Symposium - 2021 - Policing, Race, and Power

The Mark of Policing

Race and Criminal Records

Eisha Jain *

Abstract. This Essay argues that racial reckoning in policing should include a racial reckoning in the use of criminal records. Arrests alone—regardless of whether they result in convictions—create criminal records. Yet because the literature on criminal records most often focuses on prisoner reentry and on the consequences of criminal conviction, it is easy to overlook the connections between policing decisions and collateral consequences. This Essay employs the sociological framework of marking to show how criminal records entrench racial inequality stemming from policing. The marking framework recognizes that the government creates a negative credential every time it creates a record of arrest as well as conviction. Such records, in turn, trigger cascading consequences for employment, housing, immigration, and a host of other areas. The credentialing process matters because it enables and conceals race-based discrimination, and because a focus on the formal sentence often renders this discrimination invisible. This Essay considers how adopting a credentialing framework offers a way to surface, and ultimately to address, how race-based policing leaves lasting marks on over-policed communities.


This Essay considers how a moment of racial reckoning in policing could include a focus on criminal records. In the summer of 2020, the killings of George Floyd and Breonna Taylor—coming on the heels of numerous other documented incidents of police violence against racial minorities—triggered a historic social movement in the United States. 1 Larry Buchanan, Quoctrung Bui & Jugal K. Patel, Black Lives Matter May Be the Largest Movement in U.S. History, N.Y. Times (July 3, 2020), (estimating that 15 to 26 million people participated in protests in the summer of 2020); Dylan Lovan, Louisville Protests Continue in Breonna Taylor’s Name, AP News (Sept. 27, 2020), (to locate, click “View the live page”). One important insight from this movement is that the most visible forms of police violence represent only the tip of the iceberg. Police violence is neither unpredictable nor episodic. It is the product of a legal system that gives the police enormous control over communities of color.

This Essay, part of the Stanford Law Review’s Policing, Race, and Power Symposium and of the cross-journal Reckoning and Reformation Symposium, seeks to draw a connection among policing practices, criminal records, and racial inequality. There is a large and growing literature focusing on “collateral consequences,” but courts have tended not to connect these consequences to arrest practices. 2 There is a wide-ranging literature discussing unjustified collateral consequences of criminal convictions, one that is too voluminous to catalogue here. For selected contributions, see generally Jenny Roberts, The Mythical Divide Between Collateral and Direct Consequences of Criminal Convictions: Involuntary Commitment of “Sexually Violent Predators,” 93 Minn. L. Rev. 670, 678-80 (2008); Gabriel J. Chin, The New Civil Death: Rethinking Punishment in the Era of Mass Conviction, 160 U. Pa. L. Rev. 1789, 1790 (2012) (comparing collateral consequences to a form of punishment known as “civil death”); Brian M. Murray, Retributivist Reform of Collateral Consequences, 52 Conn. L. Rev. 863, 916 (2020); Brian M. Murray, Are Collateral Consequences Deserved?, 95 Notre Dame L. Rev. 1031, 1063 (2020) (analyzing collateral consequences from a retributivist perspective of desert); and Michael Pinard, Collateral Consequences of Criminal Convictions: Confronting Issues of Race and Dignity, 85 N.Y.U. L. Rev. 457, 463 (2010) (engaging in a comparative analysis of collateral consequences in the United States and other countries with similar criminal punishment practices and arguing that the “severity of collateral consequences in the United States is rooted in racial marginalization and the narrow dignity interests afforded to individuals with criminal records in the United States.”). That may be because collateral consequences tend to be associated with conviction or with the reentry of former prisoners. But collateral consequences arise from custodial criminal arrest as well as conviction. 3 For some of my own thinking on this subject, see Eisha Jain, Arrests as Regulation, 67 Stan. L. Rev. 809, 826 (2015) [hereinafter Jain, Arrests] (explaining how arrests alone, even in the absence of conviction, trigger penalties in areas such as immigration, employment, public housing, and social services). See also Eisha Jain, Proportionality and Other Misdemeanor Myths, 98 B.U. L. Rev. 953, 963 (2018) [hereinafter Jain, Proportionality] (explaining how punishments for misdemeanor arrests and convictions can be grossly disproportionate once civil penalties are taken into account); Eisha Jain, Capitalizing on Criminal Justice, 67 Duke L.J. 1381, 1391-95 (2018) (discussing how collateral consequences contribute to overcriminalization); Eisha Jain, The Interior Structure of Immigration Enforcement, 167 U. Pa. L. Rev. 1463, 1477 (2019) (discussing how linking removal to the criminal-arrest process affects immigration enforcement); Eisha Jain, Prosecuting Collateral Consequences, 104 Geo. L. J. 1206-07 (2016) (discussing how prior records of arrest affect plea bargaining); Eisha Jain, Jailhouse Immigration Screening, 70 Duke L.J. 1703, 1725-31 (2021) (showing how linking immigration screening to criminal arrest leads to extended carceral treatment within the criminal justice system in service of immigration control). Treating collateral consequences that affect reentry as paradigmatic limits our understanding of the reach of criminal records and their relationship to policing decisions.

