Why exactly racial profiling is wrong can be articulated in a number of ways. Some criticisms of racial profiling claim that it simply does not work. Against the defense of racial profiling as a rational way to focus limited law enforcement resources based on statistical data that purportedly links race and criminal offending, these critics empirically demonstrate both that the use of race as a proxy for criminality has a poor track record in producing evidence of criminal wrongdoing and that skepticism regarding its contributions to crime prevention is warranted.¹xSee Dinesh D’Souza, The End of Racism: Finding Values In An Age Of Technoaffluence (New York: Simon and Schuster, 1996), 245-88 for a defense of racial profiling as a form of rational discrimination; Randall Kennedy, Race, Crime, and the Law (New York: Vintage Books, 1998), 136-67 for a critical overview of racial profiling practices; David A. Harris, Profiles in Injustice: Why Racial Profiling Cannot Work (New York: New Press, 2002), 73-90 for a discussion of low ‘hit rates’ when police use race as a factor in making stops and searches; Bernard E. Harcourt, Against Prediction: Profiling, Policing, and Punishing in an Actuarial Age (Chicago: University of Chicago Press, 2007) for a critique of predictive techniques in policing, including racial profiling. Others focus on the harmful effects of racial profiling, such as imposing psychological trauma on the profiled individual, increasing contacts of profiled groups with the criminal justice system, reducing the profiled group’s personal mobility, increasing tensions between police and profiled communities, undermining the perceived legitimacy of the legal system, and other collateral costs of these more direct consequences.²xHarris, Profiles in Injustice, 91-128; Harcourt, Against Prediction, 145-71. Lastly, others argue that racial profiling is not merely wrong because of these consequentialist failings, but also because it violates basic constitutional rights, such as the right to be protected against unreasonable searches and seizures and the right to equal protection of the law.³xTracey Maclin, ‘Race and the Fourth Amendment,’ Vanderbilt Law Review 51 2 (March 1998): 333-93; Brando Simeo Starkey, ‘A Failure of the Fourth Amendment & Equal Protection’s Promise: How the Equal Protection Clause Can Change Discriminatory Stop and Frisk Policies,’ Michigan Journal of Race & Law 18 (2012): 131-87; Anthony C. Thompson, ‘Stopping the Usual Suspects: Race and the Fourth Amendment,’ NYU Law Review 74 (1999): 956.

In this article, I offer another account of why racial profiling is politically, morally, and legally objectionable by demonstrating how it can be understood as a violation of the presumption of innocence (hereafter: PoI). My aim is to show that an account of why racial profiling is wrong is not complete if it only focuses on the harmful consequences of or the rights invasions that result from racial profiling. Rather a comprehensive account of the objectionableness of racial profiling must focus on what is wrong about racial profiling itself, i.e., that it violates the PoI of civilians who are treated as if they engage in criminal behavior without any justification for those suspicions.

Section 1 sets forth an expansive reading of the PoI, which maintains that the PoI is relevant to pre-trial phases of the administration of the criminal law, including policing, and offers protections to citizens against unjustifiable public accusations of criminal wrongdoing. Section 2 provides reasons for why the PoI is helpful in articulating the precise wrongful nature of racial profiling in ways that other legal standards cannot. Section 3 applies the expansive interpretation of the PoI to an example of racial profiling, the New York Police Department’s (hereafter: NYPD) stop-and-frisk policy, and shows how the PoI can be used to normatively evaluate racial profiling and analytically distinguish it from other preventative public safety measures.

• 1 The Presumption of Innocence as a Protection Against Social Condemnation

  The PoI is a fundamental principle of the criminal law. Despite not being mentioned explicitly in the United States Constitution, the US Supreme Court has ruled that ‘the principle that there is a presumption of innocence in favor of the accused is the undoubted law, axiomatic and elementary, and its enforcement lies at the foundation of the administration of our criminal law.’⁴xCoffin v. United States, 156 US 432 (1895), 452. The fact that the principle is so fundamental, however, does not mean that its meaning and scope are not fiercely contested.⁵xAndrew Ashworth, ‘Four Threats to the Presumption of Innocence,’ International Journal of Evidence and Proof 10 (2006): 243-79.

  The narrow interpretation of the PoI holds that it is a common law rule that applies during a criminal trial and places the burden of proof on the prosecution to prove beyond a reasonable doubt that the defendant is guilty. On a traditional interpretation, the PoI is important for three reasons.⁶xOn the PoI’s history and value, see Coffin v. United States, 156 US 432. First, criminal punishment imposes serious hard treatment and deprivations on the convicted, which threatens basic rights to liberty, property, and privacy.⁷xAshworth, ‘Four Threats to the Presumption of Innocence,’ 246-8. Second, to limit wrongful infringements of these rights, the PoI ‘allocates the risk’ of the mistaken decisions of criminal courts, which are fallible fact-finding bodies, in favor of protecting the innocent.⁸xIbid., 248. Third, there is a massive disparity in resources between the state and the individual citizen and as a way of protecting the defendant in the face of the state’s extensive powers, the PoI places the burden on the state to prove its case to a rigorous standard, helping to create a relationship between the state and the citizen that is more amenable to liberal values of limiting state power and respecting individual rights.⁹xIbid., 249-50.

  Despite the PoI enjoying such a fundamental status, it has no relevance beyond the criminal trial per the narrow interpretation.¹⁰xWilliam F. Fox Jr, ‘The Presumption of Innocence as Constitutional Doctrine,’ Catholic University Law Review 28 (1978): 261. On the European Court of Human Rights’ increasingly broad interpretation of the PoI, see Liz Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ The Modern Law Review 76 (2013): 682-8. There is an increasing body of legal scholarship on the PoI, however, that challenges this narrow understanding.¹¹xAntony Duff, ‘Who Must Presume Whom to Be Innocent of What?,’ Netherlands Journal of Legal Philosophy 42 (2013): 170-92; Hamish Stewart, ‘The Right to Be Presumed Innocent,’ Criminal Law and Philosophy (2013): 1-14; Hock Lai Ho, ‘The Presumption of Innocence as a Human Right,’ in Criminal Evidence and Human Rights: Reimagining Common Law Procedural Traditions, ed. Paul Roberts & Jill Hunter (Oxford: Hart Publishing, 2012), 259-81; Liz Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ 681-707. This literature provides a significant philosophical inquiry that examines why the PoI is a basic principle of liberal political morality and how it fits into a more general theory of the relationship between the state and citizens, the aims of the criminal law and the state, respect for human rights, and civic obligation.¹²xAntony Duff, ‘Presumptions Broad and Narrow,’ Netherlands Journal of Legal Philosophy 42 (2013): 273-4. My understanding of the PoI in this article works with the broad interpretation that this literature sets forth. I focus on four features of the broad interpretation, i.e., its foundation, meaning, purpose, and scope.

