Perceptions of Fairness and Justice: The Shared Aims and Occasional Conflicts of Legitimacy and Moral Credibility

By: Josh Bowers* & Paul H. Robinson**

Introduction

A growing literature suggests that a criminal justice system derives practical value by generating societal perceptions of fair enforcement and adjudication.[1]  Specifically, perceptions of procedural fairness—resulting in perceptions of the system’s “legitimacy,” as the term is used—may promote systemic compliance with substantive law, cooperation with legal institutions and actors, and deference to even unfavorable outcomes.[2]  A separate literature suggests that a criminal justice system derives practical value by distributing criminal liability and punishment according to principles that track societal intuitions of justice.[3]  Specifically, perceptions of substantive justice—resulting in perceptions of the system’s “moral credibility”—would seem to promote compliance, cooperation, and deference.  By contrast, a criminal justice system perceived to be procedurally unfair or substantively unjust may provoke resistance and subversion, and may lose its capacity to harness powerful social and normative influence.[4]

This Article examines the shared aims and overlaps in operation and effect of these two criminal justice dynamics—the “legitimacy” that derives from fair adjudication and professional enforcement and the “moral credibility” that derives from just results—as well as the occasional potential for conflict.  Specifically, in this Article, we aim to isolate and define the parameters of each dynamic, to compare and examine their similarities and differences, and to explore the settings in which the two run together or (more rarely) cross-wise.  In this way, our overarching objective is to clear the air.  To date, legal scholars have tended to invoke the two dynamics too casually, to ignore one but not the other, or to conflate or confuse the two.  Thus, we intend to provide something of a primer: a useful and necessary analytic framework for ongoing debates into the advantages, limits, and dangers of moral credibility and legitimacy.  But we do not stop there.  We stake out tentative positions within these debates.  That is, we endorse the prevailing view that moral credibility and legitimacy are promising—indeed, critical—systemic enterprises, and we make a number of tentative claims about when and to what degree a system ought to pursue or prioritize each enterprise.  Particularly, we anticipate significant crime-control advantages for a system that enjoys perceptions of both moral credibility and legitimacy, but we conclude that—for empirical and theoretical reasons—moral credibility ought to be the principal objective in uncommon circumstances in which a system may effectively pursue only one.

In Part I, we explore the shared aims of legitimacy and moral credibility.  In Part II, we discuss the practices, procedures, and rules that produce, undermine, or implicate perceptions of procedural fairness and substantive justice, and we identify potential pitfalls and dangers for a criminal justice system committed to generating perceptions of fairness and justice.  In Part III, we confront the critical question of whether perceptions of fairness and justice effectively promote deference to law and legal authorities and institutions.  In Part IV, we attempt to explain why a criminal justice system may come to adopt and implement practices, standards, and rules that deviate from societal perceptions of fairness and justice.  Finally, in Part V, we examine the interesting issues raised when legitimacy and moral credibility conflict with one another, and we sketch our vision for how a system ought to resolve the tension.

I.  The Shared Aims of Legitimacy and Moral Credibility

A.     Legitimacy

In law, as in life, legitimacy is a term invoked so casually that it sometimes seems to signify little more than a vague aspiration.  However, in the criminal-justice context, the term has come to represent something more precise.  Criminologists, social psychologists, and political scientists have refined the concept to mean a “belief that legal authorities are entitled to be obeyed and that the individual ought to defer to their judgments.”[5]  In this Article, we focus principally on the work of Tom Tyler, not because he is a leading legitimacy theorist and empirical researcher, but because over the past two decades his work has generated the most attention in the legal academy, and we are particularly concerned with the ways in which legal scholars have, to date, used (and misused) his contributions.

Tyler has argued persuasively that the law’s legitimacy (or at least a perception of it) is critical to a well-functioning criminal justice system and to public safety more generally.  Specifically, effective crime control depends on volitional deference to substantive law and to its enforcement and adjudication.  And, significantly, perceptions of procedural fairness may well facilitate such deference.  The importance of the legitimacy project cannot, therefore, be oversold.  It is a terrifically promising enterprise that may serve to promote the very goals that are (and ought to be) central to criminal justice: compliance with statutory law and cooperation with legal authorities and institutions.

Procedure is legitimacy’s starting point.[6]  People come to obey the law and cooperate with legal authorities because they perceive their institutions to operate fairly.  In this way, perceptions of procedural fairness facilitate a kind of normative, as opposed to purely instrumental, crime control.[7]  Put differently, citizens of a procedurally just state comport their behavior to the substantive dictates of the law not because the state exercises coercive power (or, at least, not exclusively because of it), but because they feel a normative commitment to the state.  Unlike conventional deterrence theory, which presumes the necessity of carrots and sticks, legitimacy harnesses the power of internal commitment and volitional participation.[8]  Legitimacy replaces the Holmesian “bad man” with the “faithful man”—an individual who complies with the law not because he rationally calculates that it is in his best interest to do so but because he sees himself as a moral actor who divines that it is right to defer to legitimate authority.[9]

Critically, perceptions of procedural fairness are outcome independent.[10]  In other words, a defendant or victim need not realize her objective in order to conclude that enforcement or adjudicatory practices are legitimate.  Likewise, an ordinary citizen need not determine that the law expresses her personal notion of morality in order to accept its validity.  In this way, procedural fairness differs from outcome-driven normative and psychological approaches to criminal justice (like distributive justice generally and moral credibility specifically) that examine whether the law produces results that accord with communal intuitions of just deserts.[11]  Because the concept of procedural fairness is not dependent upon piecemeal review of substantive outcomes, positive or negative perceptions possess significant potential to motivate or undermine deference to power, thus transferring broad discretionary authority to the state.[12]  In this way, legitimacy may produce compliance and cooperation with not just an immediate enforcement effort but across codes and cases, and even actors and institutions.[13]  Thus, for legal authorities, cultivating perceptions of legitimacy is of particularly useful and flexible value.[14]

But what does the public perceive to be legitimate procedures and practices?  What minimum standards are shared across demographics and cultures?  We can provide no definitive answers to these questions in this space.  Nevertheless, a fair consensus has developed over the principal criteria that typify procedural fairness.  Legitimacy may be measured by the quality of decision making or the quality of treatment of defendants.[15]  More specifically, procedures are legitimate when they are neutral, accurate, consistent, trustworthy, and fair—when they provide opportunities for error correction and for interested parties to be heard.[16]  Legal authorities are legitimate when they act impartially, honestly, transparently, respectfully, ethically, and equitably.[17]  The criminal justice system that optimally expresses these values is not only morally defensible but also quite probably stable and effective.

B.     Moral Credibility

It has long been assumed that in determining how to distribute punishment—how much to whom?—the goals of doing justice and fighting crime inevitably conflict.  The traditional crime-control principles of deterrence, rehabilitation, and incapacitation of the dangerous would distribute criminal liability and punishment in ways quite different from the distributive principle of moral desert.  Retributivists and utilitarian crime-control advocates commonly saw their dispute as irreconcilable, and in a sense it is.  However, what has been referred to as the “empirical desert” or “moral credibility” literature has argued that, in another sense, these two fundamental aims of criminal justice may not conflict.  Doing justice may be the most effective means of fighting crime.[18]

The hitch is that it is not moral philosophy’s deontological notion of justice that has crime-control power, but rather the community’s shared principles of justice, what has been called “empirical desert.”  This turns out to be both good and bad for constructing a distributive principle for criminal liability and punishment.  On the one hand, unlike moral philosophy’s deontological desert, empirical desert can be readily operationalized—its rules and principles can be authoritatively determined through social science research into people’s shared intuitions of justice.  On the other hand, people’s shared intuitions of justice are not justice, in a transcendent sense.  People’s shared intuitions can be wrong.  In the end, however, the retributivist may find that an instrumentalist distributive principle of empirical desert will produce far more deontological desert than any other workable principle that could or would be adopted.[19]

As has been argued elsewhere, the crime-control benefits from distributing punishment according to people’s shared intuitions of justice are thought to arise from a variety of sources.[20]  Some of the system’s power to gain compliance derives from its potential to stigmatize, which can be a powerful, yet essentially cost-free, control mechanism for many offenders.  Yet a criminal law can stigmatize only if it has earned moral credibility with the community it governs.  That is, for conviction to trigger community stigmatization, the law must have earned a reputation with the community for accurately reflecting the community’s views on what deserves moral condemnation.  A criminal law with liability and punishment rules that conflict with a community’s shared intuitions of justice will undermine its moral credibility.

Another value of moral credibility comes from the fact that effective operation of a criminal justice system depends on the cooperation, or at least the acquiescence, of the system’s witnesses, jurors, police, prosecutors, judges, offenders, and others.  To the extent that people see the system as in conflict with their judgments of justice, that acquiescence and cooperation is likely to fade and be replaced with resistance and subversion.[21]  Subversion and resistance may take the form of either an impulse toward apathy or an impulse toward self-help.[22]  That is, people may turn to vigilantism in reaction to a perceived failure of justice.  More commonly, people may resist or subvert the system in less dramatic ways.  Witnesses may lose an incentive to offer their information or testimony.  Citizens may fail to report crimes in the first instance.  Jurors may disregard their jury instructions.  Police officers, prosecutors, and judges may make up their own rules.  And offenders may resist adjudication processes and punishments rather than participate in them.

An even greater power of moral credibility comes through a less obvious mechanism.  The real power to gain compliance with society’s rules of conduct lies not in the threat of official sanction but rather in the influence of the forces of social and individual moral control.[23]  It is the networks of interpersonal relationships, the social norms shared among those relationships and transmitted through those social networks, and the internalization of those norms that control people’s conduct.  The law is not irrelevant to these forces.  Criminal law plays a central role in creating and maintaining the social consensus necessary for sustaining moral norms.  In a society as diverse as ours, the criminal law may be the only society-wide mechanism that transcends cultural and ethnic differences.[24]  Thus, its most important real-world effect may be its ability to assist in the building, shaping, and maintaining of these norms.  It can help build and harness the compliance-producing power of interpersonal relationships and personal morality, but only if it has earned a reputation for moral credibility with its community.  A criminal law that has been found to be off the mark in its past condemnations and punishments can be simply dismissed as just “wrong again.”

The criminal law with moral credibility also can gain deference and compliance in the particularly difficult case of borderline or new offenses.  If the law has earned a reputation as a reliable statement of community views, people are more likely to defer to its commands as morally authoritative in those borderline cases in which the propriety of the conduct is unsettled or ambiguous in the mind of the actor.  This can be an important role.  In a society with the complex interdependencies of ours, seemingly harmless conduct can have seriously harmful consequences.  When the conduct is criminalized, one would want the citizen to respect the law even if he or she does not fully understand why it is forbidden.  Such deference is more likely where citizens have come to see the criminal law as accurate in announcing condemnable behavior.[25]

The extent of the criminal law’s effectiveness in all these respects—in harnessing the power of stigmatization, in reducing resistance and subversion to a system perceived as unjust, in facilitating, communicating, and maintaining societal norms, and in gaining compliance in borderline cases—is to a large extent dependent on the degree to which the criminal law has gained moral credibility in the minds of the citizens governed by it.  If the law assigns liability and punishment in ways that the community[26] perceives as consistent with its shared intuitions of justice, it gains deference, cooperation, and compliance.  If its judgments regularly conflict with community views, its work is undermined by those who see it as unjust.  Recent empirical studies have confirmed these effects of a system’s moral credibility.  The studies suggest that the greater the perception that the criminal law’s liability and punishment rules conflict with a person’s own judgments of justice, the less likely the person is to respect that criminal law as a moral authority and, therefore, the less likely to support, cooperate, and comply with that criminal law.[27]

II.  Shaping Reputation

What do we know about public perceptions of the legitimacy of police practices and adjudicative procedures?  What do we know about public perceptions of the moral credibility of liability and punishment rules?  In this Part, we explore what is understood about how criminal justice systems develop reputations for legitimacy or moral credibility, and we conclude that—at least when it comes to the legitimacy of specific enforcement and adjudication practices and procedures—what is understood is not yet enough.

A.     Creating Legitimacy

By now, scholars have tested legitimacy in a variety of legal contexts and even in social settings beyond the law.[28]  Our analysis is limited to a slice of the existing work that relates to two criminal-justice contexts: what constitutes professional and unprofessional enforcement practices, and what constitutes fair and unfair adjudicative procedures.  Within these narrow domains, scholars have done both substantial work and, we think, not quite enough.  They have examined perceptions of legitimacy among a host of subpopulations: suspects, defendants, witnesses, victims, and ordinary citizens.[29]  They have studied procedural fairness in courthouses, in police precincts, and on police beats.  And they have consistently found that fair treatment affects attitudes toward legal authorities (though fairness may play a greater or lesser role, depending on the particular context and circumstance).[30]  But beyond the most obviously controversial enforcement tactics and adjudicatory procedures, surprisingly little is known about which practices laypersons perceive to be most professional and unprofessional, fair and unfair.[31]  The gap is especially unfortunate because legitimacy has the potential to do its best work at the margins.  More concretely, public perceptions would seem to hold the most sway on otherwise close questions.

