The Bilateral Fourth Amendment and the Duties of Law-Abiding Persons

L. Rush Atkinson


Ordinary people take countless measures to avoid being searched by police or other government agents.  To get an inkling of how frequently the innocent alter their behavior, consider the responses of law-abiding people to recently enacted counterterrorism programs.  When faced with delays and discomfort caused by more rigorous baggage screening procedures in airports after the September 11 attacks, thousands of travelers elected to use cars instead of flying on short-haul flights; this combined response costs airlines billions of dollars and has been linked to dozens of vehicular deaths (since driving is more dangerous than flying, even in the post-9/11 era).  Similarly, in the wake of FBI prosecutions of Muslim charities for assisting terrorist organizations, private philanthropic donations to Islamic organizations decreased as a result of fears that donating would raise red flags and warrant FBI interviews.  The government’s “Terrorist Surveillance Program”—the warrantless wiretapping program—also had widespread ramifications among the innocent public; journalists, lawyers, and academics allegedly elected to forego using phones for professional business activities, out of fear their conversations were being heard by third parties.

Few, if any, of the people described above adjusted their behavior for criminal purposes; rather, their shared goal was to avoid the inconvenience and costs of what this Article calls “empty” searches—errant searches by government officials of innocent persons.  Even if you’re innocent, being searched by the police can be an unpleasant and costly venture.  Most tangibly, empty stops and searches cost innocent persons time.  The major losses, however, are intangible and include loss of bodily integrity, dignity costs, violation of personhood, loss of freedom, damage to reputation, and loss of privacy.

The precautions taken by the innocent to avoid empty searches also can be traced to the Fourth Amendment, which protects persons—even those who are innocent—only from “unreasonable” searches.  Under the Fourth Amendment, the government is only liable for damages when its agents fail to exercise the constitutionally mandated standard of “reasonableness,” which usually means probable cause or reasonable suspicion (depending on what type of search agents want to conduct) and a warrant (unless an exception applies).  If officers satisfy these two requirements but still end up “getting it wrong”—that is, searching an innocent suspect—the suspect has to bear the costs of the search.  Given this limited protection, law abiders have to take precautions in order to avoid bearing the real brunt of search-related costs.

The question this Article answers is why the Fourth Amendment only offers this limited protection, and why the innocent public must bear the costs of government investigations if the officers were “reasonable.”


The “reasonableness” limitation within the Fourth Amendment was not inevitable.  To protect persons against government intrusion, the First Congress could have implemented any of a variety of rules.  Consider the following two rules that would protect the innocent from unreasonable searches:

(1) The government shall be liable for all damage to innocent activity resulting from searches.

(2) The government shall be liable for all damage to innocent activity resulting from unreasonable searches.

These two rules constitute the most common classes of rules existing in tort law: the first is a rule of strict liability, the second a negligence rule.  The second rule also happens to be the effective rule embodied by the Reasonableness Clause of the Fourth Amendment—“The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated . . . .”  Others have noted that the Fourth Amendment effectively functions as a negligence regime, and with the diminishing importance of the Warrant Clause today, the Amendment functions more and more identically to traditional tortious regimes. A strict liability regime, like the first hypothetical rule above, offers more robust compensation for the innocent; regardless of whether a search was “reasonable” or not, a strict-liability Fourth Amendment would make the government liable for the costs of empty searches. (Of course, the government action would have to be deemed a “search” within the meaning of the term under Fourth Amendment jurisprudence.  Consequently, the innocent would still bear liability under the hypothetical rule for costs associated with minimal government intrusions.)

The writers of the Amendment eschewed a strict liability rule in the Fourth Amendment, but not because it was a foreign concept to them.  Indeed, the First Congress adopted a strict liability rule within the Bill of Rights—specifically, within the Takings Clause of the Fifth Amendment.  Under this clause, the government must compensate an injured party whose land is “taken,” regardless of the reasons for the exercise of eminent domain.  Thus, even if the seizure of land is completely reasonable, the government remains liable for the costs it has imposed.