  First, this literature examines the normative grounding of the PoI. Andrew Ashworth, for example, follows Ronald Dworkin in situating the political and moral importance of the PoI in the role it plays in the state making good on the basic right of the innocent person to not suffer the ‘deep injustice and substantial moral harm’ of wrongful conviction.¹³xAshworth, ‘Four Threats to the Presumption of Innocence,’ 247.

  Similarly, Hamish Stewart maintains that the PoI is a fundamental juridical right that is a basic part of a liberal legal order that aims to respect human freedom, which he claims includes not only rights to free expression and bodily integrity, but also a right to be without reproach.¹⁴xStewart, ‘The Right to Be Presumed Innocent,’ 2-4. Hock Lai Ho provides another human rights based account by arguing that the PoI is more than a common law rule of evidence that demands due process and a right to a fair trial, but is rather a complex of rights that protects individuals, specifically individual liberty, against state coercion.¹⁵xHo, ‘The Presumption of Innocence as a Human Right,’ 261-6. These accounts all share a view of the PoI as an integral feature of the liberal concern with basic human rights.

  Another significant attempt to ground the PoI situates it as a basic principle of civic obligation. On this account, which Antony Duff defends, the PoI is an expression of civic trust that is extended between citizens.¹⁶xDuff, ‘Who Must Presume Whom to Be Innocent of What?,’ 179-82. To presume someone is innocent, per Duff’s interpretation of its social meaning, is to presume that they are a law-abiding citizen and not an enemy who intends to do law-abiding citizens wrong.¹⁷xIbid., 180. Contra certain critics who claim that Duff’s position is merely a practical psychological maxim that suggests that it would be beneficial to view fellow citizens with trust, because it would make for a generally more positive outlook on life,¹⁸xThomas Weigend, ‘There Is Only One Presumption of Innocence,’ Netherlands Journal of Legal Philosophy 42 (2013): 201-2. I understand Duff as making the stronger claim that the trust enshrined in the PoI is both a basic right and duty of political morality. One has a right to be extended civil trust by the state, unless one’s actions justify its withdrawal, because being trusted is a basic feature of one’s status as a citizen. To not be trusted without cause, i.e., for state agents to presume that an individual is guilty, would not merely be an unpleasant or unsociable thing to do, but rather would amount to a denial of that individual’s status as a citizen. I return to this point in more detail shortly.

  Whether grounded in an account of human rights or the obligations of political morality, these accounts help to highlight a comprehensive understanding of the PoI’s purposes. The PoI’s purpose, on the narrow interpretation discussed above, is to protect against the hard treatment and deprivations of criminal punishment, such as fines and imprisonment. It is important to avoid the wrongful imposition of this hard treatment and deprivation, because it infringes on basic rights to liberty, property, and privacy. But, criminal punishment includes more than these material aspects. Criminal punishment is also a social condemnation of the convicted criminal.¹⁹xJoel Feinberg, ‘The Expressive Function of Punishment,’ The Monist 49 (1965): 397-423; Antony Duff, Punishment, Communication, and Community (Oxford: Oxford University Press, 2003); Andrew Von Hirsch, Censure and Sanctions (Oxford: Clarendon Press, 1993). The PoI is not only important to protect against wrongful imposition of the material aspects of criminal punishment, but also to avoid wrongfully condemning an innocent person. Dale Nance argues that the seriousness of condemning someone as a criminal, which is to allege that they have violated the most basic shared principles of the community, by itself justifies the strong burden of proof that the PoI places on the state to justify its censuring of a citizen.²⁰xDale A. Nance, ‘Civility and the Burden of Proof,’ Harvard Journal of Law & Public Policy 17 (1994): 656-7. Along these lines, Liz Campbell has argued that the PoI ‘seeks to prevent the state from castigating someone as a criminal before a finding of guilty and without a certain level of proof.’²¹xCampbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ 689-93. I refer to this protection against social condemnation as the communicative purpose of the PoI.

  The preceding account of the foundation, meaning, and purposes of the PoI highlights the PoI’s importance in maintaining respectful relationships between the state and citizens. On the rights based accounts, the PoI is integral to the respect of a citizen’s rights to liberty, property, and privacy, and the right to be without reproach. Accordingly, Ashworth has commented that valuing the PoI is required for the state to show ‘respect for individual dignity and autonomy.’²²xAshworth, ‘Four Threats to the Presumption of Innocence,’ 251. Focusing on the PoI’s communicative purpose makes this connection between the PoI and respect more specific.

  To see this connection, focus on the meaning of extending the PoI to fellow citizens. By presuming others to be innocent of criminal wrongdoing, one views them as being, like oneself, a moral agent with normative capacities, who can respond to the obligations formalized in the criminal law.²³xSee Duff, ‘Who Must Presume Whom to Be Innocent of What?,’ 181 on civic trust and recognizing others as ‘reason-responsive’ agents. I call this view of one’s fellow citizens moral respect. By presuming others to be innocent of criminal wrongdoing, one views them as being, like oneself, committed to the good of the broader community as someone who takes seriously and meets their most basic obligations to fellow citizens.²⁴xSee Nance, ‘Civility and the Burden of Proof,’ 653 on the principle of civility and respecting others as having membership in the political community. I call this view of one’s fellow citizens political respect. Compare this to the opposing stance, which refuses to extend the PoI. To presume someone to be guilty, without justifying cause, is to not respect them as a moral agent who can reasonably respond to the dictates of the law. This form of moral disrespect institutes a moral hierarchy between those who can and those who cannot be trusted to take moral and legal reasons seriously as guides for action. Without evidence to the contrary, what reasons are there to presume someone to be guilty of a crime if not other than that they are not the kind of being for whom moral and legal reasons have purchase?²⁵xThe feature of moral agency at stake here is not the capacity to choose, but reasonresponsiveness. If autonomy only includes the former feature, then respect for moral autonomy is not at stake (see Nance, ‘Civility and the Burden of Proof,’ 653), but autonomy is at issue if it includes the latter feature. As an act of political disrespect, not presuming others to be innocent casts them as disregarding serious obligations and, in turn, as being someone who is not committed to or, even worse, is a threat to the good of the community. The disrespect of not presuming a citizen to be innocent thus challenges that citizen’s equal moral and political status.