One difficulty is that there is no clear consensus on which mechanism or mechanisms lead people to adopt perceptions of legitimacy in the first instance.  In their seminal early study, John Thibaut and Laurens Walker indicated that individuals prefer fair procedures because such procedures are ultimately more likely to produce just outcomes.[32]  In short, a fair procedure fosters a greater likelihood of a just substantive outcome, even if that favorable outcome does not, in fact, come to pass in the immediate case.  By contrast, Tom Tyler has argued that procedural fairness has normative value wholly independent of outcome because legitimate practices convey respect for the individual and thereby promote self-esteem.[33]  More recently, Kees van den Bos, Allen Lind, and others speculated that fair procedures might influence systemic satisfaction by reducing uncertainty.[34]  In short, the question is unsettled and is just one of many ripe areas for continued study.

Other open questions concern what marks the essentials of professional enforcement and fair adjudication, whether some values are more essential than others, and whether perceptions of any such minimum standards are universal across cultures and experiences (and are therefore largely resistant to habituation, education, and attempted manipulation).  It would seem to us that perceptions of procedural fairness are more malleable than intuitions about distributive justice.  After all, we see striking consistency across cultures when it comes to perceptions of the relative severity of different types of misconduct.[35]  By contrast, we see dramatic differences in the procedural norms honored by otherwise similar liberal Western states.[36]

Beyond such rough-and-ready speculation, however, we do not attempt to answer these questions.  For present purposes, we think it enough to review the extant scholarship and highlight certain unexplored or underexplored practices and procedures, many of which are neither obviously legitimate nor illegitimate, but that present potentially fruitful areas for future examination of the question.

1.     Fair Enforcement

Professional policing regulates social behavior through fair procedures and practices.  As indicated, a fair procedure may consist of fair decision making or fair treatment.  Specifically, people are likelier to perceive police decision making as fair when officers make decisions according to readily discernible and generally applicable rules, standards, and guidelines.  Likewise, people are likelier to perceive police treatment as fair when officers behave in manners that are trustworthy, equitable, dignified, and respectful.[37]

Almost certainly, the police lose perceived legitimacy when they intentionally or willfully (or even recklessly or negligently) employ excessive force.  This is a somewhat straightforward issue, and it need not detain us long.  It goes without saying that, for any number of normative and instrumental reasons beyond perceptions of legitimacy, police should refrain from the kinds of abuses of power made infamous by the beating of Rodney King,[38] the killing of Arthur McDuffie,[39] or the sodomizing of Abner Louima.[40]  Instances of extreme and intentional illegitimate behavior, even if isolated, may profoundly impact perceptions of the legitimacy of legal authorities because “people are more strongly influenced by negative experiences than by positive experiences.”[41]  This asymmetry is not terrifically surprising.  Individuals—at least those who start out believing their institutions are legitimate—anticipate positive experiences with authority.  Consequently, their positive perceptions remain somewhat static following fair treatment but may be undercut by even a single instance of unfair treatment.  Moreover, in the digital age, instances of police abuse are likelier to be recorded and broadcast to a wider audience, as demonstrated recently by the negligent homicide of Oscar Grant by a transit officer in Oakland, California—a killing that was captured by six separate cell-phone video cameras.[42]  And, of course, such images of negative treatment are likelier to be disseminated virally than banal images of respectful or pleasant police-citizen encounters.

More difficult questions arise when police engage in run-of-the-mill unprofessional practices—when they are brusque, insensitive, rude, or dishonest.  For instance, police sometimes engage in the illegitimate practice of providing doctored testimony or police reports to justify arrests, searches, or identification and interrogation procedures—the so-called “testilying” phenomenon.[43]  More generally, police may behave impolitely, aggressively, or dismissively in their day-to-day interactions with civilians.  Again, it may be enough to say that—for reasons of human decency and procedural fairness—police should strive to interact civilly with suspects, witnesses, victims, and the general public.  But beyond such platitudes, the existing literature offers some lessons.  First, police may undermine perceptions of legitimacy by showing force rather than soliciting consent politely (even in circumstances where force is justifiable and where consent is not legally required).[44]  Second, isolated incidents of low-level unprofessionalism are likelier to influence individuals’ attitudes toward the police when the contact is citizen-initiated, as opposed to police-initiated.[45]  Third, negative perceptions of unprofessional policing may prove persistent enough to affect, in turn, perceptions of prosecutorialpractices and judicial adjudicative procedures.  Specifically, Jonathan Casper found that “aspects of police treatment . . . spill over onto defendant evaluations of their experience with courtroom personnel and their general sense of fair treatment.”[46]

The thorniest questions involve police practices that are merely controversial—practices that may be considered procedurally unfair from one view, but that also may be defensible from some valid alternative perspective or for some valid alternative reason.  It is especially interesting to consider whether the Supreme Court’s constitutional acceptance or rejection of these borderline police practices aligns with the public’s perceptions of procedural fairness.  By way of example, constitutional criminal procedure questions frequently turn on analyses of expectations, understandings, or beliefs of the so-called “reasonable man.”  Yet, the Court has done almost no work to determine whether its conceptions of the reasonable layperson dovetail with what people actually find fair in a given context.  For instance, to determine whether police have engaged in a search subject to Fourth Amendment inquiry, a court must ask whether police have intruded on a defendant’s reasonable or legitimate expectation of privacy.[47]  To reach this determination, the court must make an evaluative judgment of which activities (and in what contexts) a given society at a given time perceives to be sufficiently private to merit constitutional protection.  But the Court provides no empirical bases for its assertions of what constitutes reasonable expectations of privacy within communities.  The danger is that the average person may find the Court’s folk psychological assessment of the average person’s beliefs to be disingenuously cramped or expansive.[48]  In such circumstances, the public may come to perceive unfairness along any of three dimensions.  First, the public may consider the court to be dishonest, nontransparent, or perhaps even biased.  Second, the public may take the court to be insufficiently sensitive to societal needs for effective law enforcement.  Third, the public may construe the court to be an unaccountable and nonresponsive body that substitutes its own preferences for popularly held value judgments.

To illustrate, the Supreme Court has held that a person has no reasonable expectation of privacy in her closed trash,[49] her bank[50] and telephone records,[51] or her real property as viewed aerially.[52]  Conversely, the Court has held that a person possesses a reasonable expectation of privacy in the heat emanating from her home.[53]  These intuitions are not definitively inaccurate.  For instance, social-science findings indicate that average lay perceptions, in fact, appear to align with the Court’s determinations[54] that police implicate no reasonable expectation of privacy by (1) subjecting luggage to dog sniffs for contraband,[55] (2) tracking vehicles’ movements remotely,[56] or (3) observing private property from hovering helicopters.[57]  But, when the Court relies on intuitions, it runs a significant risk that those intuitions may be wrong.  After all, judicial intuitions are just that—divinations of societal expectations grounded in no more than a hunch; deductions about what, to use Justice Harlan’s phrasing, “society is prepared to recognize as ‘reasonable.’”[58]  Thus, it may be that the public agrees that, say, a person lacks a reasonable expectation of privacy against everyone once she has “knowingly exposed” information to anyone.[59]  But it may well be otherwise.  Indeed, the few social scientists who have studied these questions have found that the public disagrees with the Court on at least some Fourth Amendment search questions.  For instance, one study found that the public perceives the practice of allowing police to inspect bank records to be highly intrusive, notwithstanding judicial tolerance for the practice.[60]

Of course, this concern is not exclusive to constitutional questions of what courts and the public perceive to be reasonable expectations of privacy.  Analogously, in the context of Fourth Amendment consent questions, courts ask whether defendants reasonably feel free to refuse police requests.[61]  In the context of Fourth Amendment seizure questions, courts ask whether defendants reasonably feel free to leave or otherwise end police encounters.[62]  Again and again, in the context of constitutional regulation of police conduct, courts take up questions of societal perspective without the benefit of empirical guidance.  As indicated, a given court’s intuitions may be right.  Indeed, the social science indicates that lay perceptions apparently align with the judicial determination that police may not solicit consent based on false claims of legal authority to search.[63]  But, on other questions, it appears that courts may have it wrong.  For instance, public perceptions are apparently at odds with decisions that, on the one hand, prohibit further interrogation once a suspect asks for counsel[64] and, on the other, authorize undercover agents to trick a suspect into confessing to crime.[65]  In a similar vein, findings suggest that—notwithstanding the Court’s view to the contrary[66]—reasonable people rarely feel free to refuse police requests.  Specifically, one study found that eighty percent of suspects acquiesced to police search requests because they believed that the police would search even without consent.[67]  Another study found that most suspects did not feel free to leave or refuse to answer police questions in the context of street stops and bus interdiction.[68]

As these studies suggest, when courts pay insufficient attention to public perceptions, they may come to make unreasonable claims about the reasonable man.  Courts may endorse ostensible reasonable beliefs that the reasonable public does not, in fact, share—that the public, instead, perceives to be either too deferential to the criminal class or, conversely, insufficiently protective of any citizen (save for the very paranoid).  To the extent judicial intuitions deviate from the lay perspective, courts risk undermining perceptions of legitimacy both by misapplying the relevant standard and by empowering police conduct that the public may find normatively problematic.  Admittedly, courts may be less than perfectly competent to analyze and utilize social science, but good-faith efforts to do so are undoubtedly superior to empty reliance on rank speculation.

The takeaway is not just that courts should take seriously—at least as a factor—lay perceptions of fairness but that social scientists should get serious about the business of measuring those perceptions.  For instance, social scientists could measure lay perspectives toward court decisions empowering police to arrest suspects for even nonjailable offenses (like driving without a seatbelt);[69] to engage in pretextual stops and suspicion-less drug interdiction;[70] to follow a fleeing suspect without triggering constitutional inquiry;[71] and to use evidence of flight to support a finding of reasonable suspicion or probable cause.[72]  In formulating these and other investigatory rules and standards, the Court has appealed to its notions of “common sense,” of “ordinary human experience,”[73] and of the “practical considerations of everyday life”[74] without demonstrating that the public shares its notions.  The public may approve of, say, high-volume, relatively unintrusive police practices, likeTerry stops and frisks, or it may instead favor low-volume (but perhaps more intrusive) practices, like house searches.  The legitimacy project potentially has much to say on these questions.  And, by listening, the system may cultivate deference for its rules and institutions.

Significantly, the Court may not be alone in reaching unfounded empirical conclusions about public perceptions.  A number of scholars have invoked the legitimacy project to advance pet projects and advocate pet reforms.  For instance, for academics and observers troubled by the racially and economically disparate impacts of policing, it is tempting to conclude, without firm empirical bases, that urban communities find unfair aggressive order-maintenance policing and stop-and-frisk practices.[75]  This is not to say that the claim is incorrect.  To the contrary, there are strong indications that heavy-handed and targeted policing does indeed engender public disaffection and thereby may prove counterproductive (particularly in the historically disadvantaged communities that tend to be subject to high levels of enforcement).[76]  But that is not the point.  The point is that we do not definitively know.  And, in such circumstances, academics should resist the temptation to rely too casually on the legitimacy project as a fulcrum to leverage idiosyncratic preferences and conceptions of what constitutes professional policing.  Academics may appropriately offer policy prescriptions based on suggestive data, but they ought to acknowledge that the data is less than clear.

How, then, might an academic appropriately confront concerns about the perceived legitimacy of borderline practices?  The glib one-word answer is “carefully.”  But, of course, we need to say more.  The scholar must keep a focus on the complexities of legitimacy questions.  Practices are multifaceted, and the public may perceive as fair or unfair some aspects or the frequency of a given practice.  Thorough analysis of legitimacy questions, thus, demands thorough understanding of the practice in question.  By way of example, consider the aforementioned practices of order-maintenance policing and attendant stops and frisks—policing practices that have been the subject of particularly sharp criticism from scholars and civil libertarians who believe that such efforts have “reduce[d] the perceived legitimacy of the police in the eyes of the public,”[77] especially where the efforts have been concentrated in predominately poor and minority high-crime neighborhoods.[78]

A brief description of each practice is in order.  First, order-maintenance policing concentrates enforcement efforts on petty public-order offenses.  The strategy is typically based on the broken windows theory, which posits that disorder, if tolerated, may foster an environment of more serious crime.[79]  Numerous urban police departments—most notably, the New York City Police Department (“NYPD”)—have embraced order-maintenance policing over the past two decades, leading to hundreds of thousands of additional arrests for minor crimes.[80]  Second, stop-and-frisk procedures (or so-called Terry stops and frisks) consist of brief detentions and searches based on “reasonable suspicion”—a standard that is less than probable cause.[81]  Specifically, under Terry, an officer has constitutional discretion to stop a suspect and frisk him for a weapon where the officer can articulate specific facts to support a reasonable suspicion that the suspect is armed and has committed, is committing, or is about to commit a crime.[82]  Significantly, stop-and-frisk practices and order-maintenance policing are related, because a department that prioritizes public order will often come to rely heavily on stop and frisk.[83]  For instance, use of Terry stops by the NYPD skyrocketed dramatically from approximately 97,000 such stops in 2002, to more than 160,000 stops in 2003, to nearly 580,000 stops in 2009.[84]

Without question, these policing efforts can be aggressive, but have they, in fact, generated disapproval from the public?  Or is the academic perception of backlash a false academic perception of public perception?  On this central question, there is a remarkable lack of empirical evidence.  In all likelihood, the answer is nuanced.  On the one hand, there appears to be a genuine perception among at least some people in some neighborhoods that stop-and-frisk and order-maintenance policing practices are invidious and thereby unfair.[85]  On the other hand, most people approve of Terry stops and frisks.[86]  Ultimately, there is probably significant variation in perspective within and across populations.[87]  Specifically, even if the general public favors these practices, residents of high-crime neighborhoods would seem to be more conflicted.  These residents internalize directly both the costs and benefits of policing and crime, and they appear to harbor anxieties about each.[88]  In such circumstances, police walk a fine line between impressions of callous disregard and repressive overreach.  And, to effectively chart this course, police could benefit from rigorous empirical study of the levels of enforcement that targeted communities perceive to be too much or not enough.  But, to date, there has been too much shouting and too little study.

