely on suspicion of an offense, that offense must be criminal.”183 The court found that, because Massachusetts had decriminalized possession of less than one ounce of marijuana,184 to order Cruz to exit the vehicle, the police officers would have needed reasonable suspicion that he possessed more than an ounce.185 Mere odor, even when combined with the defendant’s statement that he had smoked earlier that day, was not sufficient.186 Consequently, the court held that the crack cocaine should have been suppressed as the fruits of an illegal seizure.187 Because the initial “stop” of the vehicle was justified by the parking violation,188 officers were permitted to ask the defendant whether he had been smoking marijuana, but only for the purpose of issuing a civil citation.189
Readers may legitimately wonder about the practical implications of Duncan and Cruz as proxies for the consequences of implementing our model more broadly. Because Duncan was decided more than a decade ago, Washington’s experience is more revealing. On the positive side, Duncan has not caused Washington’s legislature to ratchet up its classification of minor offenses in order to expand the scope of permissible predicates for Terry stops. Indeed, the legislature has since authorized a ballot measure, which voters subsequently approved, to legalize possession of small amounts of marijuana. The Seattle Police Department (SPD), in turn, recognized that public consumption of marijuana is merely an infraction, the policing of which constitutes the Department’s “lowest law enforcement priority.”190 Additionally, marijuana odor does not provide probable cause for a vehicle search.191 SPD recently went so far as to return confiscated marijuana to street dealers because it fell below the legally prohibited amount.192 Duncan itself has been followed by Washington appellate courts193 and extended to parking infractions.194 It has also been cited in the Washington Association of Prosecuting Attorneys’ manual on searches and seizures, which instructs prosecutors and officers that Terry stops may not be based on suspicion of non-traffic infractions.195
On the other hand, evidence that Duncan has directly affected police practices is lacking. That can be blamed in part on the fact that the SPD did not keep data on Terry stops until it agreed to do so pursuant to a recent settlement agreement with the Department of Justice (DOJ).196Consequently, it is impossible to compare police practices before and after Duncan. We do know, however, that SPD policies in recent years have been troubling. The DOJ’s summary of its investigative findings states that “SPD’s policy and practices blur the line between a social contact or casual encounter, and a temporary investigatory detention pursuant to Terry v. Ohio.”197 It also noted that according to a 2009 report by SPD’s Office of Professional Accountability, forty percent of those subjected to stop-and-frisks believed officers lacked a reasonable ground for stopping them.198 In recommendations for a proposed consent decree, the ACLU similarly found a perception that SPD “engage[s] in contacts with people of color on suspicion of minor infractions or misdemeanors.”199 It recommended that a host of civil violations and misdemeanors—including “jaywalking and other pedestrian infractions, obstruction, disorderly [conduct], littering, and pedestrian interference”—be added to SPD’s “Lowest Law Enforcement Priority” list.200 The settlement agreement between the city and DOJ requires SPD to revise its manual to “prohibit investigatory stops where the officer lacks reasonable suspicion that a person has been, is, or is about to be engaged in the commission of a crime.”201 While this voluntary revision is welcome, SPD’s history provides all the more reason for courts to rigorously scrutinize the purported justification for Terry stops and insist on a clear distinction between noncriminal and criminal behavior.
B. The Petty Offense Distinction
The Supreme Court has never considered whether a minor criminal offense like a petty misdemeanor is sufficient to justify a Terry stop-and-frisk. However, its rationale in United States v. Hensley lends some support to our petty offense distinction. That lower federal courts and state courts have invoked Hensley in invalidating stops for completed misdemeanors further underscores the practical utility and doctrinal soundness of our model.
In Hensley, the Court considered a motor vehicle stop based on an officer’s reasonable suspicion of a completed felony. On December 4, 1981, two armed men robbed a tavern in St. Bernard, Ohio.202 Based on an informant’s tip, police officers issued a “wanted flyer” for Thomas Hensley for the purposes of investigating the aggravated robbery.203 On December 16, 1981, police officers spotted Hensley in a white Cadillac convertible.204 Confusion ensued, however, about whether there was a warrant outstanding for Hensley’s arrest, which clearly would have permitted a stop. Before a police dispatcher could confirm whether Hensley had an outstanding warrant, officers stopped him and searched his car.205 The search produced three handguns; Hensley was arrested, charged, and convicted of being a felon in possession of firearms.206
Prior to Hensley, theSupremeCourt had not considered the lawfulness of investigatory stops for completed, as opposed to imminent or ongoing, crimes.207 The Hensley Court determined that such stops were in fact lawful, but explicitly cabined its holding to completed felonies only.208 While declining to define the precise limits on investigatory stops for all completed crimes,209 the Court advised that any test to identify such limits would have to balance “the nature and quality of the intrusion on personal security against the importance of the governmental interests alleged to justify the intrusion.”210
Hensley emphasized the particular threat that felonies pose to public safety. The Court explained that “the law enforcement interests at stake in these circumstances outweigh the individual’s interest to be free of a stop and detention that is no more extensive than permissible in the investigation of imminent or ongoing crimes.”211 Hence, by negative implication at least, the Court’s opinion suggests minor crimes may not pose a threat significant enough to justify Terry stop-and-frisks.