  The real parting of ways between the narrow and more expansive accounts of the PoI occurs when the question at stake is the PoI’s scope. Whereas the narrow reading maintains that the PoI is only applicable to the criminal trial, the expansive reading argues that it is relevant to earlier parts of the criminal process. These expansive accounts have looked to apply the PoI, for example, to the justification of pre-trial detention and bail, decisions of the prosecutor to bring formal criminal charges, restrictions on police investigations, public declarations by state officials concerning the guilt of those who have not been convicted, and the treatment of ex-offenders.²⁶xDuff, ‘Who Must Presume Whom to Be Innocent of What?,’ 174-9, 182-92; Antony Duff, ‘Pre-Trial Detention and the Presumption of Innocence,’ in Prevention and the Limits of The Criminal Law, ed. Andrew Ashworth, Lucia Zedner, & Patrick Tomlin (Oxford: Oxford University Press, 2013), 115-32; Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence’; Stewart, ‘The Right to Be Presumed Innocent,’ 5-11. If the US Supreme Court is right to claim that the PoI ‘lies at the foundation of the administration of our criminal law,’ a legitimate question is raised about why the PoI would only be relevant at one stage of the criminal law’s administration. The motivation behind attempts to extend the PoI beyond the criminal trial is that the protections the PoI offers against burdensome state coercion and unjustifiable social condemnation at trial are also important at other stages in the criminal process.

  For example, Duff has argued that the PoI not only protects defendants against the material, psychological, and normative burdens of unjust conviction, but also protects citizens against the burdens of being put on trial without cause.²⁷xDuff, ‘Who Must Presume Whom to Be Innocent of What?,’ 174-80. Duff demonstrates that, before the criminal trial, the PoI requires that the state have a case, supported by sufficient evidence, which is serious enough to merit bringing formal charges against a citizen and imposing the burdens of having to answer those charges. Just as the PoI places restrictions on when the state can turn a citizen into a defendant by bringing formal charges, I claim that the PoI limits the conditions under which the state can turn a citizen into a suspect by placing them under formal investigation. The PoI expresses the need for justifying coercively imposing on a citizen the burdens of, not only being prosecuted and punished, but also investigated for criminal wrongdoing. The PoI, as I understand it in the remainder of this paper, is relevant to issues of police enforcement of the law and constrains the ways that police can investigate citizens.²⁸xOn the importance of the PoI’s scope extending to the pre-trial phase, see Stewart, ‘The Right to Be Presumed Innocent,’ 5-9.

  It is necessary to respond to two important objections to this expansion of the PoI’s scope before proceeding to apply it. The first objection contends that the PoI is too rigorous of a standard to impose on the police before they investigate or arrest a citizen or on the prosecution before they bring formal charges.²⁹xI want to thank one of the anonymous reviewers of this article for articulating this objection. In fact, the standard imposed on police and prosecutors respectively to engage in such official state acts is far less demanding than what the PoI requires in its standard application to the criminal trial. As Ashworth and Ho have argued, however, no single standard of proof, such as the reasonable doubt standard, is required by the PoI and many different standards of proof are consistent with its purposes.³⁰xAshworth, ‘Four Threats to the Presumption of Innocence,’ 250-1; Ho, ‘The Presumption of Innocence as a Human Right,’ 261-2. What requires the more demanding reasonable doubt standard is, in the trial context, the need to justify punishment and social censure, concern for the moral harm of wrongful conviction, the fallibility of the court’s fact-finding abilities, and the imbalance of power between the state and defendant.³¹xAshworth, ‘Four Threats to the Presumption of Innocence,’ 250-1. This means that it is possible for the PoI to require one standard of proof in one context and a higher or lower standard in a different context.

  This is what Duff tries to capture when he differentiates between defeating and merely qualifying the PoI.³²xDuff, ‘Who Must Presume Whom to Be Innocent of What?,’ 175. The former is accomplished when the state proves guilt beyond a reasonable doubt in a criminal trial, whereas the latter is accomplished, when the police have met the requisite standard to investigate or arrest or when the prosecutor has a case that is sufficiently strong to justify bringing criminal charges. Defeating the PoI justifies conviction and the resulting consequences, whereas qualifying the PoI does not justify treating a citizen as guilty, but rather justifies treating them as a suspect or defendant, depending on how much evidence has been produced.³³xIbid.

  The second objection is that the PoI is not needed to protect citizens from burdensome state coercion at these earlier stages in the criminal process. This objection proposes that other rights, such as the right to liberty, property, and privacy, can protect citizens from unwarranted investigation, prosecution, and pre-trial restrictions.³⁴xWeigend, ‘There Is Only One Presumption of Innocence,’ 198-9. Instead of the PoI, a proportionality test would be used that rules out infringements of basic rights and liberties, for example, if police want to search a home or detain a suspect, if those infringements do not serve some legitimate purpose.³⁵xIbid., 199. Aspects of this test would consider the severity of the crime being investigated and the degree of suspicion, given available evidence, against a suspect.³⁶xIbid.

  For example, one could argue the PoI is not needed to protect citizens from burdensome police searches, because the Fourth Amendment to the US Constitution already provides such protections. There is a sense in which this objection is correct in pointing to a possible redundancy of the PoI at this point in the criminal process. Contra the objection, this redundancy exists, however, because the extant legal standards, such as the Fourth Amendment, already embody the PoI’s normative demands. One feature that can make a police search unreasonable is if it lacks particularized suspicion that a specific individual is engaged in criminal activity.³⁷xTerry v. Ohio, 392 US 1 (1968), 27. This constitutional requirement restricts police from presuming citizens are guilty, which is what they would be doing if they searched or detained citizens for investigative purposes without particularized suspicion supported by evidence. If the concern raised by the objection is that the specific language of the PoI is not necessary to raise this kind of Fourth Amendment challenge, then that is granted. But the important point still holds that the normative requirements of the PoI, which places a burden of proof on the state before it coercively interferes with citizens by administering the criminal law, is already enshrined in requirements of probable cause or reasonable suspicion, i.e., burdens of proof on state agents in their law enforcement activities.

  This claim that the PoI is redundant or unnecessary, however, is not unique to objections to attempts to expand it beyond its normal scope. In fact, a common theme in American jurisprudence is to question whether, even during the criminal trial, the PoI is a necessary jury instruction, if the burden of proof and reasonable doubt standard are already explained.³⁸xSee Fox Jr, ‘The Presumption of Innocence as Constitutional Doctrine,’ 262-6 for a survey of these discussions. The objection that the PoI is duplicative, because, for example, a burden of proof standard is already placed on the police power to search suspects by the Fourth Amendment, is not a new challenge to the PoI, but rather a reformulation of the same challenge in a different arena of the criminal law. My contention here is that the normative content of the PoI is already captured by legal standards that apply in pre-trial phases of the criminal process. It is beyond the scope of my paper to examine in which instances there would be both analytic and normative benefits to introducing explicit language of the PoI into those legal standards. I turn now, however, to an example in which the protections of the PoI, albeit at times overlapping with, are not fully exhausted by existing legal protections.

