The pitfalls of stop-and-frisk analysis
Of the various exceptions to the 4th Amendment's warrant requirement, there's probably none more frequently used, and certainly none more frequently litigated, than the "stop and frisk" doctrine created 42 years ago in Terry v. Ohio. The 8th District's decision last week in State v. White highlights three fundamental mistakes that courts make in those cases.
The facts in White are simple, and typical: the police, while patrolling a "high drug activity" area, observe four black males standing in front of a vacant, boarded-up building. The males see them, too, and start to walk away, and one "threw an object to the ground between two cars parked in the driveway of the property." The police, "concerned with officer safety because the object thrown could have been a weapon," ordered the men to the ground; they found six rocks of crack cocaine on the ground where the object had been thrown, and a small package of crack by one of the suspects' legs.
The court's decision affirming the denial of the motion to suppress is problematic; basically, the decision holds that throwing something to the ground in a "high crime" area gives the police reasonable suspicion to conduct not only a stop, but a frisk for weapons. Still, the basic problem is not the result but the process used to reach that result. Courts analyzing stop-and-frisk issues make three fundamental analytical errors.
The first is what might be termed "sample selection bias." Both the trial and the appellate courts gave significance to the fact that the males were seen on the property at a vacant, boarded-up house, although there was nothing to indicate that the males were doing anything more than standing around. The poor areas in Cleveland are dotted with vacant, boarded-up houses, and people standing around in front of them is hardly a novelty. Let's say that for every five such groups investigated by the police, one is engaged in criminal activity. No trial court, however, is ever going to hear about the other four: since no arrests were made, no cases were brought. As far as the judge is concerned, the presence of young males on vacant property is always indicative of criminal activity, because he only hears of cases where it was. A "furtive movement" can be an effort to hide drugs, or simply an effort to retrieve the vehicle registration from the glove compartment as a police officer approaches. Even if the latter occurs ten times for every one the former does, the latter will never show up in court; it will never become part of the dataset the judge uses in determining whether a "reasonable suspicion of criminal activity" can be inferred from the action.
Judges complicate this problem further by making assumptions about connections that the empirical evidence doesn't support, or even rejects. The connection between guns and drugs in White is an example. While there are abundant quotations in the case law to the effect that weapons and narcotics are inextricably linked, like this one from the 2nd District, the factual basis for that claim, other than TV cop shows, is not readily apparent, at least as it pertains to street dealers and consumers. The vast majority of drug cases I've handled have not involved a weapon, and the actual experience of other attorneys I've talked to corresponds to that in State v. Jones, where the 8th threw out a frisk of two suspects stopped for trying to flag down cars, noting that the officer acknowledged that of the several hundred drug arrests he’d made, only “five or six” turned up weapons.
But if empirical evidence is lacking as to various aspects of stop-and-frisk cases, it's not lacking as to the prevalence of that tactic, and its utility as a crime-fighting weapon. New York keeps detailed statistics on the stops and frisks that its police forces conduct, and the numbers are revelatory. In the first three months of 2009, New York's finest stopped 171,094 people. Less than 6% of those stops resulted in an arrest. Put another way, of every 17 people stopped by the police, 16 of them turned out to have done nothing.
Another number gleaned from the data is not surprising to anyone familiar with law enforcement: 87% of those stopped were African-American or Hispanic.
It has become de rigeur for judges to denounce the "substantial societal costs" imposed by the exclusionary rule: Scalia did it Hudson v. Michigan in 2006 (a diatribe that was largely echoed by the Ohio Supreme Court in State v. Oliver), and Roberts did it just last term in Herring v. US.
Again, though, that is because judges always see those costs; they don't see the costs imposed by an illegal search when it proves fruitless. Police and prosecutors routinely lament the lack of cooperation they get in investigating and prosecuting crimes in certain neighborhoods. Might that lack of cooperation stem from police practices that the residents of those neighborhoods find oppressive, such as the routine stop and detention of young black males on the flimsiest of pretexts? We know the societal cost involved in freeing a criminal because the evidence against him is suppressed. We don't know the societal cost in having to drop a prosecution because witnesses are unwilling to cooperate with someone they perceive as their oppressor. But that cost is there.
There are two ways to approach the 4th Amendment -- as an expression of a core liberty concept, or as a manual on police procedures, prescribing what law enforcement officers can and cannot do in various situations. The first offers a more consistent method of analysis, and the second poses the risk that the court will venture off into areas where its assumptions are untested and the full results unanticipated or unknown.