The difference between the structure of the Fourth Amendment and the Takings Clause is more than just semantics.  Had the Fourth Amendment been crafted in the same form as the Takings Clause’s image, the government would be constitutionally required to compensate innocent persons for every house search, every Terry stop, every police frisk, every airport or subway stop, every DUI checkpoint, and so on.  Moreover, compensation would be required regardless of the level of suspicion the police had before effectuating the search, eliminating the ability for police to insulate themselves from liability by claiming probable cause or reasonable suspicion.  The difference between these two constitutional provisions, therefore, reinforces the earlier question posed: why does the Fourth Amendment only protect innocent persons from “unreasonable” government searches?


The most satisfying explanation for the “reasonableness” caveat within the Fourth Amendment starts by considering why errant searches of innocent persons occur.  Certainly, sometimes police are just reckless and search people willy-nilly, figuring that by thrashing the bushes, criminals will be exposed.  More often, however, police respond to suspicious evidence. This evidence, in turn, comes in two flavors.  Some factors—such as race, gender, and age— are exogenous indicia, meaning a suspect cannot affect whether the factor is present or not.  More commonly, however, suspicious factors are endogenous, meaning that suspects can in fact alter the odds of the factor being present.  Flight from police, for instance, is endogenous because a suspect can choose whether to flee or not.  Other endogenous factors include the wearing of gang colors,  utterance of incriminating statements, and possession of materials that, while not illegal, are typically associated with criminal activity (such as large amounts of cash, laundry detergent in a vehicle, and ski masks).

Most indicia of criminality are endogenous, meaning that in most cases, innocent suspects can affect the chances of there being an “errant search.”  Most descriptions of the Fourth Amendment assume that only the government can affect the probability of empty searches.  In reality, persons potentially subject to search—including the innocent—can take a variety of actions to limit the likelihood of searches.  Refraining from standing on street corners helps avoid being mistaken for a drug dealer; not carrying Arabic texts on an airplane is thought to reduce suspicion of being a terrorist; avoiding luxury cars lessens suspicion that one is driving a stolen car.

The law-and-economics school refers to conditions where both the injurer (here, the government) and victim (here, the innocent suspect) have the ability to control the rate at which accidents (here, the empty search) occur as “bilateral” (or “two-actor”) scenarios.  Previously, economic analyses of bilateral conditions have focused mainly on bilateral conditions in the torts context—cases such as when both a pedestrian and a vehicle driver can take precautions to avoid a collision.  But the analysis applies with equal force in the Fourth Amendment context.

When both the injurer and the victim can take steps to avoid accidents, optimal rules that minimize the chances of accidents must take into account the likely responses of two parties.  Regulators must also account for the fact that actors will not just respond exclusively to the legal rules set by the government; actors in bilateral conditions will also shape their behavior based on how they believe other actors will respond.  If a pedestrian believes a driver will exercise an abundance of caution to avoid hitting pedestrians, the pedestrian has less incentive to exercise her own care; if the same pedestrian thinks a driver will blow through crosswalks without care, the pedestrian will rationally use greater vigilance when crossing.

Based on the likely strategic responses of actors in a bilateral scenario, a strict liability regime cannot induce actors to exercise the socially optimal level of care.  This result applies in the search as well as tort context.  Consider, for instance, the hypothetical rule offered above:  The government is liable for all damage to innocent activity resulting from searches.  Because the innocent will be fully compensated by the government for all losses resulting from empty searches, law abiders are indifferent to the number of empty searches that occur—they lose nothing.  As a result, the innocent will take no precautions to avoid empty searches, refusing to reduce their suspicious behavior in any way.  In order to avoid liability for searching the increasing number of suspicious-but-innocent people, the police may respond by taking too much care—that is, searching only when they have significant suspicion of a suspect’s guilt.  As a consequence, the level of care will be socially inoptimal because too few searches will be performed, and not enough criminals will be apprehended.

In bilateral scenarios, negligence regimes generally perform better than strict liability rules in inducing both parties to take the socially optimal level of care.  Under a negligence rule, the injurer takes the degree of care required by the judicially established standard of care, meaning that here the government will search only when it satisfies the required level of suspicion; law abiders, now facing the possibility of bearing the costs of the empty searches, reduce their suspicious activity in order to reduce the chances of the police satisfying that threshold.  Negligence, then, can induce both parties to adapt their behavior, a result strict liability regimes fail to achieve.  If the judiciary is able to set the required level of care at the socially optimal level, it will impel not only government but also the innocent to take the socially optimal level of care.