• 2 The Presumption of Innocence as a Resource for Examining Racial Profiling

  In what remains, I demonstrate how this expansive interpretation of the PoI is both interpretively useful as an analytic tool in explaining the precise wrongful character of racial profiling and as a normative standard in evaluating the legitimacy of racial profiling as a law enforcement practice. The three most relevant features of the expansive interpretation of the PoI for these purposes are that (1) it is normatively relevant in pretrial stages of the criminal process and (2) it places a burden of proof on state agents in those earlier stages that (3) protects citizens from unjustifiable accusations of criminal suspicion or guilt. There are three reasons to pursue the PoI as a tool with which to analyze and evaluate racial profiling despite the challenge of expanding the PoI beyond its traditional context.

  First, the PoI is already a normative language that Americans use to capture what is wrong with racial profiling. For example, a popular book on a high profile example of racial profiling is titled The Presumption of Guilt and challenges racial profiling as a violation of the PoI.³⁹xCharles J. Ogletree, The Presumption of Guilt: The Arrest of Henry Louis Gates Jr. and Race, Class, and Crime in America (New York: Palgrave Macmillan, 2012). This framing of objections to racial profiling should come as no surprise. Even though the US Supreme Court has a narrow interpretation of the PoI as a legal standard, the PoI has a broad purchase amongst Americans who see the principle that one is ‘innocent until proven guilty’ as a basic constraint on blaming practices. One might object to this recourse to the PoI to challenge racial profiling as a sloppy extension of the PoI in an attempt to garner rhetorical benefit from the PoI given its universally cherished status.⁴⁰xWeigend, ‘There Is Only One Presumption of Innocence,’ 193. If the expansive reading of the PoI is compelling in demonstrating the PoI’s broad normative relevance, however, then perhaps something more is at stake than the rhetorical deployment of an important principle.

  Second, the PoI should be explored as a lens with which to examine racial profiling, because of the problems with other existing legal principles to provide protection against racial profiling. The Fourth Amendment protections against unreasonable police searches could be a useful instrument for challenging racial profiling, but the US Supreme Court has either ignored issues of race completely in evaluating Fourth Amendment claims or, more recently, explicitly denied their relevance to Fourth Amendment jurisprudence.⁴¹xWhren v. United States, 517 US 806 (1996), 813; see Thompson, ‘Stopping the Usual Suspects,’ for a critique of this feature of Fourth Amendment jurisprudence. The protections offered by the equal protection clause could offer another venue for challenges to the police use of race as a proxy for increased risk of criminality, but the equal protection clause has proven quite inadequate in offering such protection.⁴²xStarkey, ‘A Failure of the Fourth Amendment & Equal Protection’s Promise: How the Equal Protection Clause Can Change Discriminatory Stop and Frisk Policies.’ This is not to say that these constitutional standards could not protect citizens from racial profiling if interpreted in different ways, but rather that the tendencies in the US Supreme Court’s rulings has been to be deferent to the claims of police that the use of race is necessary to meeting law enforcement purposes.⁴³xKennedy, Race, Crime, and the Law, 136-67. Given the limited success of existing legal standards, exploring the PoI as a possible standard with which to evaluate racial profiling is an important project.

  Third, there are aspects of what is wrong with racial profiling that are not captured, for example, by a Fourth Amendment challenge. Part of what is wrong with racial profiling is that it results in an unjustified police search that infringes individual rights to privacy, property, and personal security. Racial profiling is not only objectionable, however, because of the resulting wrongful search or detention for questioning, but also racial profiling is in itself wrong. Specifically, the state’s labeling of a specific racial group as having special criminal tendencies is wrong on its own terms, even if it did not in fact result in a wrongful stop, search, or detention for investigative purposes. Whereas the Fourth Amendment could offer protection against the police searches and seizures, it does not provide any protection against the spurious accusation by state agents that the citizen who is profiled is engaged in criminal wrongdoing.⁴⁴xCompare to Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ 692. The promise of the PoI is that it provides a way to challenge these spurious accusations, because, on the expansive interpretation, one of the PoI’s basic purposes is to protect against the state’s undeserving social condemnation of citizens.⁴⁵xIbid. Campbell suggests that an expressivist interpretation of the equal protection clause might provide a similar outlet for this kind of challenge. I do not evaluate that possibility here. I now turn to a specific example of racial profiling to demonstrate how the PoI helps to make salient the precise wrong of racial profiling.

• 3 The Presumption of Innocence and the NYPD’s Stop-and-Frisk Policy

  The NYPD’s stop-and-frisk policy is an example of racial profiling that violates the PoI. A stop-and-frisk is a type of field investigation in which a police officer stops, detains, questions, and, in some cases, frisks an individual who they suspect of criminal wrongdoing. The NYPD began aggressively using the tactic in 1994 as part of a comprehensive overhaul of law enforcement strategies.⁴⁶xFranklin E. Zimring, The City That Became Safe: New York’s Lessons for Urban Crime and Its Control (Oxford: Oxford University Press, 2013), 100-150. More recently, however, from 2002 to 2011, there was a sharp increase in stops per year from 314,000 to a high of 686,000.⁴⁷xSee Floyd v. City of New York, 813 F. Supp. 2d 417 (Dist. Court, SD New York 2011), 31-4 for statistics regarding NYPD stop-and-frisk policies over the past decade. There is disagreement about the underlying philosophy of policing that motivates the active use of stop-and-frisk. Some situate it as an outgrowth of Broken Windows or Order-Maintenance approaches to policing.⁴⁸xSee Jeffrey Fagan & Garth Davies, ‘Street Stops and Broken Windows: Terry, Race and Disorder in New York City,’ Fordham Urban Law Journal 28 (2000): 464-76, for an examination of this claim. Others suggest that stop patterns do not fit these strategies, but instead indicate that the goal is to locate individuals with outstanding warrants and remove them from the community.⁴⁹xZimring, The City That Became Safe, 128-9, 145-6, 150. The most straightforward justification of the project, which city officials favor, is that aggressive stop-and-frisk policies reduce the level of violent crime in the city by deterring potential criminals from carrying a weapon for fear of being arrested for a weapons violation.⁵⁰x‘Mayor Bloomberg Delivers Address On Public Safety to NYPD Leadership,’ The Official Website of the City of New York, accessed December 11, 2013, www1.nyc.gov/office-of-the-mayor/news/151-13/mayor-bloomberg-deliversaddress-public-safety-nypd-leadership. The actual contribution of stop-and-frisk policies to lowered crime rates in New York City is an empirically contested issue.⁵¹xOn stop-and-frisk’s contribution to crime prevention, see Zimring, The City That Became Safe, 144-7, 148-50. On racial profiling as a crime control measure, see Harris, Profiles in Injustice, 73-90; Harcourt, Against Prediction, 195-214.