Finally, questions about the perceived legitimacy of stop-and-frisk and order-maintenance policing are complicated by questions about the perceived legitimacy of racial profiling—yet another nuanced question.  And a further complication is that the public may fail to differentiate what is and is not racial profiling in the first instance.  Specifically, for present purposes, we take racial profiling to be the practice of using race qua race as a factor in enforcement decisions.[89]  This practice is, therefore, to be distinguished from other types of profiling that merely may correlate to race or ethnicity.[90]  On the one hand, the public seems to overwhelmingly support the use of drug-courier and terrorist profiles.[91]  On the other hand, there appears to be something closer to cross-demographic consensus that racial profiling, as we define it, is illegitimate.[92]  Making matters more complicated still, different demographic groups tend to disagree about whether a particular practice is motivated by race (and therefore an instance of what we take to be genuine racial profiling).[93]  For instance, one study indicated that although whites tended to find racial profiling problematic they were less likely to perceive borderline police conduct to be discriminatory.[94]  Perhaps for that reason, perceptions by whites of normatively problematic police conduct were not shown to influence levels of deference in a statistically significant way, whereas such perceptions were shown to influence deference levels amongst minorities.[95]  Put simply, minorities appear likelier to believe that illegitimate police practices are not just unfair in the general sense, but also biased against them.

And two final observations that further muddy the analyses of racial profiling specifically and order-maintenance policing and stop-and-frisk practices generally: First, there also appear to be generational gaps in perceptions of profiling.[96]  Second, perceptions of profiling—and controversial police practices more generally—are elastic.  At distinct historical moments, tactics like profiling may look different to the discerning public.[97]  Put differently, perceptions of profiling have shifted over time depending on the salient reference point.  For example, we could compare responses to the beating of Rodney King, to the 9/11 terrorist attacks,[98] to the problem of illegal immigration,[99] to the violence associated with the Mexican drug trade.

Ultimately, the extent to which any of these three related policing practices—order-maintenance policing, stop-and-frisk practices, and racial profiling—negatively influence communal perceptions of legitimacy turns on a definition of the relevant community.  In the separate context of empirical desert, we think it makes sense to define the relevant community as the populace covered by a contemplated liability or punishment rule, because the reach of substantive criminal law typically extends all the way to the state’s borders.  But discretionary enforcement practices are more often developed and implemented locally and, accordingly, should remain locally responsive.[100]  Thus, in the context of legitimacy, a narrower definition of community may be warranted, but it is not clear how to go about coherently narrowing community to some subset of the sovereign whole.[101]  Moreover, it is not even clear whether the appropriate measure of community ought to be geographic, cultural, or sociodemographic.  Undoubtedly, there are defensible (or at least workable) definitions, but questions will remain over whose perceptions should matter when opinions are split within a given community.  For instance, we might prioritize the perceptions of those most affected by police practices, those most affected by crime, those constituting a democratic majority, those who are members of historically subordinated minorities, or those who are likelier to defer to legitimate authority (if such groups are identifiable).

Finally, because of the malleability of perceptions of fairness, legitimacy advocates must reconcile themselves to the fact that a system premised even partially on legitimacy may come to adopt procedural rules and standards that may vary from place to place, community to community, and time to time—a scenario that some may find especially problematic when it comes to purportedly nationally applicable standards and rules.[102]  Such variation is not indefensible—and, in fact, may be desirable, even in the constitutional context—but it may require advocates to stake out cognizable positions on seemingly unrelated questions, like the feasibility and appropriateness of theories of localism and popular constitutionalism.[103]  In short, the issues at play are inexorably complex.  But, significantly, they are not insoluble.  Academics and reformers ought to pursue legitimacy-based arguments, but they ought to take care to ensure that their arguments are sufficiently theorized before they may effectively invoke perceptions of legitimacy to resolve controversial public-policy questions.

2.     Fair Adjudication

The legitimacy project could also provide insight into lay perceptions about procedural rules and standards that regulate prosecutors and courts.  For instance, it might be useful to inquire into whether laypersons favor inclusion of an actual-innocence verdict option; whether they are troubled by the lack of a constitutional claim for actual innocence; or whether they perceive as fair such things as standing requirements, exhaustion doctrines, peremptory challenges, speedy-trial rules, double jeopardy, the inability of prosecutors to comment on a defendant’s exercise of her right to remain silent, or the inability of jurors to learn of sentencing consequences pre-verdict.  With respect to all of these procedural rules and standards, social scientists could ask whether there are disconnects between adjudicative practices and public perceptions of fairness, and whether any such rifts undermine perceptions of systemic legitimacy more generally and ultimately deference to law and legal authorities.

One of the most contentious procedural rules is the exclusionary rule.[104]  And, because it is so controversial, it has been the subject of some promising research into its perceived legitimacy.  Perhaps somewhat surprisingly, one study found that an overwhelming majority of respondents disapproved even of using illegally obtained evidence for impeachment purposes (a maneuver that the Supreme Court constitutionally has authorized).[105]  More recently, Kenworthey Bilz found that public perceptions of the rule may be parsed according to the particular rationale offered for its use.[106]  According to Bilz, the public endorses the rule when it is used to promote the integrity of the criminal justice system, but less so when it is used to promote deterrence only—a significant finding that runs counter to the conventional (and almost exclusive) judicial reliance on the deterrence rationale.[107]  Of course, this raises the question of whether the exclusionary rule serves aggregate systemic integrity because conceptually the rule can only advance integrity along one dimension by sacrificing it along another.

Concretely, the exclusionary rule limits the ill-gotten gains of the state by granting an undeserved windfall to factually guilty defendants.  On one reading, such a tradeoff is no affront to systemic integrity; it is the price of integrity—the cost of honoring the presumption of innocence and the corresponding allocation of the burden of proof to the state.  But, significantly, it may not be perceived as such by the public.  The paradox of the exclusionary rule is that the court must tolerate the illegality of one party or another: the state that builds a case unlawfully or the guilty defendant who gets off scot-free.  To determine which is the lesser evil, it would be wise to consider the perceptions of the public.  And, to that end, more empirical study is needed.  For instance, social scientists might inquire whether people consider the exclusionary rule to be procedurally unfair where (as is almost certainly the case) it leads to unequal treatment and under-enforcement for reasons unrelated to desert or factual guilt.  Or they might ask a more nuanced question: whether people believe that crime severity should factor into decisions of whether the exclusionary rule applies—a position the Court has steadfastly rejected.[108]  At a minimum, the research thus far supports the notion that courts should probably give more credence (or at least some lip-service) to the expressive integrity justification for the exclusionary rule, and should perhaps rely less upon the commonly invoked utilitarian deterrence justification that the public apparently feels is comparatively less important.

And the exclusionary rule is not the only procedure designed to both regulate executive actors and to promote integrity.  The same could be said of, say, speedy trial[109] and double jeopardy[110] rules and standards, which also provide undeserved windfalls to at least some guilty defendants.  Public perceptions of these rules and standards may turn on the degree and type of state fault and precisely what bad acts the defendant is alleged to have done.  For example, when it comes to double jeopardy, the public may be likelier to perceive as unfair and unjust intentional prosecutorial efforts to file successive charges that just pass the formulaic Blockburgertest over unintentional prosecutorial slip-ups that just fail it.[111]  Likewise, when it comes to speedy trial, the public may disapprove of exploitative shorter delays over inadvertent or even negligent longer ones.[112]  Or it may be that, in each instance, the public perceives fault to be relatively unimportant to the question of whether the trial delay or successive prosecution is procedurally just.  Again, it may depend, to a degree, on whether public perceptions of fairness turn principally on sanctioning (and thereby deterring) abusive exercises of executive power or on promoting the integrity of the criminal justice system more generally.  In short, there is a lot to unpack and much room for further study.

Comparatively, researchers have done substantially more work on public perceptions of bargained dispositions, the dominant adjudicatory practice in American criminal justice.[113]  These studies reveal that most Americans disapprove of plea bargaining—as many as four-fifths, according to some studies.[114]  However, it is unclear whether these negative public perceptions of the practice are products of procedural or substantive objections—that is, whether the public disapproves of the lack of trials and attendant formal process and transparency (legitimacy objections) or of the perceived lenient, harsh, or unequal outcomes (distributive-justice objections).[115]  By contrast, practitioners overwhelmingly and predictably approve of plea bargaining as a necessary tool to efficiently manage high caseloads.[116]  Somewhat more surprising are studies that have found that defendants may approve of bargaining and may derive similar, or even greater, satisfaction following pleas than trials.[117]  The reasons are unclear.  It may be that defendants appreciate the ability to exercise some dominion over their own fates—a kind of “process control” that “may foster a greater sense of participation.”[118]  It may be that the very informality of the plea process makes it more comprehensible to defendants than professionally dominated and highly technical rule-bound trials.[119]  It may be that defendants appreciate the certainty of guilty pleas or (like practitioners) the efficiency of avoiding the process costs of trials.[120]  Or it may be that defendants are more satisfied with sentencing outcomes after pleas than trials.

Plainly, there is much more to explore concerning lay perceptions of plea bargaining and whether and to what degree those perceptions are shaped by procedural practices, as opposed to substantive results.  Additionally, there are a number of corollary or subsidiary bargaining practices that have received almost no attention at all.  For instance, it is unclear what laypersons think of arguably strained judicial constructions of “voluntariness” in the plea-bargaining context.[121]  Likewise, it is unclear what laypersons think of so-called Alford pleas—equivocal pleas in which defendants accept guilt while protesting innocence.[122]  And it is also unclear what laypersons think of other nontraditional pleas, like cooperation agreements.[123]  Under cooperation agreements, cooperating defendants are ultimately punished not according to desert, but according to the crime-control value of the help that they provide to legal authorities.[124]  Indeed, some of the most culpable defendants may receive the most significant discounts because they have the most information to sell.  These utilitarian bargains may serve instrumental goals,[125] but only at retributive costs that the public may perceive to be unfair and unjust.[126]  Finally, it is unclear what laypersons think of reforms intended to provide victims with more significant roles in plea bargaining and adjudication generally.[127]

B.     Creating Moral Credibility

Studies confirm that laypeople think of criminal liability and punishment in terms of desert—the moral blameworthiness of the offender—and not in terms of other principles, such as general deterrence and incapacitation, which have been so popular with system designers during the past several decades.[128]  Thus, people naturally expect that a criminal justice system will distribute criminal liability and punishment so as to do justice.