The “bilateral” nature of errant searches provides a satisfying explanation of why the Fourth Amendment adopts a negligence regime:  The Amendment presumes that it is regulating a bilateral, rather than unilateral, relationship where both the searcher and the searched can be induced to take care to avoid empty searches.  Similarly, the difference between the Takings Clause and the Fourth Amendment can be explained by reference to this law-and-economics analysis.  In cases of eminent domain, the landowning “victim” whose land is seized was envisioned by the First Congress to have little control over the exercise of eminent domain exercised and the probability of the land seized.  Given that eminent domain is primarily a legislative decision, the owner’s power over the likelihood of eminent domain’s exercise is nearly non-existent, because he has only one vote.  The takings context, therefore, is presumed to be a “unilateral” scenario; in such cases, there is no need to take the “victim” into account, and therefore strict liability regimes offer a simpler way of achieving the optimal result.


Taking the analysis above out of the law-and-economics world and into a more doctrinal one, the bilateral Fourth Amendment can be understood as the Constitution’s imposition of reciprocal duties upon both the police and the general citizenry:  police are obliged not to search unless there is adequate suspicion, and individuals are obliged to avoid cumulative behavior that is inherently suspicious.  Put another way, both actors have a duty to act reasonably.  The claim here is stronger than the rights conception of the bilateral Fourth Amendment because it adds a normative element to the analysis.  Not only does the Fourth Amendment in practice regulate the activity of law-abiding persons and channel them away from certain actions, but it should regulate these activities.  Just as before, comparing the Fourth Amendment to torts regimes illustrates the point; the dissimilarities between the relationships regulated by these fields, however, reveals the rationale behind the “duty of the innocent” conceptualization.

In tortious negligence regimes, the victim is not thought of as having a duty to conform to some standard of care.  If a trespasser is only liable for disobeying “no trespassing” signs, landowners who want to deter trespassers will erect more signs than they would in a regime holding trespassers strictly liable regardless of signage.  Sign-posting is a rational response to an enacted legal rule, but a landowner has no duty to enact signs.  After all, the landowning “victim” is the one who bears the brunt of the costs when a trespass occurs, so the decision to enact signs should fall within her discretion.

In the search context, however, the entire costs of empty searches are not initially borne by the innocent as they are by victims of torts.  The stigma of the Fourth Amendment searches—the costs to dignity and personhood—are imposed primarily upon the law-abiding persons errantly searched by police.  But other costs arising from empty searches are more dispersed.  Empty searches, for instance, increase the costs of finding criminals; if the police must search innocent persons to find a criminal, it drains government resources and reduces law enforcement’s effectiveness.  Reducing the effectiveness of the police force, in turn, increases the amount of crime the rest of the population bears.

Because the costs of empty searches are not completely internalized by the innocent victims, we have a scenario different-in-kind from tortious victims.  Even though the innocent rarely intend to inhibit law enforcement, activities that are inherently suspicious have this effect and therefore require regulation. And were it possible, we would enjoin certain innocent activities in order to avoid empty searches that waste resources and impede law enforcement (indeed, this last point is reinforced in the next section).  For these reasons, it is proper to think of the innocent as having a genuine duty to act reasonably, even when we do not impose such a responsibility on analogous victims.

Federal courts have made little reference to the masking behavior of the innocent, the social costs that such behavior imposes, or any obligations that the innocent might bear.  Weighing the negative externalities of law-abiding behavior, however, provides support for the burdens already placed upon the innocent.  In Illinois v. Wardlow, for instance, the Court considered whether “unprovoked flight” from the police could by itself amount to reasonable suspicion. Chief Justice Rehnquist’s opinion focused on the probative value of flight and suggested there were few innocent explanations for such behavior.  The majority overlooked a second point in its favor: even if flight is innocent, the social costs that law abiders cause by fleeing officers are undeniable; innocent flight distracts law enforcement from other possible criminal activity in the area, minimizing the effectiveness of any patrol.  The social effect of flight, therefore, affirms the notion that law abiders should bear costs that their activities generate.