  Between 2004 and 2012, the NYPD made 4.4 million stops, 52 percent of which resulted in frisks.⁵²xSee above at note 47 for the following statistics. Only 6 percent of stops resulted in a summons and another 6 percent resulted in an arrest. A weapon was found in only 1.5 percent of frisks. This means that in 88 percent of the 4.4 million stops no evidence of criminal wrongdoing was found to justify further law enforcement action and in 98.5 percent of the 2.3 million frisks no weapon was found. The low hit rates call into question the criteria used for making stops. The most telling statistic in this regard is that the most common reasons indicated by police for making a stop are ‘Furtive Movements’ and ‘Area Has High Incidence of Report Offense of Type Under Investigation,’ yet stops were more likely to result in an arrest when these were not given as reasons for making the stop. This means that the most common reasons for stops are ‘weak indicators of criminal activity.’⁵³xIbid., 34.

  In addition to low hit rates, the NYPD stop-and-frisk policy also disproportionately impacts minorities.⁵⁴xFor racial disparity statistics in the following discussion, see Floyd v. City of New York, 813 F. Supp. 2d 417, 30-60; also, see Fagan & Davies, ‘Street Stops and Broken Windows,’ 475-96. New York City’s population is, roughly, 23 percent black, 29 percent Hispanic, and 33 percent white, yet 52 percent of those stopped are black, 31 percent are Hispanic, and 10 percent are white. Black and Hispanic people are also more likely to be subject to the use of force, even though white people are more likely to be found with weapons or contraband. Other important evidence of racial disparity is that the rate of stops for a given geographical unit is best predicted by the racial composition of that unit, not the rate of crime.⁵⁵xFloyd v. City of New York, 813 F. Supp. 2d 417, 59. Race is the best predictor of stop rates within a geographical area as well, even after controlling for all other relevant factors.⁵⁶xIbid. Perhaps most troubling of all, despite being disproportionately stopped, stops of black people less frequently result in further law enforcement action than those of white people, indicating that the reasons for which black people are stopped are less founded than those given for stopping white people.⁵⁷xIbid., 60.

  NYPD stop-and-frisk policy can be seen as a form of either indirect or direct racial profiling. The claim that the policy is an indirect form of racial profiling, a less controversial claim, focuses on its use of generalized categories of suspicion, such as ‘Furtive Movements’ and ‘High Crime Area’ to justify stops. As David Harris has argued, not only do such generalized categories not meet the requirement of particularized suspicion, but they are also subject to bias and likely result in disproportionate stops of minorities.⁵⁸xDavid A. Harris, ‘Particularized Suspicion, Categorical Judgments: Supreme Court Rhetoric versus Lower Court Reality under Terry v. Ohio,’ St. John’s Law Review 72 (1998): 1017-19. The reasons that the NYPD policy results in disproportionate racial impacts, however, is most likely not an unintended consequence of using subjective and vague categories for making stops. Rather, this racial disparity is more likely the result of an informal policy of direct racial profiling, which they call targeting ‘the right people.’⁵⁹xFloyd v. City of New York, 813 F. Supp. 2d 417, 81-8 for evidence of this informal policy. The policy focuses stops based on general demographic information from local suspect data.⁶⁰xOn racial profiling methods generally, see Harris, Profiles in Injustice, 1-72. If an individual is a member of a racial group that features heavily in crime reports, then this policy instructs officers to subject that individual to increased scrutiny. The NYPD policy is different than using race as one feature of a detailed description of a specific suspect, which is constitutionally permissible and does not raise the same issues at stake here. Instead what the NYPD policy does is take general demographic data about the race of crime suspects and use it as a generic criminal profile to justify stops. This generic profile usually amounts to nothing more specific than young black and Hispanic males in their teens to early twenties. In effect, the policy of targeting the right people imposes a de facto status of suspect on an entire group, selected in terms of age and race.

  The NYPD’s stop-and-frisk policy violates the Fourth Amendment protection against unreasonable search and seizure. To make a stop, police must have reasonable suspicion that crime is afoot, even if they lack probable cause.⁶¹xFor these standards and relevant case law, see Floyd v. City of New York, 813 F. Supp. 2d 417, 18-26; Terry v. Ohio, 392 US 1. Reasonable suspicion is an objective standard that requires articulable facts and reasonable inferences from those facts, not mere subjective hunches, to justify the suspicion of criminal wrongdoing. Reasonable suspicion also requires particularized suspicion that a specific individual is engaged in criminal activity.⁶²xQuestioning whether US courts enforce these last two standards, see Thompson, ‘Stopping the Usual Suspects,’ 962-73; Harris, ‘Particularized Suspicion, Categorical Judgements.’ For exemptions to the individualized suspicion requirement, see Michigan Dept. of State Police v. Sitz, 496 US 444 (1990); United States v. Martinez-Fuerte, et al., 428 US 543 (1976). The use of generalized categories, such as ‘Furtive Movements,’ and race-based suspicion runs afoul of the particularized suspicion requirement. If the argument made earlier that the particularized suspicion requirement is an embodiment of the PoI’s normative content by placing a burden of proof on the police before they search and detain citizens, then in this sense the NYPD stop-and-frisk policy can be understood as violating the PoI, if not as an explicit legal standard, than at least as a broad normative principle.

  But, the NYPD policy can be seen as a more direct violation of the PoI that is in no way duplicative of Fourth Amendment protections. The US Supreme Court has recognized that one of the PoI’s basic functions is the ‘purging’ function.⁶³xUnited States v. Thaxton, 483 F. 2d 1071 (Court of Appeals, 5th Circuit 1973), 1073; Taylor v. Kentucky, 436 US 478 (1978). This purging function provides the defendant a ‘clean slate’ to begin the trial by reminding the jury to set aside any suspicions of or prejudices toward the defendant that might result from being arrested and charged.⁶⁴xUnited States v. Thaxton, 483 F. 2d 1071, 1074. With its purging function, the PoI sets constraints on the deliberation of the fact-finder during the criminal trial. These constraints render illegitimate any process that judges a defendant to be guilty by starting with predetermined conceptions regarding the defendant’s guilt.

  This purging function is relevant to the deliberative process of police determinations of who to stop-and-frisk. Just as the PoI reminds the judge or jury to provide a defendant with a clean slate before it makes a judgment that the defendant can be punished, the PoI can instruct the police to set aside their unfounded suspicions, prejudices, stereotypes, and biases before it makes a judgment that there is enough evidence to investigate a citizen for criminal wrongdoing.⁶⁵xOn racial stereotyping and policing, see Thompson, ‘Stopping the Usual Suspects,’ 983-91. Whereas the PoI, if applied to the police, would make the police’s deliberative procedures an object of legal scrutiny, the US Supreme Court has held that such subjective evaluations of the police’s thought processes are not relevant to Fourth Amendment considerations.⁶⁶xWhren v. United States, 517 US 806, 811-3. This key difference is another reason for seeing promise in the PoI as a normative standard for evaluating racial profiling.