However, studies have shown that current liability and punishment rules commonly undermine the criminal law’s reputation for doing justice.  One recent study showed that a wide range of modern crime-control doctrines treat cases in ways that dramatically conflict with laypeople’s intuitions of justice.[129]  The conflict exists for such standard doctrines as “three strikes” and other habitual offender statutes, high penalties for drug offenses, adult prosecution of juveniles, abolition or narrowing of the insanity defense, strict liability, felony murder, and criminalization of regulatory violations.  The conflicts were shown to undermine the criminal law’s moral credibility with the subjects.[130]  Previous studies had results consistent with those results.[131]

It appears, then, that to build moral credibility, the criminal law must avoid conflict with the community’s principles of justice.  While there is still much work to be done, current research tells us something about community views on a wide variety of criminal law issues, including studies on: objective requirements of attempt; creating a criminal risk; objective requirements of complicity; omission liability; use of deadly force in self-defense; use of force in defense of property; citizens’ law enforcement authority; mistake or accident defenses; culpability requirements generally; culpability requirements for complicity; voluntary intoxication; individualization of the objective standard of negligence, insanity, immaturity, and involuntary intoxication; duress and entrapment defenses; felony murder; causation requirements; and punishment for multiple offenses.[132]  Other studies have examined lay intuitions on whether guilt should be determined according to objectivist or subjectivist views of criminality,[133] on competing theories of blackmail,[134] on offense grading distinctions,[135] and on competing theories of justification.[136]

The studies make clear that current criminal law regularly deviates from the community’s justice judgments.  And with each instance, the law risks undermining its moral credibility with the community.  While many of the deviations are sufficiently minor to have little impact on their own, the cumulative effect of the many deviations can have a substantial practical impact, as Part III discusses.  Consider this summary of studies, as reported by Robinson and Darley.[137]

The studies report many sources of law-community conflict in relation to the criminal law’s secondary prohibitions, such as inchoate offenses, omission liability, and complicity liability.[138]  Modern American criminal codes commonly would impose liability where laypersons would not, or would impose considerably less.  Consider a variety of examples of such conflicts.  Subjects found the dangerous proximity test embodied in the common law a more proper test for attempt liability than the substantial step test of the Model Penal Code (“Code”), upon which most modern American criminal codes are based.[139]  Subjects gave more weight to renunciation of a criminal attempt even when that attempt had progressed far enough to trigger liability.[140]  The respondents believed that unsuccessful attempts to assist in a crime called for little or no liability,[141] which is more consistent with the common law rule than the Code’s treatment, which imposes full offense liability.[142]  Subjects gave a person who intends murder, but creates only a slight risk of causing a slight harm, a sentence greatly reduced from the Code’s treatment of the act as attempted murder.[143]  Subjects did not give an accomplice to crime the level of liability equal to that they assigned to the principal, contrary to the Code’s approach.[144]  Finally, subjects found a person who failed in his or her duty to rescue a person from death to be somewhat culpable but not for murder, again in conflict with the Code.[145]

The conflicts moved in the other direction as well, with respondents imposing greater liability than that imposed by modern codes, or assigning liability in cases in which the codes do not.  For example, subjects assigned liability to individuals who develop a settled intention to commit a crime, even though modern codes assign none unless that intention is translated into action.[146]  Likewise, respondents imposed some liability on persons who failed to assist a person in distress, although modern American codes typically do not.[147]

These points of conflict are really just symptoms of conflicts in broad perspective between laypersons and the Code.  Note that respondents consistently assigned reduced liabilities to individuals who play secondary roles in the commission of a crime, even though the Code assigns them liabilities equivalent to the principal.  Also, subjects assigned reduced liability for the violation of secondary prohibitions (such as attempt), even though the Code assigns liabilities equivalent to those assigned for the substantive offense.  That is, the results suggest that the community is more “harm-oriented” than is the Code, at least in grading the seriousness of wrongdoing.[148]

The criminal law’s justification defenses show a variety of points of conflict with community views.[149]  The subjects considered the justifications to be more compelling than the modern legal codes in a variety of instances.  When force is used in self-defense, in defense of property, or to apprehend a fleeing offender, respondents frequently assigned no liability in cases to which the Code did.[150]  Even when respondents assigned liability, they typically would grade the violation as considerably lower than would the Code.[151]  The differences between community standards and criminal codes becomes apparent when studies examine what people think is justifiable in defense of property, and even more apparent in the cases involving a citizen who uses force to apprehend a criminal fleeing a crime or to apprehend a person thought to be a criminal fleeing a crime.  The respondents were willing to tolerate the use of more force than the Code permits, and assigned lesser sanctions to defendants for the use of force even in the instances in which they felt that some blame ought to accrue.[152]

Studies of the culpability requirements for offenses revealed several points of conflict between community views and criminal law.[153]  For example, there is a general approval of the law’s tendency to make a major differentiation between reckless and negligent commission of an offense.  Subjects imposed significantly different liability assignments depending on whether the individual was reckless with respect to the various elements specified as relevant by modern codes, or merely negligent.[154]  Most codes assign no higher liability for an offense (other than homicide) committed knowingly or purposefully than one committed recklessly.  Subjects, in contrast, assigned higher liability for higher culpability than recklessness.[155]  Subjects also distinguished recklessly committed offenses from negligently committed ones, but unlike the general code treatment, sometimes assigned significant liability even for negligently committed offenses.[156]  In general, to accurately reflect communal judgments, drafters should redraft codes to have liability vary with the culpability level of the offender.

A similar result emerged when studies examined the culpability requirements for complicity.  The codes set purposefulness as to assisting the principal as the minimum requirement for complicity, but respondents were willing to assign liability to a person who knowingly, or even recklessly, assists.  Respondents assigned different and lower levels of liability as the culpability level decreased, which again suggests that the community would prefer differences in grading based upon differences in culpability.[157]

The voluntary intoxication studies found a pattern of judgments that was broadly consistent with the legal treatment of the cases.  Specifically, “codes commonly use negligence as the trigger point for the attribution of liability, and so do subjects, thus supporting the codes’ adoption of that standard rather than one triggered by a higher degree of culpability.”[158]  However, respondents were considerably influenced by the degree of pre-intoxication culpability that the person had with respect to commission of the offense, while the codes typically do not consider that factor.[159]

A person is negligent if he or she disregards possible risks that a reasonable or prudent individual would consider.  “Traditionally codes have treated this as an objective standard, not to be varied as a function of, for instance, the lower intelligence of the person whose conduct is being judged.”[160]  However, some modern codes have partially individualized the standard, directing the decision maker to take account of some personal characteristics of the offender.[161]  There is no developed theory of which factors should be taken into account; “[i]nstead, judges (and, to a lesser extent, juries) are allowed to determine which attributes should be considered.”[162]  A study seeking to discover which attributes subjects considered to be appropriate for individualization of the reasonable person standard gave complex results: it found both lowered and raised standards.[163]  This suggests that the modern trend toward individualizing the objective reasonable person standard has support among the community.  But the absence of an obvious principle tends to confirm the practical need for some greater guidance on the issue.  Leaving judges to decide ad hoc, as is now the case, is not likely to generate results consistent with community views.

The studies show that systematic conflict with the community also exists within doctrines of excuse.[164]  While the subjects recognized mental incapacity as a valid basis for exculpation, they preferred formulations of the insanity defense that recognize both a control and cognitive deficit; respondents seem to conclude that dysfunction of either type is a valid trigger for exculpation.^[165]  Yet a majority of state codes recognize cognitive dysfunction but not control dysfunction as a basis for the defense.[166]

The studies recognized duress scenarios as providing at least a mitigation of liability; the degree of mitigation is a function of the respondent’s perception of the degree of coercion in the particular situation.[167]  Codes instead set a single cut-off standard in which the offender either gets a full defense or full liability.  Respondents used similar considerations to evaluate entrapment cases, rejecting as unimportant the Code’s formulation of the defense, which requires that it be given only when a police agent supplies the inducement that leads the person to commit the crime.[168]  This suggests that they would be happy to have the entrapment defense disappear as a separate defense, and be subsumed under the duress defense.[169]

Studies found that respondents’ intuitions about the appropriate grading for different variants of offenses differ from those reflected in the modern legal codes.[170]  For example, respondents regarded forcible rapes as similar in grade irrespective of the parties’ prior relationship, while the Code varies the offense grade by prior relationship.[171]  With regard to felony murder, respondents preferred what might be termed a “felony-manslaughter rule” (with a standard “accomplice discount”) rather than the modern approach that extends murder liability to all participants.[172]

It is clear that respondents agreed with the general tendency in modern legal codes to distinguish grades of offenses within as well as between offenses, and typically favored additional grading distinctions not made by the Code.  For example, in causation studies, subjects graded various causal contributions to a death differently, assigning less liability to persons who intended to kill but did not succeed in directly doing so, although death later occurs in some more indirect way.[173]  The subjects’ interest in more specific grading also is similar to their conflict with the tendency of modern codes toward dichotomous judgments as compared to the respondents’ more continuous judgments.  As noted above, they see the degree of contribution in complicity, or how far the offender has progressed in an attempt, or the strength of the causal connection in homicide cases as grounds for adjusting liability along a continuum, as opposed to the law’s judgment to set a fixed point on the continuum that will judge the conduct to be either complicity or not, an attempt or not, full causal accountability or none.

The problems of sentencing for multiple offenses also suggest a structural change in current law.  Presently, multiple offenses typically are dealt with either by concurrent sentences, which impose no additional sentence for a second offense, or by a consecutive sentence, which effectively doubles the penalty.  Respondents, in contrast, assigned sentences for a second offense that added to but did not double the sentence for the first offense, and continued this pattern for further offenses.  Their approach matches that of the United States Sentencing Commission guidelines.[174]  If the law were to better reflect community views, that approach would be more widely adopted.

C.     The Problem of Perception

Perceptions mediate the normative force of procedural fairness and empirical desert, but perceptions, of course, may be wrong.  By way of historical example, trials by ordeal and other irrational modes of adjudication were probably considered legitimate during the early Middle Ages, even though they almost certainly were not.[175]

In other words, the legitimacy project for its part does not actually demand that procedures be fair, only that they appear to be.  And the moral credibility project for its part does not actually require that substantive rules produce just results, in a transcendent sense, only that they reflect people’s shared moral intuitions.  An element of relativism thereby creeps in: there may be moral truth about the distributive justice of a given rule or the fairness of a given procedure, but popular perceptions of fairness and justice may be otherwise.[176]  In this way, the emphasis on perception raises three potential problems.  First, a fair procedure or just rule may be misconstrued as unfair or unjust (false negatives).  Second, an unfair procedure or unjust rule may be misconstrued as fair or just (false positives).  Finally, a questionable but nontransparent procedure or rule may not be perceived at all.

To some extent, these concerns may be beside the point for present purposes, as our principal focus is whether, when, and how perceptions of fairness and justice facilitate effective crime control—that is, our focus is on the implications of perceptions, as opposed to their truth.  In any event, one of us is highly skeptical whether deontological facts are operationalizable.[177]  Assuming there are moral absolutes (a big assumption), we believe that tapping lay perceptions is the best (albeit imperfect) way to come closest to discovering them.  Nevertheless, we must acknowledge the potential for lay perceptions to fall short of the mark.  This is a particular concern when it comes to perceptions of procedural fairness because they are more likely to be malleable cultural constructs and are therefore less likely to accurately reflect some higher truth.[178]

In an effort to not sell short the objections, we think it appropriate to say a bit more about each potential problem of perception—that is, false negatives, false positives, and nontransparency.  First, as to false negatives, we concede the inevitability of the problem.  Even a credible and just government may commit some salient misstep, and such a blunder may trigger a pernicious spillover effect that leads citizens to misperceive as unfair or unjust even normatively defensible governmental actions, standards, or rules.[179]  These misperceptions could thereafter lead to a loss of deference—for instance, the nullification of a justifiable prosecution or the violation of a justifiable law.  Some scholars have posited that this is precisely what happened domestically during the Vietnam conflict: frustrations with the government’s foreign policies contributed to the counterculture’s perception that the nation’s drug laws were similarly unjust.[180]

Second, the reverse—false positives—may also exist.  Under some circumstances, the most problematic forms of racial profiling may implicate the danger.  Specifically, consider the example of an officer who stops and searches an individual after drawing unwarranted inferences of culpability based solely on skin color.  Imagine further that the officer employs a soft tone and exhibits a gentle demeanor.  For obvious reasons, courteousness is commendable, but the fear is that the officer may use polite words and kind manner to manipulate the subtler aspects of police-citizen interactions to create a false perception that she is not engaged in racial profiling, even when she is.[181]  In short, one set of legitimate practices may mask others that potentially are illegitimate.  Put differently, police may play professional and unprofessional practices off of one another.  As Tyler and Wakslak found:

[P]olice behavior shapes the attributions people make . . . .  Those who believe that the police are neutral are less likely to feel profiled . . . .  Those who experience high quality interpersonal treatment—politeness, respect, acknowledgment of their rights—are also less likely to feel that they have been profiled.[182]

This tendency of people to confuse respectful policing for unbiased policing may be good for police and even for public safety, particularly if profiling is a somewhat intractable product of unconscious motivation, but it is not categorically good to the degree we are committed to more than just normative crime control—to the degree we care also about the fairness and equity of police procedures in reality.

This is not to say that police should succumb to rude or abrasive impulses.  It is desirable for police to be courteous, but not when civility is window dressing; it is desirable for police to give reasons for actions, but not when reasons are pretexts.  It is, therefore, somewhat chilling that certain legitimacy advocates endorse steps like “mitigat[ing] by courteous behavior and an explanation” stops of people who “have done nothing wrong.”[183]  Such advice amounts to the perceptual tail wagging the dog.  If police stop people for doing nothing wrong, we should look for feasible ways to minimize that practice, not to merely change public perceptions of it.  We should want normatively troubling police conduct exposed, not hidden.  Moreover, in this context, false perceptions of procedural fairness may prove particularly harmful because perceptions of fairness (unlike moral intuitions) would seem to be more often culturally constructed and accordingly potentially polarizing.  Specifically, a social group that has grown accustomed to inordinate police stops may perceive racial profiling where another group perceives polite policing.[184]  No matter who is right in a given case, the consequences are divisive.