It may seem perverse to describe the Fourth Amendment as imposing a duty upon the law-abiding public, given that the innocent are thought to be the primary rights bearers under the Amendment and the government the typical party upon which the Bill of Rights imposes obligations.  But the appreciation that innocent activity masks criminality gives us reason to think that this behavior should be regulated and sometimes even stopped.  Ineffective searches are not simply the fault of law enforcement, despite the general assumption that police are to blame, and the searched are not the only party to be hurt by ineffective searches.  Instead, both the searcher and the searched bear responsibility for reducing empty searches.


The bilateral nature of searches, however, also influences the optimality of the specific rules used to establish suspicion.  In other words, law abiders’ ability to adjust their behavior based on the likelihood of being searched should inform our very notion of what “suspicion” is.

Traditionally, the evaluation of suspicion has been an exercise of assessing, as Judge Randolph of the D.C. Circuit once wrote, “conditional probabilities:  if one event occurs, how likely is it that another event will occur?” Such an approach derives from the intuition that if the police simultaneously observe two pieces of evidence that independently suggest a suspect committed a crime (e.g., bloody clothes and a fingerprint at the crime scene), observing both pieces of evidence gives us more confidence of her guilt than finding only one (e.g., the fingerprint alone).

Establishing suspicion through conditional probabilities has obvious benefits because it allows the police to search suspects that have not left a “smoking gun.”  This approach, however, has its own costly consequences.  For one, use of conditional probabilities affects the measures the innocent must take to avoid costly empty searches.  Because these measures can be so dramatic, a policymaker must consider how parties at risk of being searched will react to rules governing suspicion in order to fashion a socially optimal rule.  In certain cases, the cost of chilled innocent activities, combined with the costs of empty searches, will outweigh the value of searching those with the distinguishing characteristic in question.  For instance, consider the factor “associates with known criminals.”  This characteristic increases the likelihood that searching a suspect will uncover criminal evidence, though its probative value is too weak to be the only grounds for a search requiring probable cause or reasonable suspicion.  Its weak probative value, however, means it could be considered as evidence in a jurisdiction that employs conditional probabilities.  In the same jurisdiction, however, some innocent persons will reduce their associations with known criminals in order to avoid empty police searches.  The consequences of law abiders’ response could have an isolating effect upon at-risk members of society, increasing the chance of recidivism and escalating criminality.  So while using the evidence may lead to an additional search that reveals crime, allowing consideration of the evidence may have the net effect of increasing criminality.

These types of perverse outcomes do not mean that conditional probability analyses need to be abandoned; rather, evidentiary rules pertaining to suspicion merely need to be tempered with an inquiry about what innocent behavior is likely to be curbed if an activity is considered “suspicious,” and what the social value of that curbed innocent behavior is.  The social value of some curbed activity is marginal, and the crime reduced by considering an activity “suspicious” is likely greater than the social value of the marginal activity reduced. On the other hand, other equivocal activities have greater social value.  Ignoring the subject completely will lead to perverse rules about suspicion, and therefore some inquiry must be made.


This Article identifies how the innocent respond to rules fashioned by the Supreme Court and some implications of this reactionary ability. The bilateral nature of searches, and specifically the ability of law abiders to reduce their chances of being erroneously searched, provides a central reason why the Fourth Amendment must regulate searches under a reasonableness regime rather than a strict liability regime.  However, once we understand that a rule of reasonableness is necessary, the standard of care set by the Court—what we know doctrinally as probable cause or reasonable suspicion—is only socially optimal if the Court appreciates the effects its rules have on the innocent.  Consequently, the Court must fashion rules that weigh both the probative value of the activity and the marginal activity that will be lost if the Court accepts certain evidence as grounds to satisfy the standard of care.


L. Rush Atkinson is a Graduate Fellow at the Center for the Administration of Criminal Law at the NYU School of Law.

Copyright © 2011 Georgetown Law Journal

Post a Comment (all fields are required)

You must be logged in to post a comment.