  If the PoI demands that police view citizens with a clean slate before making judgments about criminal suspicion, then the NYPD’s decision-making procedure for determining whom to stop violates the PoI. First, the NYPD permits the use of generalized categories for making stops, which means that officers are not evaluating the conduct of individual citizens in particular, but are rather making broad categorical judgments about whether they are, for example, in a high crime area or exhibiting evasive behavior. Such general categories, especially when they are subjective, vague, or weak indicators of criminality, do not offer citizens under suspicion a clean slate. Second, the NYPD’s policy of targeting ‘the right people’ encourages the police to view certain racial groups, specifically young black and Hispanic men, with increased suspicion. This instruction to racially target police suspicion is officially and routinely past down from supervisors to patrol officers.⁶⁷xFloyd v. City of New York, 813 F. Supp. 2d 417, 813, 71-7; 84-5. The subjection of racial groups to increased suspicion is an unrecoverable tainting of the clean slate. Third, the deliberative process of patrol officers is further contaminated given the intense institutional pressure placed on them to make high numbers of stops without any counterbalancing checks on the legality of those stops.⁶⁸xIbid., 60-99. In these three ways, the NYPD’s policy can be seen as violating the PoI’s deliberative standards that regulate determinations of how to exercise police power over citizens.

  I am not claiming that any police suspicion of a citizen is a violation of the PoI. If that were the case, then an expansive interpretation of the PoI would rule out any police investigations whatsoever, rendering that expansive interpretation completely untenable. Rather, the PoI as a constraint on police investigations only requires that police accusations of criminal suspicion be made after a burden of proof is first met and that the deliberative procedure for meeting that burden is itself legitimate, i.e., it must begin its evaluation of the citizen’s conduct from a clean slate. The PoI is thus a constraint on the state imposed transition from being an innocent citizen under no suspicion to a suspect, i.e., a citizen who is subject to police investigation because of evidence calling into doubt the citizen’s factual innocence, even though final determinations of legal innocence and guilt are reserved until an actual criminal trial.

  On this account, not every stop-and-frisk is a violation of a citizen’s PoI. The PoI is only violated if the investigating officer did not substantively and deliberatively satisfy the requisite burden of proof for making that stop. It is theoretically possible for certain general categories to satisfy the PoI’s deliberative requirements if they are empirically supported, strong indicators of criminal activity, and helpful to officers in evaluating a citizen’s conduct. The general categories most commonly used by the NYPD, however, do not meet these conditions. General categories should also be subject to extra scrutiny to test if they are in fact applied in a disparate fashion or manipulated in some way. If the NYPD is merely using these general categories as a veneer to dress up a policy of direct racial profiling, however, then its failures to meet the standards of the PoI are even worse. Targeting criminal investigations based on status features, such as race, can never be legitimate under the PoI, because it so thoroughly ruins the possibility of an investigation of a citizen ever starting from a clean slate.

  The NYPD’s policy of racial profiling violates the PoI in a second way. This is by expressing social condemnation of the targeted racial group as having innate criminal tendencies. Although the reprobation at stake in being investigated is not as strong as that which attends a criminal conviction, state sanctioned declarations that a targeted group has criminal tendencies is a sufficiently strong accusation to merit the communicative protections of the PoI.⁶⁹xSee Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ 694-706 for a typology of state practices that criminally label along a continuum of suspicion. Although racial profiling is different than an explicit pronouncement by state officials that a specific individual is guilty, which falls neatly under the ‘reputation-related’ protections of the PoI,⁷⁰xStefan Trechsel & Sarah J Summers, Human Rights in Criminal Proceedings (Oxford: Oxford University Press, 2005), 178-91. it nonetheless is a state act that communicates serious suspicions regarding judgments concerning criminality.

  First, state officials within the NYPD explicitly discuss focusing suspicion on specific racial groups under the policy of targeting ‘the right people’ from meetings of high-level commanders to daily roll call. Even if made behind closed doors, such stigmatizing pronouncements involved in the policy deliberations of a state institution do not sit easily with the expressive protections of the PoI. Second, the public nature of stop-and-frisk encounters renders them inherently expressive acts. Being stopped, detained, frisked, perhaps handcuffed and lined up against a wall, and on occasion subjected to the use of force in open view of one’s fellow citizens is a public declaration of official suspicion that one is likely engaged in criminal behavior. The social meaning of such public acts of state force take on their full meaning against the historical backdrop of issues of policing and race in the United States. A disturbing feature of that history has been a long tradition of associating blackness with criminality, discriminatory law enforcement practices, and a criminal justice system that has operated with de facto and, at times, de jure racialized presumptions of guilt.⁷¹xKhalil Gibran Muhammad, The Condemnation of Blackness: Race, Crime, and the Making of Modern Urban America (Cambridge, MA: Harvard University Press, 2010); Kennedy, Race, Crime, and the Law, 76-135. Although it is not articulated in words, a stop-and-frisk is a public expression of state sanctioned suspicion. When stops-and-frisks are repeatedly performed in public with deeply unequal racial patterns, the NYPD stop-and-frisk policy becomes a stigmatizing theater in which state power is used to cast suspicions on young black and Hispanic men.

  The PoI’s protections demand that public censure by state agents is limited by having met some standard of proof. As is clear by the hit rate data above, no such standards are met in the NYPD’s policy of racial profiling, which means that the policy of targeting ‘the right people’ in fact makes spurious accusations of criminal wrongdoing against innocent civilians. Recalling the earlier discussion that demonstrated how violations of the PoI communicate disrespect by failing to recognize the violated individual’s equal moral and political status, the inherently inegalitarian and politically exclusionary nature of the NYPD’s policy is made apparent. On this account, racial profiling’s public expression of unfounded criminal suspicion is in itself a wrongful practice. The harmful effects on profiled groups, for example, whether they suffer any psychological harm because of this stigmatization, are not necessary to explain why the practice is objectionable.⁷²xCampbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ 690-1 makes this important distinction.