But what if the public is aware of racial profiling yet approves of it?  As indicated, we think (but do not know definitively) that this is not the case.  We believe that the public shares the Court’s perspective that racial discrimination demands a special justification—what the Court refers to as strict scrutiny.[185]  Indeed, we suspect that it is precisely for this reason that police are so unwilling to acknowledge racial profiling.  That is, police recognize that the public typically perceives the practice to be illegitimate, even if it is sometimes (or even typically) instrumentally effective.  But, as indicated, perceptions are elastic and may change over time.[186]  For instance, in the wake of, say, future terrorist attacks or immigration crises, a majority may come to more firmly tolerate or even welcome profiling.[187]  In such circumstances, popular considerations of systemic legitimacy may fail adequately to constrain excessive state power.  On this reading, approval of racial profiling could describe a false positive in the face of more pressing antisubordination and countermajoritarian principles.  In this way, genuine invidious practices may illustrate the outer-bounds of normative theories that permit popular intuition to shape criminal justice policy.

There are examples of adjudicative false positives as well.  For instance, a defense attorney may successfully push an ill-advised and self-serving plea deal on her client simply by communicating persuasively and politely and thereby cloaking self-dealing in a professional facade.[188]  By way of further example, several academics have recommended reforms (which we endorse) intended to grant crime victims greater opportunity to be heard.[189]  But there are sound reasons—most notably, efficiency and equal treatment[190]—to provide victims with only a “voice” and not with guarantees of “decision control.”[191]  From the standpoint of the legitimacy project, this is not a problem: “voice” procedures enhance perceptions of legitimacy, irrespective of whether the voices are heeded.[192]  Concretely, “process control” is all that matters—that is, the ability to shape the manner in which information is conveyed to the decision maker.[193]  But when an adjudicative procedure has no potential to affect outcome, the perception that it is fair may be chimerical.  The argument here is that such a procedure does little more than exploit the cognitive biases that lead people to believe that merely by participating they can affect uncontrollable events.[194]  Of course, a ready response to this is that victim impact statements and similar reforms do, in fact, have at least the capacity (albeit not the authority) to affect adjudicatory decisions.  But we should at least be cognizant of the concern that a preference for “process control” may create opportunities for legal authorities to package procedures and practices in perceptually appealing ways, even if the packages are empty in fact.[195]  Put simply, seemingly fair and just procedures may just serve institutional ends at the expense of fairness and justice.

Another false positive may be found in the administration of juvenile justice in the years before the Court’s influential decision in In re Gault, which extended many constitutional procedural rights to respondents in delinquency hearings.[196]  Significantly, the Court noted that juvenile courts had come to adopt arbitrary and unfair procedures by propagating a hollow perception of the proceeding as “one in which a fatherly judge touched the heart and conscience of the erring youth by talking over his problems, by paternal advice and admonition.”[197]

At a macro-level, the greatest risk of false positives undoubtedly arises out of corrupt governments that are erroneously perceived to be fair and just.  An unjust society will perceive its unjust procedures and rules to be just—even where these procedures and rules merely reproduce that society’s backward worldview.  Put differently, bankrupt intuitions may do nothing to check bankrupt institutions.  To the extent that people buy into the illogic and injustice of, say, a fascist or racist system—such as Nazi Germany or the Jim Crow South—levels of deference remain unaffected (at least among the unsubordinated classes).

Do the dangers of false negatives and positives mean that we should abandon empirical desert and legitimacy?  Not at all.  In the first instance, people’s intuitions are probably right far more often than they are wrong.  In any event, the fact that perceptions may be wrong weighs in favor of pursuing both projects in an effort to increase the likelihood that they will not only foster normative crime control but that they will check each other’s work.  Ideally, citizens may come to reject a corrupt regime’s procedures as illegitimate, even if they erroneously perceive its liability and punishment rules to be just.  Or they may come to reject its rules as unjust even if they erroneously perceive its means to be fair.  This underscores an intriguing fact: perfect deference may be undesirable because it is through disobedience and not deference that the legitimacy and moral credibility of the state are tested.[198]  Instances of lawlessness force the unjust state to advertise its immorality and injustice, and they provide the just state with opportunities to demonstrate the normative worth of its rules and institutions.

Third, and finally, legitimacy and moral credibility have no normative force if the public fails to even perceive the rule or procedure in question.  This is less of a concern with liability and punishment rules than with policing and adjudicatory procedures because substantive rules tend to be more transparent than procedures.[199]  And this is less of a concern with policing than with adjudication because people have more contact with police than courts.  Still, there are concerns.  For example, we have raised a number of potentially illegitimate, low-visibility procedures, standards, and rules—statutes of limitations, the exclusionary rule, standing requirements, speedy-trial rules, exhaustion doctrines, peremptory challenges, and Alford pleas.  But, ultimately, any arguments about the normative authority of empirical desert and procedural fairness with respect to these procedures, standards, and rules must be tempered by the fact that the general public may not even be aware that they exist.

III.  Producing Deference

Certainly, fairness and justice are worthwhile independent of their consequences, but these values are all the more desirable if they also facilitate effective crime control.  Does a system’s reputation for being fair and doing justice make the system more successful at fighting crime?  What evidence exists to support claims of such positive practical effects?

A.     The Effects of Undermining Legitimacy

The advantages of producing deference are obvious, and legitimacy is a particularly appealing mechanism to achieve the objective because perceptions of legitimacy possess the potential to produce deference broadly.  A populace that finds its legal authorities and institutions to be legitimate ought to be likelier to acquiesce to even those rules and enforcement measures that fail to generate categorical (or even very much) support.

The bulk of studies drawing this link between perceptions of legitimacy and deference have examined the question through the lens of police practices.  For example, Tyler has found that racial profiling affects compliance among minorities, but not among whites, who are less likely to perceive the practice.[200]  Another study found that even garden-variety police unprofessionalism may affect deference.  Specifically, John McCluskey found that police are more likely to generate compliance when they interact with individuals courteously and patiently.[201]  Likewise, McCulskey linked forceful entry tactics with noncompliance and noncooperation—an important finding that calls into question the efficacy of hard-nosed police tactics such as “blue swarming,” in which police use strength in numbers to compel compliance, rather than politely soliciting cooperation through even-toned requests.[202]

These and other studies provide robust support for the null hypothesis that perceptions of legitimacy do, in fact, correlate with higher levels of cooperation and lower rates of recidivism.[203]  But questions of causation remain open, and confounding variables may exist.  For instance, in Tyler’s seminal study, subjects were asked to self-report both their perceptions of the criminal justice system and their commission of petty offenses.[204]  From the results, Tyler inferred a link between perception and compliance.[205]  The inference is no doubt strong, but, of course, it could be that law-abiding respondents are likelier to perceive the criminal justice system to be fair.  Or it could be that the respondents with less faith in the criminal justice system were more willing to self-report violations, but not necessarily likelier to commit them.  Tyler mitigated these concerns by using a longitudinal study design that measures deference at two points (as compared to a cross-sectional study that would have provided only a snapshot in time), but the concerns, nonetheless, remain.[206]

A separate study similarly found that individuals were less likely to recidivate if they perceived their arrests to be legitimate.[207]  But, as stated before, it could be that people who perceive arrests to be legitimate are also people for whom misconduct is aberrational.  Again, the causal arrows may flow the other way: subjective perceptions of legitimacy may be produced by moral commitment to law and legal authorities—not vice versa.[208]  Or it could be that police officers act differently toward agreeable suspects (those who exhibit decency, contrition, or congeniality), and these suspects thereafter perceive their interactions with police to be legitimate.  If these internal attributes correlate with deference to law and legal authorities, then fair treatment may be incidental to recidivism rates.

Additionally, some commentators have highlighted high rates of acquittal in certain poor and minority communities—the so-called “Bronx Jury” phenomenon—as evidence that negative systemic perceptions translate to diminished willingness to cooperate with legal authorities.[209]  But higher acquittal rates do not necessarily represent lawless jury nullification; they may simply reflect good-faith application of a legally defensible (albeit distinct) culturally constructed conception of what constitutes proof beyond a reasonable doubt.[210]  And even if historically disadvantaged groups do, in fact, nullify at higher rates, uncooperative juror conduct may not be motivated by perceptions of procedural unfairness; it could be a product of perceptions of substantive injustice or something else entirely.[211]

Finally, as indicated, with respect to order-maintenance policing, academics claim (with fair intuitive support) that heavy-handed police tactics have engendered backlash in some predominately poor and minority neighborhoods.[212]  Yet crime rates have dropped dramatically during the same period, including (and often especially) in many of these same neighborhoods.  It could be, then, that perceptions of illegitimacy have had little or no effect.[213]  Or it could be that crime declines would have been substantially greater had it not been for perceptions of procedural injustice.  Or it could be that, as an initial matter, order-maintenance policing is not in fact perceived to be illegitimate or that it has had no discernible effect on the crime decline one way or another.[214]  We simply do not have certain answers to these questions.

Perhaps most importantly, many legitimacy studies fail to get to the specific question of deference in the first instance.  Instead, they measure only satisfaction with the law or legal authorities and extrapolate from there.[215]  The inductive leap from satisfaction to deference is strong, but it is still a leap.  Perception is one transitive step removed from compliance with law and cooperation with legal authorities.  Other studies fail to parse out what precisely is meant by deference.  Specifically, perceptions of legitimacy may influence compliance with substantive law, or they may facilitate cooperation with enforcement and adjudication procedures, or they may do both, or they may do neither.  And a lack of cooperation itself may be sliced into weak or strong versions: apathy or outright defiance.  The better studies consider the question of deference directly, but even some of these studies base their findings solely on self-reports of how respondents think they would respond to hypothetical unfair treatment.[216]

Ultimately, then, for all the research indicating that procedurally just treatment enhances systemic approval, there is less to indicate that legitimate procedures do, in fact, influence rates of compliance with substantive legal rules and cooperation with legal authorities.[217]  But we should not be too cynical.  Claims of deference may be nonfalsifiable, but in settings where public-policy choices must be made, policy makers may be wise to assume that perceptions of legitimacy do have a sizable effect on rates of compliance and cooperation (at least in some settings).  If nothing else, Tyler’s central premise is appealing and intuitively sound as a matter of pure analytic reasoning: “If people view compliance with the law as appropriate because of their attitudes about how they should behave, they will voluntarily assume the obligation to follow legal rules.  They will feel personally committed to obeying the law, irrespective of whether they risk punishment for breaking the law.”[218]

B.     The Effects of Undermining Moral Credibility

What evidence exists that differences in the criminal law’s moral credibility with the community will track differences in the effectiveness of the criminal justice system?  The most obvious hint may come from common sense.  If citizens see the criminal justice system as unjust, what motivation would they have to assist it—volunteering as a witness, assisting investigations and enforcement—or to defer to it—following its rules even when prosecution is unlikely, following the jury instructions given, joining in the stigmatization of conduct labeled criminal even when its condemnability is unclear, or internalizing the societal norms that criminal law embodies?  Common sense tells us that people are more likely to resist and subvert a criminal justice system that they see as unjust than they are to assist and defer to it.[219]

Anecdotal evidence supports this commonsense view.  The notoriously unjust Soviet criminal justice system, ruled by coercion and threat and tainted by politics and ideology, earned little moral credibility (or legitimacy) with the public.[220]  As Peter Solomon concludes, “Soviet experience demonstrates that indiscriminate and coercive use of the criminal law approaching naked repression discredits both the law and the regime that sponsors it.”[221]  Russian crime rates increased under Soviet rule.[222]  Further, when overwhelming state control was lost with the collapse of the Soviet Union, crime spiked.[223]  Having destroyed its moral credibility with the community, the system had little ability to control conduct through the forces of social or normative influence.[224]

What recent research has shown is that this effect is not limited to instances of dramatically bad reputations, as in the Soviet Union.  Even minor changes in moral credibility incrementally affect people’s willingness to acquiesce, assist, and defer to the criminal law.  One technique used in social science research on such issues is an experiment in which subjects are told of injustices in the current criminal justice system that they did not previously know about, and are then tested to see whether the new information changes their view of the system and their willingness to assist and defer to it.  Obviously the subjects came to the experiment with decades of prior information about the system—its operation is a common feature of news reports—and they no doubt had a pre-existing view as to the system’s moral credibility.  Within this context, there is a limit to how much an experimenter can change a subject’s view in the brief period of the experiment.  Yet, the studies show that even incremental changes in perceptions of justness can and do incrementally change people’s willingness to assist and to defer.