Both federal and state courts have relied on Hensley to invalidate the use of Terry stops for completed misdemeanors. Although these cases also involve completed offenses, their holdings similarly emphasize law enforcement’s comparably weaker interest in preventing and prosecuting misdemeanors, as opposed to felonies.
The Ninth Circuit in United States v. Grigg, for example, held that mere suspicion of a completed misdemeanor could not justify a Terry stop when the offense posed only a minimal threat to public safety.212 The case involved the misdemeanor offense of playing a car stereo too loudly, a crime easily distinguished from that in Hensley: “[I]t is difficult to imagine a less threatening offense than playing one’s car stereo at an excessive volume. The absence of any danger to any person arising from the misdemeanor noise violation here does not support detaining the suspect as promptly as possible.”213
The Griggcourt acknowledged that the “nature of the misdemeanor” must be taken into account for Fourth Amendment purposes.214 But it “decline[d] to adopt a per se standard that police may not conduct a Terry stop to investigate a person in connection with a past completed misdemeanor simply because of the formal classification of the offense.”215 Instead, the court proposed a case-by-case balancing test that considers the “totality of the circumstances.”216
State courts have reached similar conclusions regarding the significance of offense-severity. In Blaisdell v. Commissioner of Public Safety, for example, the Minnesota Court of Appeals invalidated the stop of a vehicle whose driver was suspected of having committed a “no-pay” theft from a gas station.217 The theft was a misdemeanor offense under Minnesota law.218 Drawing on Hensley, the court recognized that misdemeanors are inherently less severe than felonies.219 The court explained:
Obviously, and by definition, misdemeanor offenses are punished less severely than gross misdemeanors or felonies. Additionally, the legislature has provided that an officer may not make a warrantless arrest for a misdemeanor unless the offense is committed in the officer’s presence. We consider this to be a legislative recognition that the public concerns served by warrantless misdemeanor arrests are in some degree outweighed by concerns for personal security and liberty. At the very least, because misdemeanor offenses are considered less serious crimes than felonies and because police cannot arrest for misdemeanors unless the offense is committed in their presence, the public concerns served by seizures to investigate past misdemeanors are less grave than the concerns served by seizures to investigate past felonies and gross misdemeanors.220
For these reasons, the court in Blaisdell imposed a per se rule that Terry stops are impermissible for past crimes that do not meet the statutory definition of a gross misdemeanor.221
The distinction between completed and ongoing crimes is undoubtedly relevant to determining whether police action is reasonable under the Fourth Amendment, but it should not be courts’ only, or even predominant, concern. True, the exigencies Officer McFadden encountered in Terry are absent where the suspected offense has already occurred.222 But the government nonetheless has a “strong interest in solving crimes and bringing offenders to justice.”223 Offense-severity seems to us a far more relevant barometer of the state’s interest in preventing and prosecuting crime than does the suspected offense’s temporal proximity to the search or seizure. And Hensley, Grigg, and Blaisdell all recognized this role for offense-severity in analyzing the Terry stops in question. Furthermore, while lower court decisions since Hensley have tended to focus exclusively on the felony/misdemeanor distinction, the petty offense distinction provides a more administrable dividing line for constitutional purposes, as we have suggested in this Note. That is because, in contrast to the felony/misdemeanor distinction, which the Court has previously described as “highly technical” and “arbitrary,”224 the petty offense distinction is readily identifiable and capable of easy implementation both on the street and in the courtroom.
Our Note has several important implications. First, defense lawyers should be more aggressive in challenging the lawfulness of stops based on suspicion of infractions and other minor offenses. Our sense is that the problem is exacerbated by the fact that much objectionable conduct escapes judicial review entirely. Second, courts should make clear that reasonable suspicion of civil violations does not justify Terry stops under the Fourth Amendment or equivalent state constitutional provisions. While decisions invalidating Terry stops on the basis of state statutes would be welcome, courts should go further by establishing a constitutional floor below which states may not go. Third, frisks that occur subsequent to stops for minor offenses should be subject to heightened judicial scrutiny. Logically, the severity of an offense should inform an officer’s determination of a suspect’s potential dangerousness. Finally, opponents of stop-and-frisk practices should give serious consideration to lobbying state legislatures to decriminalize certain offenses—trespass, disorderly conduct, and drug possession, for example—that commonly serve as bases for intrusive Terry stops.
To be sure, applying an offense-severity distinction to Terry stops is not a panacea for all that ails Fourth Amendment law. But it is a step in the right direction. At a minimum, distinguishing merely infractionary conduct from that which is criminal would assuage some of the concerns of Terry’s opponents. Doing so might also restore faith in the initial rationale for the Terry exception: ensuring police are not engaged in fishing expeditions, but are instead focused on preventing and solving the most serious crimes.