  Despite these normative concerns, defenders of the NYPD stop-and-frisk policy claim that it is no more difficult to justify than a checkpoint used for enforcing laws regarding driving under the influence, which already has the approval of the US Supreme Court.⁷³xSee note 50 above. Per the defense, both measures operate according to the same preventative rationality that seeks to deter the targeted offense, for example, driving under the influence (hereafter: DUI) or carrying concealed weapons. The PoI can help to distinguish the NYPD’s stop-and-frisk policy from indiscriminate preventative public safety measures, such as DUI checkpoints. Using a PoI test, two important differences exist between these practices, despite both raising similar issues with rights to privacy and protection from unreasonable searches. Although both indiscriminate and discriminate practices violate the PoI’s clean slate protections, in the former no one is afforded a clean slate and everyone is subjected to increased suspicion, but in the latter case only some people are not afforded a clean slate. This inequality distinguishes the measures, i.e., that in one case the PoI’s clean slate protections are equally sacrificed for the pursuit of a social goal, whereas in the latter case only certain groups sacrifice those protections, and results in a second difference. This second difference is that indiscriminate measures do not express distrust of anyone in particular and so no one is singled out and stigmatized, whereas measures like the NYPD stop-and-frisk policy explicitly communicate official suspicion of targeted racial groups. These key differences render a defense of the NYPD’s stop-and-frisk policy that draws an analogy to DUI checkpoints unsatisfying. I leave aside any investigation of indiscriminate public safety measures, besides the important point that whatever issues such measures raise with the PoI will be different than the discriminatory policy under examination here.

• 4 Conclusion

  I have argued that that the PoI is relevant in at least two ways to critically examining racial profiling. First, the PoI is interpretatively useful insofar as it brings to the fore the precise wrongful character of racial profiling as a disrespectful expression by state agents of suspicion toward targeted racial groups. Second, the PoI is normatively useful in evaluating the legitimacy of racial profiling by placing under scrutiny both (a) the deliberative process of state agents in their decisions to use police power by asking whether that process affords citizens a clean, unbiased slate before that power is exercised and (b) the social meaning of racial profiling as a public expression of race-based suspicions of likely criminal wrongdoing. If the preceding discussion is compelling, then more work is warranted in exploring the promise of the PoI as a normative standard that regulates the relationship between the police and citizens, specifically when it concerns controversial practices such as racial profiling.⁷⁴xI am thankful to participants in the 2013 Netherlands Journal of Legal Philosophy conference on the presumption of innocence and the anonymous reviewers of the Netherlands Journal of Legal Philosophy for providing valuable feedback on earlier versions of this paper.

Noten

• 1 See Dinesh D’Souza, The End of Racism: Finding Values In An Age Of Technoaffluence (New York: Simon and Schuster, 1996), 245-88 for a defense of racial profiling as a form of rational discrimination; Randall Kennedy, Race, Crime, and the Law (New York: Vintage Books, 1998), 136-67 for a critical overview of racial profiling practices; David A. Harris, Profiles in Injustice: Why Racial Profiling Cannot Work (New York: New Press, 2002), 73-90 for a discussion of low ‘hit rates’ when police use race as a factor in making stops and searches; Bernard E. Harcourt, Against Prediction: Profiling, Policing, and Punishing in an Actuarial Age (Chicago: University of Chicago Press, 2007) for a critique of predictive techniques in policing, including racial profiling.

• 2 Harris, Profiles in Injustice, 91-128; Harcourt, Against Prediction, 145-71.

• 3 Tracey Maclin, ‘Race and the Fourth Amendment,’ Vanderbilt Law Review 51 2 (March 1998): 333-93; Brando Simeo Starkey, ‘A Failure of the Fourth Amendment & Equal Protection’s Promise: How the Equal Protection Clause Can Change Discriminatory Stop and Frisk Policies,’ Michigan Journal of Race & Law 18 (2012): 131-87; Anthony C. Thompson, ‘Stopping the Usual Suspects: Race and the Fourth Amendment,’ NYU Law Review 74 (1999): 956.

• 4 Coffin v. United States, 156 US 432 (1895), 452.

• 5 Andrew Ashworth, ‘Four Threats to the Presumption of Innocence,’ International Journal of Evidence and Proof 10 (2006): 243-79.

• 6 On the PoI’s history and value, see Coffin v. United States, 156 US 432.

• 7 Ashworth, ‘Four Threats to the Presumption of Innocence,’ 246-8.

• 8 Ibid., 248.

• 9 Ibid., 249-50.

• 10 William F. Fox Jr, ‘The Presumption of Innocence as Constitutional Doctrine,’ Catholic University Law Review 28 (1978): 261. On the European Court of Human Rights’ increasingly broad interpretation of the PoI, see Liz Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ The Modern Law Review 76 (2013): 682-8.

• 11 Antony Duff, ‘Who Must Presume Whom to Be Innocent of What?,’ Netherlands Journal of Legal Philosophy 42 (2013): 170-92; Hamish Stewart, ‘The Right to Be Presumed Innocent,’ Criminal Law and Philosophy (2013): 1-14; Hock Lai Ho, ‘The Presumption of Innocence as a Human Right,’ in Criminal Evidence and Human Rights: Reimagining Common Law Procedural Traditions, ed. Paul Roberts & Jill Hunter (Oxford: Hart Publishing, 2012), 259-81; Liz Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ 681-707.

• 12 Antony Duff, ‘Presumptions Broad and Narrow,’ Netherlands Journal of Legal Philosophy 42 (2013): 273-4.

• 13 Ashworth, ‘Four Threats to the Presumption of Innocence,’ 247.

• 14 Stewart, ‘The Right to Be Presumed Innocent,’ 2-4.

• 15 Ho, ‘The Presumption of Innocence as a Human Right,’ 261-6.

• 16 Duff, ‘Who Must Presume Whom to Be Innocent of What?,’ 179-82.

• 17 Ibid., 180.

• 18 Thomas Weigend, ‘There Is Only One Presumption of Innocence,’ Netherlands Journal of Legal Philosophy 42 (2013): 201-2.

• 19 Joel Feinberg, ‘The Expressive Function of Punishment,’ The Monist 49 (1965): 397-423; Antony Duff, Punishment, Communication, and Community (Oxford: Oxford University Press, 2003); Andrew Von Hirsch, Censure and Sanctions (Oxford: Clarendon Press, 1993).

• 20 Dale A. Nance, ‘Civility and the Burden of Proof,’ Harvard Journal of Law & Public Policy 17 (1994): 656-7.

• 21 Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ 689-93.

• 22 Ashworth, ‘Four Threats to the Presumption of Innocence,’ 251.

• 23 See Duff, ‘Who Must Presume Whom to Be Innocent of What?,’ 181 on civic trust and recognizing others as ‘reason-responsive’ agents.

• 24 See Nance, ‘Civility and the Burden of Proof,’ 653 on the principle of civility and respecting others as having membership in the political community.