In the most recent and direct study,[225] researchers exposed adult subjects to a range of real-world cases relying upon modern crime-control doctrines that produced results inconsistent with the subjects’ notions of justice.  Subjects’ willingness to assist and to defer when faced with a number of specific kinds of situations was tested before and after this exposure.[226]  Subjects were asked, for example, whether they would report various offenses by other persons; turn in evidence to the police; report their own accidental violation; conclude that the prohibition of certain conduct by the law meant that the conduct was in fact morally condemnable; or conclude that the law’s imposition of a serious penalty meant that the conduct at issue was in fact morally condemnable.[227]  In a separate study, the same researchers tested the same willingness of subjects to assist and to defer by using two random groups: one that had been exposed to the unjust cases and one that had not.[228]  Both studies showed similar results: subjects exposed to the unjust cases were less willing to assist and to defer to the criminal law.[229]

Another study tested for the same effects using pre-existing data from large population surveys collected by others.  Existing national databases were examined to see what, if any, relation the system’s moral credibility had on the effective operation of the criminal justice system.[230]  One large national survey database involved telephone interviews with Americans who recently participated in criminal court proceedings.[231]  Bivariate and multivariate statistical techniques were used to determine whether their responses to the survey questions were consistent with a view that the system’s moral credibility increased their willingness to defer to the courts to resolve a similar case in the future.[232]  The findings support the study conclusions reported above, which is significant because it involves people who have actual experience with the criminal courts.  Respondents with criminal court experience who viewed their community courts as morally credible in dealing with criminal cases (specifically those involving violence, drugs or alcohol, and delinquency) expressed a greater willingness to defer to the criminal justice system in the future.[233]  Subjects who perceived failures in the criminal justice system were significantly less likely to say they would defer to the system in the future.[234]

A number of previous studies, described below, confirm related points.  Some show the existence of a relationship between an individual’s disbelief in the morality of a particular law and his or her willingness to obey that law.  Other studies show that the degree to which people report that they have obeyed a law in the past and plan to obey it in the future correlates with the degree to which they judge that law to be morally valid.  Still other studies go further to show how perceptions of injustice might lead to more generalized flouting of the law.

A number of studies have focused on how beliefs about the morality of a particular law can affect compliance with it.  In one such study, Herbert Jacob showed a greater relation between compliance and a law’s perceived moral correctness than between compliance and perceived likelihood of punishment for violating it.[235]  He concluded that “[t]he relationship between compliance and legitimacy [perceived moral correctness] appears to be considerably stronger than the one between compliance and perceptions of severity or certainty of sanctions.”[236]  In another study, Grasmick and Green found similarly that “three independent variables—moral commitment, perceived threat of legal punishment and threat of social disapproval—appear to constitute a concise and probably exhaustive set of factors which inhibit illegal behavior.”[237]  Likewise, Silberman’s study of 147 undergraduates at a small private university, suggested that “[w]hen public sentiment in general disapproves [of a] given offense, it is relatively unlikely to occur.  Similarly, serious criminal activity is less likely to occur among those who show a high degree of moral commitment, even though these individuals might commit less serious offenses.”[238]

These studies demonstrate that perceptions of the moral correctness of particular laws can affect compliance with them.  Other studies have gone further to show how perceptions of the immorality of a particular law or of some act of the criminal justice system can lead to more generalized effects on compliance.  One, in particular, is a recent investigation by Janice Nadler that looked at how knowledge of injustices by the criminal justice system can affect intentions to comply with the law.[239]  Nadler found that subjects exposed to cases that they viewed as unjust were more likely in a subsequent mock trial to engage in juror nullification.[240]

A similar study by Greene presented unjust cases and also examined their effect on subjects’ attitudes.[241]  Greene reached conclusions similar to Nadler’s.  Subjects who had read cases in which the legal system behaved in ways counter to their moral intuitions rated themselves “more likely to take steps aimed at changing the law, less likely to cooperate with police, more likely to join a vigilante or watch group, and less likely to use the law to guide behavior.”[242]  He further concluded that “overall, subjects appeared less likely to give the law the benefit of any doubt after reading cases where the law was at odds with their intuitions.”[243]

A more recent study by Mullen and Nadler showed how the perception of moral injustice in the legal system can increase rates of deviant behavior.[244]  The researchers found that exposure to outcomes that are inconsistent with a person’s strongly held moral beliefs increases the likelihood of him or her engaging in deviant behavior.[245]

Finally, people’s common compliance with tax laws raises interesting issues related to these points.  Large numbers of American citizens pay their taxes even though the penalty for tax evasion is not great and the probability of detection is trivial, thus, the expected cost of such a crime is quite small.[246]  For these reasons, many legal scholars believe that the threat of official sanction does not explain why such large numbers of citizens pay taxes.[247]  A survey by Karyl Kinsey sheds some light on the underlying forces.[248]  When people reported that a friend or coworker, after contact with the IRS, had been made to pay more taxes than they properly owed, the people thought the tax laws generally were less fair and were more likely to intend to cheat on their taxes in the future.  Nadler, in reviewing the study, commented that “[t]he results of the tax study suggest that exposure to reports of an unjust legal outcome in a particular situation might lead to lower perceived fairness of the law more generally, which in turn can lead to noncompliance with the law in the future.”[249]

Taken together, these studies suggest that knowledge of systematic injustice produced by the criminal justice system, particularly when it is intentional, can have a range of deleterious effects on people’s attitudes and on their behavior.  People are less likely to comply with laws they perceive to be unjust.  They may also be less likely to comply with the law generally when they perceive the criminal justice system as tolerating such injustice.  The studies also show that these effects are not limited to noncompliance, but apply generally to undermine cooperation and assistance with the legal system.  Further, perceptions of injustice undermining the system’s moral credibility can also affect its ability to harness the normative forces of social influence and the internalization of norms.  The flip side, of course, is that if the criminal justice system reflects ordinary perceptions of justice, it can take advantage of a range of psychological mechanisms that serve to increase assistance, cooperation, compliance, and deference.

IV.  Countervailing Interests: The Attraction of Unfairness and Injustice

Given the practical benefits that flow from building a reputation for fairness and justice, why has the current criminal justice system adopted practices, procedures, and rules that conflict with these values?  Sometimes the reason is simply a failure to appreciate the detrimental effects, such as those examined in the preceding Part.  Alternatively, those who shape the rules and practices simply may not realize just how unfair or unjust they are perceived to be.  After all, there may be a disconnect between lay and professional perspectives, because the viewpoints of police, prosecutors, and judges are shaded and shaped by their professional training and experience.[250]  Put differently, what technocrats perceive to be fair and just is not necessarily what laypersons perceive to be fair and just.  Criminal-justice functionaries may simply be too institutionalized to tap and assess their own intuitions as means to effectively decipher prevailing lay beliefs.[251]

But in circumstances where experts do accurately decipher lay perceptions, there may be countervailing instrumental advantages that weigh in favor of maintaining even those procedures and rules that the public perceives to be unfair or unjust.  In other words, there may be independent reasons—good-faith justifications—for the criminal justice system to adopt and implement procedures and rules that undermine lay perceptions of fairness and justice.

Tensions are endemic to a criminal justice system that responds to discrete (and sometimes competing) justifications for punishment.  For instance, incapacitation may demand longer sentences for less volitional actors who are more likely to recidivate.  By contrast, desert may call for less punishment for these same people because they lack the capacity to control their conduct.  But such gulfs are not always unbridgeable.  That is why legitimacy and moral credibility are such promising enterprises.  Specifically, an expressive justice system exploits seemingly elusive common ground between instrumental and deontic impulses.[252]  By reflecting lay perceptions of fairness and justice, such a system does better to optimize crime control than a system that relies exclusively on conventional instrumental deterrence as a distributive principle, and it does better to approximate deontological desert (to the extent there is any such absolute) than a system that structures its rules and procedures around the intuitions and preferences of moral philosophers.[253]  But the common ground is, of course, not limitless, and hard decisions may follow.

A.     Enforcement Practices

Consider the tradeoffs at work in setting enforcement practices.  Keeping with the earlier example of order-maintenance policing, the practice carries with it a number of institutional advantages that stand apart from normative crime control.  Specifically, order-maintenance policing may improve the quality of life in high-crime areas not only by responding to the immediate disorder but also by facilitating a high volume of searches and arrests that help police discover evidence of more serious past or present crimes and compile a database that police may use to investigate future crimes.[254]  One of us has criticized aggressive use of order-maintenance policing as a low-cost, high-volume (and therefore easily over-used) mechanism to stop, search, and arrest.[255]  As indicated, it is unclear whether the public agrees with the criticism—that is, whether it perceives aggressive order-maintenance policing to be unfair and, if so, under what circumstances.  But, significantly, even if the public agrees, its misapprehensions would and should not end the analysis.  A final determination of whether and how often the approach ought to be used must take into account the independent institutional police advantages.  Ultimately, the institutional advantages of order-maintenance policing may trump even the costs that flow from genuine perceptions of unfairness.[256]

Moreover, a system must measure a normatively problematic practice against the available alternatives.  For example, one of us has made the argument that aggressive use of order-maintenance policing is a less intrusive enforcement strategy than historical exercises of rough justice.[257]  Specifically, police are no longer permitted to crack skulls or to exploit formless loitering statutes, but they can lawfully stop and frisk.[258]  Indeed, a prominent legal historian has argued that the Supreme Court was aware of this tradeoff, holding vagrancy statutes unconstitutional only after formulating Terry as a viable substitute.[259]  On this reading, Terry may be an equally effective but less invasive (and therefore more defensible) enforcement strategy, even if the public fails to perceive it as such, which, as indicated, is an open question.[260]

With respect to any police practice, the core question ought to be whether a court appropriately balances individual liberties and effective law enforcement.  If a court does so, then its decision is proper, even if the public fails to perceive that the balance is adequately struck.  But, as discussed, effective law enforcement is probably undermined (at least to a degree) when the public believes authorities are behaving unfairly.[261]  Thus, perceptions are critical to the balance, but not necessarily determinative.  For example, the public may disagree with the Court’s ruling that police are authorized to arrest individuals for even nonjailable offenses.  But, as Richard Frase has argued, officers could not so readily engage in order-maintenance policing without the power.[262]  Thus, independent institutional advantages may trump lay perceptions of unfairness.  Likewise, the public may feel that the Court has been too deferential in delineating the permissible scope and bases for Terry stops and frisks.  But such deference may be necessary to optimize crime control and, again, to effectuate order-maintenance policing.  As stated before, the Court’s constitutional focus is on the balance between liberty and order,[263] and we think that to be the right focus.  Our bottom line is simply that lay perceptions ought to matter to this balance.  Thus, the Court ought to consider as a factor what, for example, the public thinks of pretextual stops,[264] exigency and consent searches,[265]plain-view seizures,[266] and the use of evidence of flight and presence in high-crime neighborhoods as relevant to determinations of the existence of reasonable suspicion and probable cause.[267]  And, if the public disapproves of any of these practices, standards, and rules, then the Court probably ought to defer to public preferences for equally effective alternatives, provided such alternatives exist.  But, at the same time, the Court ought to endorse unpopular practices, standards, and rules where there are sufficient good-faith reasons to continue them.  For instance, the Court ought not disturb lightly those practices critical to the nation’s war on terrorism—whatever the public may think of the practices in question.  In this way, perceptions may do their best work as a kind of tiebreaker.

B.     Adjudication Procedures

Just as a system may have plausible reasons to adopt police practices that contravene popular perceptions of fairness, so too may may implement adjudicatory practices that advance some overriding justifiable alternative purpose that cannot effectively be served by procedures that accord narrowly with lay preferences.  Concretely, adjudicatory ends that are “legitimate” in the colloquial sense (that is, defensible) may counterintuitively run up against the public’s conceptions of legitimate (that is, procedurally fair) adjudicatory practices.

To understand the conflicting values potentially at play, consider the exclusionary rule and other rules and standards intended to deter official misconduct.  A system may justifiably settle on an exclusionary rule that deters police misconduct, even if people perceive the rule to provide guilty defendants with unfair windfalls that promote unequal treatment and underenforcement.[268]  Conversely, a system may justifiably limit the reach of an exclusionary rule that provides no added deterrent, even if people desire a more robust rule.[269]  Likewise, a system may justifiably adopt a double jeopardy rule that deters prosecutorial harassment or a speedy trial rule that deters prosecutorial foot-dragging, even if the public feels such rules are, alternatively, insufficiently or overly protective.  The takeaway is not that perceptions of legitimacy ought to trump other considerations or that other considerations ought to trump legitimacy—only that there ought to be a balance of the competing objectives and values at play.  And there are empirical questions not just about public perceptions but also about whether alternative instrumental ends are well served by deviating from public perceptions.  Ultimately, then, perceptions matter, but they are not all that matters.  The problem, at present, is that it is not clear whether the exclusionary rule is even perceived to be unfair (and, if so, in what circumstances), much less whether any perceptions of unfairness are outweighed by good-faith justifications for keeping or, alternatively, limiting or scrapping the rule.

An apt analogy may be to the frequently drawn distinction between conduct and decision rules.  Categorical substantive conduct rules are intended to shape lay behavior and reflect moral intuitions, while procedurally oriented decision rules are intended to optimally constrain state power and/or soften rigid application of conduct rules.[270]  Concretely, conduct and decision rules serve separate values.  But, critically, they must coexist in a justice system.  Thus, the system may justifiably endorse and adopt, say, an exclusionary rule even if the public sharply disfavors the consequent Type II errors (that is, wrongful dismissals and acquittals).  In such circumstances, legitimacy costs are the necessary and inevitable byproduct of “acoustic leakage” between desirable conduct rules that track lay intuitions and (otherwise) desirable decision rules that effectively cabin authority and facilitate individualized adjudication.[271]

The exclusionary rule is but one example of an adjudicatory procedure that may be justifiably kept or discarded for reasons independent of lay perceptions of fairness.  The following is a rough-and-ready (and far from exhaustive) list of additional examples of potentially defensible adjudicatory rules, standards, and practices that, nevertheless, may deviate from public perceptions of fairness.  First, a well-functioning justice system almost certainly must abide some amount of Type I error (that is, wrongful conviction), even if people find fair only an evidentiary standard of proof beyond all doubt.[272]  Second, a system may justifiably consider the psychological harm to a child victim of sex abuse, even if people perceive it to be unfair to restrict the ability of a defendant to confront his accuser face-to-face.[273]  Third, a system may permit lawyers to make unfettered nondiscriminatory peremptory challenges to help ensure that impartial juries are empaneled.[274]  Fourth, a system may procedurally bar defendants and convicts from pursuing many types of innocence claims to promote finality, certainty, efficiency, and rule utility.[275]  Fifth, a system may prohibit trial jurors from learning sentencing consequences pre-verdict to advance rule-of-law considerations and to minimize nullification concerns.[276]  Sixth, a system may exclude victims from the adjudicatory process to promote efficiency, professionalism, and uniform decision-making.[277]

Finally, no analysis of procedural rules and standards would be complete without considering plea bargaining—the most prevalent adjudicatory practice in American criminal justice.  Plea bargaining promotes certainty, autonomy, and, even perhaps, proportionality in an age in which determinate sentencing laws are insufficiently tailored to desert.[278]  More than anything else, plea bargaining facilitates the efficient resolution of criminal cases—a benefit that is not just desirable but perceived to be necessary in a criminal justice system that features high caseloads, costly rules of procedure, and inadequate resources to provide trials to even a sizable fraction of defendants.[279]  Indeed, the instrumental arguments in favor of plea bargaining are so powerful that many critics have abandoned efforts to abolish the institution.[280]  As such, the legitimacy project’s better focus may be targeted reforms of discrete plea-bargain practices that the public may find to be particularly unfair.  For example, the public may perceive Alford pleas to be more illegitimate than conventional pleas, and, significantly, the instrumental benefits of the pleas are comparatively less clear-cut.  Specifically, Alford pleas are said to contribute to expeditious case processing and to provide a mechanism for innocent defendants to avail themselves rationally of the advantages of bargaining, but the pleas may concurrently facilitate wrongful convictions and thereby undermine the systemic search for truth.[281]  Another potential area for reform involves cooperation agreements, which are intended to serve crime-control objectives by permitting law enforcement to penetrate the upper echelons of criminal syndicates, but which may lead not to the capture of big fish but only to the capture of bigger schools of small fish.[282]  Finally, many jurisdictions prohibit judicial involvement with plea bargaining,[283] notwithstanding findings showing that people perceive plea bargaining to be more legitimate the more it comes to resemble mediation.[284]  On the one hand, such bars to judicial participation may promote neutrality in the supervision of guilty pleas (or trials in the unlikely event the parties fail to consummate pleas).[285]  On the other hand, judicial participation may check prosecutors who may use superior bargaining power to leverage inequitable pleas.  In short, plea bargaining is no one-dimensional practice, and, accordingly, perceptions of legitimacy may have a lot to say about discrete aspects, even if the practice is—as it were—too big to fail.

C.     Liability and Punishment Rules

Some deviations from community views of justice may reflect simply a failure to understand those views.  Serious empirical research on the issue is relatively recent.[286]  However, in many other instances, an admitted deviation from empirical desert is justified by some specific interest that it advances.  Below we consider two sorts of justifications: first, rules whose drafters acknowledge the value of a desert distribution but who argue that, given the complexities of the real world, that desert goal is sometimes best achieved by adopting a legal rule that on its face may not seem to put desert first; second, rules that openly reject desert in favor of some other interest, including reducing future crime, or promoting an interest unrelated to criminal justice, such as assisting international diplomacy possibly by providing immunity to diplomats.  Consider the deviations from desert reported and documented by Robinson and Cahill.[287]

Some deviations from desert are adopted out of fear that a more desert-based rule would be subject to easy manipulation and abuse, and thus would produce less justice, not more.[288]  Rules of this sort include the criminal law’s near-universal common rejection of a reasonable mistake of law defense, and the rejection of an insanity defense by some states and limitations on its availability in other states.[289]  It also is common for states to ignore the individual characteristics of a defendant in making liability judgments, as in judging provocation or negligence, including ignoring the person’s incapacities that make it difficult, if not impossible, for the person to have avoided the violation.[290]

A deviation rule also may be adopted because a more desert-based rule would encounter evidentiary problems, which would reduce the reliability of the process.  For example, statutes of limitation were adopted to avoid the dangers of stale evidence.[291]  Strict liability is imposed in cases where culpability is likely to exist but may be difficult to prove.[292]  And coerced confessions and uncounseled lineups are excluded to avoid false recriminations.

Another reason for a rule that on its face would seem to promote failures of justice is its tendency to promote justice in many other cases, often cases seen as more important, where the failure of justice would be more outrageous.  Plea bargains and witness immunity are examples.  They may be granted for quite serious offenses if the cooperation thereby gained will allow the successful prosecution of even more serious offenses by others.

These deviation rules do risk undermining the system’s moral credibility, and they ought to be maintained only if there is no other, nondeviation means of achieving the objective.  A careful review of these rules and their effects may suggest that not all are fully justified in their present form.  Nonetheless, some of these deviation doctrines will stand up to that scrutiny because they really do promote justice.[293]  Good intentions count a good deal in setting reputation, so it would be worthwhile for the criminal justice system to make clear its desert-based rationales in adopting doctrines that appear to deviate from desert.  It can be important to the system’s moral credibility that it is in any case trying to do justice as best it can in a complex world.

Other doctrines that deviate from desert do openly sacrifice desert, typically to pursue some other interest thought to be important.  Most obvious are the doctrines distributing criminal liability and punishment to optimize general deterrence or incapacitation of the dangerous, which deviate from desert to advance those traditional coercive crime-control programs.[294]

General deterrence may present crime-control opportunities because of its potential to affect an entire population of potential offenders.  The deviation from desert in a single case or a small group of cases might be enough to send an effective deterrent message to a very large group of potential offenders.  On the other hand, as one of us has shown elsewhere,[295] it is likely that one or more of the prerequisites for a deterrent effect will be missing, thereby subverting the possibility of such striking gains.

A deterrence-based rule can have no effect unless the target audience knows of the rule, directly or indirectly, yet the studies show that such knowledge of legal rules is weak, even among those who have special reasons to learn those rules.  Further, even if the target knows of the rule, a rule can have no effect unless the target is a rational calculator who can and will choose to act in his rational self-interests.  Yet the majority of the people most likely to need the coercive deterrent threat are not such rational calculators.  Finally, even if the target knows the rule and is a rational calculator, he will not be deterred unless his rational calculations lead him to believe that the risk of committing the offense outweighs its benefits.  Yet, again, the data suggests that the low conviction rates that exist for almost all offenses make the risk of punishment in most situations sufficiently low that most targets will ignore it.  Having a criminal justice system will deter, but manipulating liability and punishment rules within that system will work only in the atypical cases in which all three of these prerequisites to effective deterrence will be satisfied.

Add to this the fact that there is already a deterrent threat inherent in just punishment; deterrence-based rules can do better than desert-based rules only in those cases in which the former deviate from desert—do injustice or cause a failure of justice—yet it is in just these deviation cases that the deterrence program is at its weakest.  People assume the law is as they think it should be—which they think is a desert-based rule—hence they will know of a different, deterrence-based rule only if the system has undertaken a special education campaign.  And it is in these deviation-from-desert cases in which people—citizens, jurors, witnesses, police, prosecutors, and judges—are most likely to subvert and resist the system rather than to help it.

Indeed, it does not follow that a deviation from desert is justified even in those instances in which the deterrence prerequisites are satisfied.  A deviation is justified only if the general deterrent effect would be so great as to outweigh even the long-term detriment to the criminal justice system’s moral credibility from such a deliberate choice to do injustice or to fail to do justice.  Even a single well-publicized case that conflicts with community intuitions of justice (and recall that well-publicized cases are often the most useful for general deterrence) can have a serious detrimental effect on the criminal justice system’s reputation because the deliberateness of the deviation reveals the system’s lack of full commitment to doing justice.

Because incapacitation of the dangerous is so effective at preventing future crimes by the individual offenders detained, it may present substantial crime-control opportunities.  But it is not necessarily true that those opportunities will regularly justify deviation from desert.  First, of course, such deviations undermine the criminal law’s moral credibility.  But beyond that, such preventive detention faces other hurdles.  The cost of a deviation from desert can be justified only if its crime control cannot be achieved through nondeviation means.  In this instance, the possibility for civil preventive detention of dangerous persons means that incapacitation cannot justify undermining the moral credibility of the criminal justice system.  And, as one of us has argued elsewhere, such an open civil preventive system is likely to be both fairer to detainees and more protective and less costly for the community.[296]

Beyond these crime-control interests of general deterrence and incapacitation, other criminal justice related interests are also offered as justification for deviations.  As Parts II.A and III.A make clear, fairness in reaching a result, not just the justness of the result, can be important to society.  At stake here is not only the deontological interest in fairness but also the practical interest in the power of legitimacy discussed previously.  As Part V explains, the demands of fairness may suggest procedures or practices that tend to frustrate doing justice.[297]

In preparation for Part V, let us give a few examples to illustrate the point.  The legality principle bars conviction for offense conduct that was not specifically described in a previously existing prohibition, even if most people, including the offender, believed that the conduct was prohibited.  The exclusionary rule may bar the use of clearly reliable evidence in order to discourage police from engaging in unauthorized searches or seizures, even if such exclusion lets a clearly guilty offender go free.  Speedy trial rules, designed to discourage prosecutorial delay, can have a similar effect.  The bar against “double jeopardy” operates to limit prosecutorial abuse through repeated prosecutions, even if it means the clearly guilty will escape the punishment they deserve.  The entrapment defense, which is designed to discourage overzealous police, can give a defense even if the offender is a career criminal looking for an opportunity to commit the offense.  Like the exclusionary rule, the entrapment defense, especially the objective police-misconduct formulation of it as appears in the Code,[298] seeks to control police overreaching.

Still other deviation rules are justified on grounds unrelated to the criminal justice system,[299] as with diplomatic and official immunity, which are said to promote international interchange and governmental independence, respectively.  Similarly, non-criminal-justice interests are being advanced when the unique condemnatory power of criminal conviction is used to boost the prohibition of minor regulatory violations.

Each of these deviation rules may have some justification, but there is also reason to believe that each incurs a cost to the criminal justice system by undermining its moral credibility.  The detrimental effects of such reduction in moral credibility suggests that each deviation rule merits reevaluation to determine whether its benefits outweigh its costs and whether the same benefits might be as effectively produced by a nondeviation means.

V.  The Occasional Conflicts Between Legitimacy and Moral Credibility

The previous discussions have made clear that perceptions of fairness in enforcement and adjudication are distinct from and independent of perceptions of justice in the distribution of liability and punishment.  This is not to say that the two dynamics—legitimacy and moral credibility—are unrelated.  To the contrary, they are often mutually reinforcing.[300]  Significantly, however, they are not always symbiotic.  A procedurally fair system may generate seriously unjust results, and a procedurally unfair system may nonetheless produce just results.  In short, the police practices, criminal adjudication procedures, and criminal liability rules within a jurisdiction may be in very different states.  More importantly for our purposes, not only is it possible for fair practices and procedures and just punishment to be on different tracks, but sometimes they are on a collision course.  Specifically, practices and procedures that advance fairness sometimes can undermine justice.  And enforcement practices and adjudication procedures that would most effectively advance justice may be seen as unfair.

A.     Points of Tension

Consider, for example, such stalwarts of the American criminal justice system as the prohibition against allowing prosecutors to rely on illegally seized evidence, retry acquitted defendants, or delay trials as best suits effective prosecution.  The rights against double jeopardy and to a speedy trial, as well as the exclusionary rule, all have been constitutionally enshrined to some extent.  Yet it may well be that the virtues that drive these procedural rules are not accuracy in truth finding or reliability in doing justice.  On the contrary, each of these rules, and many others, can easily frustrate justice. [301]

The exclusionary rule can exclude reliable evidence that allows the perpetrator of even a serious offense to go free, a result that cannot help but draw the criminal justice system into disrepute, at least with regard to its commitment for doing justice.  In the case of Larry Eyler, for example, police suspected Eyler of a string of gruesome killings of young gay men.  When a state trooper just happened upon Eyler parked on the side of the highway preparing for another kill of a young hitchhiker, he became suspicious, called headquarters and heard of prior suspicions, and took Eyler to the station, probably saving the hitchhiker’s life.  A search of Eyler’s vehicle turned up conclusive proof of his previous crimes, but the court excluded the evidence because the search was unlawful.[302]  Eyler was released to kill again, and indeed did so before subsequently being captured and convicted for the later crime.[303]  Many may wonder whether this frustration of justice, together with its high cost in human life, is worth the benefits that the exclusionary rule offers.