• 25 The feature of moral agency at stake here is not the capacity to choose, but reasonresponsiveness. If autonomy only includes the former feature, then respect for moral autonomy is not at stake (see Nance, ‘Civility and the Burden of Proof,’ 653), but autonomy is at issue if it includes the latter feature.

• 26 Duff, ‘Who Must Presume Whom to Be Innocent of What?,’ 174-9, 182-92; Antony Duff, ‘Pre-Trial Detention and the Presumption of Innocence,’ in Prevention and the Limits of The Criminal Law, ed. Andrew Ashworth, Lucia Zedner, & Patrick Tomlin (Oxford: Oxford University Press, 2013), 115-32; Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence’; Stewart, ‘The Right to Be Presumed Innocent,’ 5-11.

• 27 Duff, ‘Who Must Presume Whom to Be Innocent of What?,’ 174-80.

• 28 On the importance of the PoI’s scope extending to the pre-trial phase, see Stewart, ‘The Right to Be Presumed Innocent,’ 5-9.

• 29 I want to thank one of the anonymous reviewers of this article for articulating this objection.

• 30 Ashworth, ‘Four Threats to the Presumption of Innocence,’ 250-1; Ho, ‘The Presumption of Innocence as a Human Right,’ 261-2.

• 31 Ashworth, ‘Four Threats to the Presumption of Innocence,’ 250-1.

• 32 Duff, ‘Who Must Presume Whom to Be Innocent of What?,’ 175.

• 33 Ibid.

• 34 Weigend, ‘There Is Only One Presumption of Innocence,’ 198-9.

• 35 Ibid., 199.

• 36 Ibid.

• 37 Terry v. Ohio, 392 US 1 (1968), 27.

• 38 See Fox Jr, ‘The Presumption of Innocence as Constitutional Doctrine,’ 262-6 for a survey of these discussions.

• 39 Charles J. Ogletree, The Presumption of Guilt: The Arrest of Henry Louis Gates Jr. and Race, Class, and Crime in America (New York: Palgrave Macmillan, 2012).

• 40 Weigend, ‘There Is Only One Presumption of Innocence,’ 193.

• 41 Whren v. United States, 517 US 806 (1996), 813; see Thompson, ‘Stopping the Usual Suspects,’ for a critique of this feature of Fourth Amendment jurisprudence.

• 42 Starkey, ‘A Failure of the Fourth Amendment & Equal Protection’s Promise: How the Equal Protection Clause Can Change Discriminatory Stop and Frisk Policies.’

• 43 Kennedy, Race, Crime, and the Law, 136-67.

• 44 Compare to Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ 692.

• 45 Ibid. Campbell suggests that an expressivist interpretation of the equal protection clause might provide a similar outlet for this kind of challenge. I do not evaluate that possibility here.

• 46 Franklin E. Zimring, The City That Became Safe: New York’s Lessons for Urban Crime and Its Control (Oxford: Oxford University Press, 2013), 100-150.

• 47 See Floyd v. City of New York, 813 F. Supp. 2d 417 (Dist. Court, SD New York 2011), 31-4 for statistics regarding NYPD stop-and-frisk policies over the past decade.

• 48 See Jeffrey Fagan & Garth Davies, ‘Street Stops and Broken Windows: Terry, Race and Disorder in New York City,’ Fordham Urban Law Journal 28 (2000): 464-76, for an examination of this claim.

• 49 Zimring, The City That Became Safe, 128-9, 145-6, 150.

• 50 ‘Mayor Bloomberg Delivers Address On Public Safety to NYPD Leadership,’ The Official Website of the City of New York, accessed December 11, 2013, www1.nyc.gov/office-of-the-mayor/news/151-13/mayor-bloomberg-deliversaddress-public-safety-nypd-leadership.

• 51 On stop-and-frisk’s contribution to crime prevention, see Zimring, The City That Became Safe, 144-7, 148-50. On racial profiling as a crime control measure, see Harris, Profiles in Injustice, 73-90; Harcourt, Against Prediction, 195-214.

• 52 See above at note 47 for the following statistics.

• 53 Ibid., 34.

• 54 For racial disparity statistics in the following discussion, see Floyd v. City of New York, 813 F. Supp. 2d 417, 30-60; also, see Fagan & Davies, ‘Street Stops and Broken Windows,’ 475-96.

• 55 Floyd v. City of New York, 813 F. Supp. 2d 417, 59.

• 56 Ibid.

• 57 Ibid., 60.

• 58 David A. Harris, ‘Particularized Suspicion, Categorical Judgments: Supreme Court Rhetoric versus Lower Court Reality under Terry v. Ohio,’ St. John’s Law Review 72 (1998): 1017-19.

• 59 Floyd v. City of New York, 813 F. Supp. 2d 417, 81-8 for evidence of this informal policy.

• 60 On racial profiling methods generally, see Harris, Profiles in Injustice, 1-72.

• 61 For these standards and relevant case law, see Floyd v. City of New York, 813 F. Supp. 2d 417, 18-26; Terry v. Ohio, 392 US 1.

• 62 Questioning whether US courts enforce these last two standards, see Thompson, ‘Stopping the Usual Suspects,’ 962-73; Harris, ‘Particularized Suspicion, Categorical Judgements.’ For exemptions to the individualized suspicion requirement, see Michigan Dept. of State Police v. Sitz, 496 US 444 (1990); United States v. Martinez-Fuerte, et al., 428 US 543 (1976).

• 63 United States v. Thaxton, 483 F. 2d 1071 (Court of Appeals, 5th Circuit 1973), 1073; Taylor v. Kentucky, 436 US 478 (1978).

• 64 United States v. Thaxton, 483 F. 2d 1071, 1074.

• 65 On racial stereotyping and policing, see Thompson, ‘Stopping the Usual Suspects,’ 983-91.

• 66 Whren v. United States, 517 US 806, 811-3.

• 67 Floyd v. City of New York, 813 F. Supp. 2d 417, 813, 71-7; 84-5.

• 68 Ibid., 60-99.

• 69 See Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ 694-706 for a typology of state practices that criminally label along a continuum of suspicion.

• 70 Stefan Trechsel & Sarah J Summers, Human Rights in Criminal Proceedings (Oxford: Oxford University Press, 2005), 178-91.

• 71 Khalil Gibran Muhammad, The Condemnation of Blackness: Race, Crime, and the Making of Modern Urban America (Cambridge, MA: Harvard University Press, 2010); Kennedy, Race, Crime, and the Law, 76-135.

• 72 Campbell, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence,’ 690-1 makes this important distinction.

• 73 See note 50 above.

• 74 I am thankful to participants in the 2013 Netherlands Journal of Legal Philosophy conference on the presumption of innocence and the anonymous reviewers of the Netherlands Journal of Legal Philosophy for providing valuable feedback on earlier versions of this paper.