The double jeopardy bar may present a similar situation.  In the case of Melvin Ignatow, for example, Brenda Schaefer was brutally raped, tortured, and killed by Ignatow and his former girlfriend.  At trial, the girlfriend testified for the prosecution but came off as an unreliable witness, and Ignatow simply lied his way to reasonable doubt.  He was acquitted.  Ten months later, as the new owners of Ignatow’s former house were putting down new carpeting, they found film taped inside a floor duct.  When the film was developed, it provided a grisly record of Ignatow’s horrendous offense, yet Ignatow could not be retried for the murder.[304]  Again, this gross failure of justice is likely to undermine in many peoples’ minds the system’s commitment to doing justice.[305]

Or imagine that an Eyler or an Ignatow is released because of a speedy trial violation, a statute of limitations has run, or the text of an offense statute was ambiguous (even though the defendant knew his conduct was wrong).[306]  The fairness interests may be clear—speedy trial rights, statutes of limitations, and the legality principle are common and well established—but the justice costs can be significant.

Nonetheless, there are good reasons to insist on adhering to the conventional standards and rules that are premised on fairness concerns.  First, and obviously, fairness is an important value in itself.  But there are practical crime-control reasons beyond this, as Parts II.A and III.A have shown.  But one can say more.  For example, the system’s adherence to these fairness rules, even in such costly cases, advertises the extent of its commitment to them.  Indeed, it is the costs of undermining justice in discrete cases that may do the most to advertise just how devoted the system is to these fairness interests.  If the system is willing to follow such rules, even when they undermine justice in such egregious cases, the message says, then citizens can have confidence that the rules certainly will be followed in the more common, less egregious cases.  That demonstration of high commitment enhances the system’s legitimacy, with its consequent benefits of greater deference and compliance.

However, one can imagine ways in which a society might strike a different balance between fairness and justice on these, and other, issues.  A system might limit application of the rules, perhaps by applying them less rigorously in cases of serious offenses, as some have suggested.[307]  Or a system might shift to alternative procedures that could effectively advance fairness interests without jeopardizing justice—for example, by replacing the exclusionary rule with a robust civil-compensation or administrative-disciplinary regime that could punish police for unlawful searches of any individual (and not just for unlawful searches of accused offenders).[308]  Or a system might narrow application of rules and standards in circumstances where the threat of injustice is high, but the threat of unfairness is low.  For example, the system might bar application of double jeopardy when a defendant’s deceptive conduct helped generate the original acquittal.[309]

B.     Resolving the Conflict

As the last Subpart demonstrates, although a system should strive to realize both values, this may not always be possible.  Specifically, in the previous Subpart, we explored tensions between legitimacy and moral credibility and identified a number of discrete rules and standards—the exclusionary rule, speedy trial guarantees, and protections against double jeopardy—that may be defensible on fairness grounds even where they promote injustice.  More generally, the question arises: Which objective is superior where a system might achieve only one?  We can look at this question in either of two ways.  We can attempt to resolve the deontological tensions between fairness and justice generally, which is the larger debate that takes up a noticeable part of moral philosophy, criminal procedure, and criminal law theory scholarship.  Or we can attempt to resolve the narrower question of which value ought to be preferred where the goal is gaining deference to and compliance with the criminal justice system.  For most of this Article, our focus has been on the narrow question, but, significantly, our answer would generally be the same under either perspective: although legitimacy may be the superior value in discrete circumstances (as discussed above and below), moral credibility is more often to be preferred in unfortunate circumstances where a system may optimize only one.

Overall, Tyler seems to concede that moral credibility has a much greater effect in shaping compliance than does legitimacy,[310] but no doubt the answer is more complex than to always prefer moral credibility.  It may depend on the setting.  In different contexts, one or the other justification carries potentially greater normative force.  For instance, in the order-maintenance policing context, the state cannot rely on moral credibility because many of the governing laws are borderline regulatory public-order crimes that lack inherent normative punch.[311]  Instead, legitimacy is the sole source of genuine normative power.  In fact, legitimacy may be the sole effective source of any power because traditional carrots and sticks are particularly insufficient for deterring commonplace borderline crime.  Police must necessarily be selective because offenses of this kind are typically so prevalent.[312]  This creates a further complication: selective enforcement may feed perceptions of bias and illegitimacy, which may generate more disobedience, which may lead to even more selective enforcement, which may further feed perceptions of illegitimacy.[313]  The situation is delicate and potentially counterproductive.  Procedural fairness is left alone to do the heavy lifting without the backstops of moral credibility and instrumental deterrence, and the enforcement process, if perceived to be unfair, may succeed only in making the load heavier.[314]  If nothing else, the potential for a negative feedback loop provides yet another powerful reason to reconsider the degree to which we rely on the criminal law to achieve regulatory ends.[315]

Comparatively, when it comes to mala in se crimes, even if the public were to perceive legal authorities to be somewhat illegitimate, and even if an instrumental approach were unable to deter the rational bad actor, ordinary individuals would still tend to comply for the simple reason that ordinary individuals are not, in the main, bad actors.  They obey murder statutes and cooperate with murder investigations and prosecutions (discounting, of course, fears of reprisal) because their moral aversion to homicide is that strong.[316]  With serious mala in se crimes, moral credibility alone may prove somewhat effective, even with low levels of legitimate (or even much of any) enforcement.[317]

In sum, procedural fairness is more important to the enforcement of regulatory crime, while moral credibility is more important to the enforcement of conventional crime.  This may be reason alone to emphasize moral credibility over legitimacy (in the event that a criminal justice system were competent to emphasize only one) because moral credibility operates best within the traditional criminal law domain of mala in se offenses, while legitimacy operates best in only regulatory domains in which state objectives might be achieved equally or nearly as well through civil law or other means.[318]

The question comes down to which concept ultimately provides a better screen.  For a number of reasons, we believe that moral credibility should be expected to more effectively motivate optimal deference.  First, moral credibility entails a concrete assessment of the substantive law or enforcement effort at hand.  Legitimacy, by contrast, entails institutional analyses that operate at higher levels of abstraction.  To produce deference, then, legitimacy is mediated by a cognitive move: the prospective offender must contemplate illegal conduct, then consider what she thinks of the set of procedures used to enforce it, and then decide whether to engage in the forbidden conduct based on her feelings—not toward the law itself—but toward the system that prescribes it.  For moral credibility to produce deference, the prospective offender need only contemplate the illegal conduct and then consider what she thinks of that conduct.  The questions are discrete and coherent: Should I comply with this law?  Is this law morally justified?  Should I cooperate with this prosecution?  Is this prosecution morally justified?  We expect that an individual is likelier to refrain from behavior that she finds immoral than from behavior that she deems neither right nor wrong but that just so happens to be proscribed by legitimate authority.[319]  As Sarah Lawsky has observed: “[A]ssessments of distributive justice might lead to compliance (or noncompliance) directly, without being mediated by an increased belief in legitimacy.”[320]  Indeed, studies have shown that “whereas assessments of procedural justice tend to influence views and beliefs . . . assessments of distributive justice tend to influence behavior.”[321]

Second, moral judgments are innately comprehensible.  We are all social beings with moral compasses that we instinctively consult.[322]  We require no auxiliary understanding to access perceptions of just deserts.  But we must achieve a certain level of socialization for legitimacy to do its work.  We must grasp the objectives, structure, and methods of a justice system and the implications of its procedural practices and strategic choices.  For this reason, perceptions of moral credibility are not just easier to tap, they are likelier to be right.  Fewer external variables cloud our moral valuations.  By contrast, perceptions of legitimacy may fail to reflect reality in fact because these perceptions may be based on incomplete or inaccurate information about supplementary matters of enforcement and adjudication.  Indeed, the fact that perceptions of procedural fairness rely on more than intuition may explain the somewhat greater dissensus that we think we see on questions of procedural as opposed to distributive justice.[323]

Third, legitimacy is an umbrella concept that covers everything from discourteous to discriminatory behavior.  And even if we lack a comprehensive rank ordering of legitimacy criteria, it simply stands to reason that mere rudeness is less likely to undermine deference than perceived immorality.

Fourth, perceptions of procedural fairness are more likely to be socially constructed than perceptions of substantive justice.  Therefore, the corrupt state may more easily manipulate the legitimacy project to serve its bad ends, as we detail below.

Thus, we land in a somewhat different place than Tyler and Darley.  They emphasized legitimacy over moral credibility, arguing that perceptions of procedural fairness are superior because they are more global and thus have the potential to provide legal authorities with broad discretionary power.  Legitimacy provides greater and more reliable authority to legal officials than does morality, because legal officials have discretionary authority to decide what is appropriate.  Within the scope of their prescribed roles, the police and courts make decisions, and citizens believe that they ought to obey those decisions.  Because legitimacy invests authorities with discretionary authority, it is a more flexible form of social value upon which to base the operation of the legal system.  “With morality, the discretion rests with citizens, who decide whether or not the law corresponds to their moral values.”[324]

Moral credibility, by contrast, asks only the narrow question of what the public thinks of a particular law (or at most a particular application of a particular law) and thereby provides less free space for unconstrained decision making.  In short, it provides only a one-off check.

But where Tyler and Darley saw a detriment to the moral-credibility project, we see its chief virtue.  Tyler and Darley may be right that perceptions of legitimacy provide authorities with leeway to make unpopular decisions that are, nevertheless, correct.[325]  But the point presupposes that the decisions are, in fact, correct.  What if their decisions are incorrect or, worse still, corrupt?  As indicated, a false perception of legitimacy may motivate unwarranted deference, and we ought not to want the public to blindly acquiesce to governmental conduct simply because the state has established an ersatz reserve of good will or trust.  Bad results may come to pass when the public no longer makes or acts upon specific judgments as to the appropriateness of discrete governmental action.

To put a finer point on it, we might imagine two systems.  First, in a legitimate system that has no reputation for moral credibility, people will obey the law uncritically because legal authorities ought to be obeyed, leaving unanalyzed the question of whether the law is itself normatively defensible law.  Because police and prosecutors have vested interests in cultivating discretionary power, it is unsurprising that legal authorities should favor such a system.[326]  But should the rest of us?  Not if the best interests of legal authorities fail to align with the public interest.  Second, in a morally credible system that has no reputation for legitimacy, people will be indifferent to legal authorities but will behave morally because it is right to behave morally.  Significantly, in such circumstances, the public will be forever checking and rechecking legal measures, and may choose to defy the state should it break loose from its normative moorings.  It is no flaw that, as Tyler accurately observed, “Morality can lead to compliance with laws, but it can also work against it.”[327]  Rather, the individual may (and typically should) feel compelled to deviate from even the legitimate system that tries to implement and enforce an isolated unjust or immoral liability or punishment rule.  We should want the system to cultivate deference for its morally laden directives.  We may even want the system to cultivate deference for its amoral or morally ambiguous directives.  But we should not want the system to cultivate deference for its immoral directives.

In sum, we think that, descriptively, moral intuitions pack more punch, and, normatively, this is as it should be.[328]

Conclusion

A growing literature on procedural fairness suggests that there is practical value in enhancing a criminal justice system’s “legitimacy” with the community.  A separate literature suggests that there is practical value in enhancing the system’s “moral credibility” with the community it governs by distributing criminal liability and punishment according to principles that track the community’s shared intuitions of justice.  In this Article, we have examined the shared aims and the similarities in the operation and effect of these two criminal justice dynamics as well as their occasional differences in effect and their potential for conflict.

Among other things, we have concluded that the normative influences of the two dynamics are indeed similar, and that they may be mutually reinforcing.  On the other hand, the extent of our knowledge about the two dynamics is different.  While the “legitimacy” dynamic is the better known, and is more frequently used as a justification by scholars, we know less about what practices and procedures will produce legitimacy than we do about what liability and punishment rules will produce moral credibility.  Similarly, at present, there is less empirical support for the claimed beneficial practical effects of legitimacy in producing deference and compliance than there is for moral credibility doing the same.

While the benefits of perceived legitimacy and moral credibility go beyond the deontological to include the practical benefits of advancing effective crime control, it is also true that plausible and good-faith arguments, generally utilitarian in nature, can be made in support of practices, procedures, or rules that are perceived as unfair or unjust.  However, we argue that a system should deviate from the community’s notions of fairness and justice only when: first, that deviation achieves a goal that cannot be achieved through nondeviation means; and, second, the crime-control benefits of the deviation outweigh the crime-control costs inherent in undermining the system’s legitimacy and moral credibility.

Finally, we have shown that sometimes there is even tension between the dynamics of legitimacy and moral credibility, as with such doctrines as double jeopardy, the exclusionary rule, speedy trial, and the legality principle.  While the effect of moral credibility in producing cooperation and deference may be greater than that of legitimacy, the choice between the two is more complex, commonly dependent upon context.  Sometimes, legitimacy is to be prioritized.  More often, we think moral credibility is the superior value.  Happily, it is typically the case that legitimacy and moral credibility work together to support one another in harnessing the powerful forces of social and normative influence in gaining deference and compliance.