This Essay employs the sociological concept of marking to recognize how policing decisions entrench racial inequality. 4 Literature in law and sociology has developed an account of how criminal records function as a “credential” or “mark.” See James B. Jacobs, The Eternal Criminal Record 2 (2015) (comparing a criminal record to a “negative curriculum vitae” that “contains only disreputable information”); James Jacobs & Tamara Crepet, The Expanding Scope, Use, and Availability of Criminal Records, 11 N.Y.U. J. Legis. & Pub. Pol’y 177, 177 & n.3 (2008); Devah Pager, The Mark of a Criminal Record, 108 Am. J. Soc. 937, 939, 942 (2003); Devah Pager, Marked: Race, Crime, and Finding Work in an Era of Mass Incarceration 4-5 (2007) [hereinafter Pager, Marked] (arguing that in the context of criminal convictions, the “power of the credential lies in its recognition as an official and legitimate means of evaluating and classifying individuals”); see also Becky Pettit & Bruce Western, Mass Imprisonment and the Life Course: Race and Class Inequality in U.S. Incarceration, 69 Am. Soc. Rev. 151, 156, 165 (2004) (discussing how “[h]istorically, going to prison was a marker of extreme deviance, reserved for violent and incorrigible offenders,” but that it has turned into a common life marker for low-socioeconomic-status Black men, which leads to diminished life opportunity); James B. Jacobs, Mass Incarceration and the Proliferation of Criminal Records, 3 U. St. Thomas L.J. 387, 413 (2006). A marking framework treats an arrest or conviction as a form of public credentialing. The credentialing process matters because it translates a particular experience into a formally recognized status. For example, an individual could receive a world-class education through her public library, but without the credential of a degree, she would struggle to formalize that knowledge and demonstrate its value. Similarly, a criminal record formalizes the experience of being arrested. It translates a single police officer’s decision to arrest into a marker that has lasting social meaning.

Framing a criminal record as a negative credential offers a way to visualize more expansive legal interventions than focusing on convictions alone. The key questions under a marking framework relate to why, when, and how negative credentials are made and used. The credentialing framework also reveals that not all credentials have the same social meaning. Racial minorities who are the most likely to be subject to criminal arrest may also experience more lasting consequences from their marks.
This Essay discusses how a marking framework can be employed to create a more expansive vocabulary for recognizing and beginning to redress the role of racial discrimination in criminal records. It proceeds as follows. Part I situates the analysis in the context of racialized police violence. Part II shifts from racial discrimination in policing to racial discrimination in the use of criminal records. Part III considers how a credentialing framework could apply to criminal records, and it discusses possibilities for reform.

I. Race and the Discretion to Arrest

In the summer of 2020, seventeen-year-old Darnella Frazier recorded police officer Derek Chauvin killing George Floyd by kneeling on his neck. 5 Joshua Nevett, George Floyd: The Personal Cost of Filming Police Brutality, BBC News (June 11, 2020),; Sarah Mervosh & Nicholas Bogel-Burroughs, Why Derek Chauvin Was Charged with Third-Degree Murder, N.Y. Times (updated June 29, 2020),; Laurel Wamsley, Derek Chauvin Found Guilty of George Floyd’s Murder, NPR (Apr. 20, 2021 5:37 PM ET), The video raised obvious questions about the abuse of government power. Why did the officer use such excessive force? Why did none of the observing officers intervene? Why was Mr. Floyd even subjected to a full custodial arrest for such a minor violation—allegedly paying for cigarettes with a counterfeit $20 bill? 6 Matt Furber, Audra D.S. Burch, & Frances Robles, What Happened in the Chaotic Moments Before George Floyd Died, N.Y. Times (updated June 10, 2020),

The record of Mr. Floyd’s last moments is far from alone in raising these questions. A large body of scholarship conceptualizes police violence as a problem of law. 7 See, e.g., Alexandra Natapoff, Atwater and the Misdemeanor Carceral State, 133 Harv. L. Rev. F. 147, 151 (2020) (“[W]hen the Supreme Court elevates carceral values over individual liberty and privacy, it puts its thumb on the scale in favor of punitive and inegalitarian police practices.”); Amna A. Akbar, An Abolitionist Horizon for (Police) Reform, 108 Calif. L. Rev. 1781, 1791-93 (2020) (“A combined municipal-state-federal legal architecture permits routine police violence by granting police discretion over when and how to arrest or deploy force . . . .”). Constitutional doctrine gives police significant discretion to engage in stops and arrests. In Whren v. United States, the Supreme Court held that police officers may make pretextual traffic stops motivated by racial profiling, so long as there is probable cause for the underlying offense. 8 See Whren v. United States, 517 U.S. 806, 817-18 (1996); see also Alexandra Natapoff, A Stop Is Just a Stop: Terry’s Formalism, 15 Ohio St. J. Crim. L. 113, 114 (2017); Devon W. Carbado, (E)racing the Fourth Amendment, 100 Mich. L. Rev. 946, 1033 (2002). In Terry v. Ohio, the Court held that police may briefly engage in a stop as long as the police officer has “reasonable” suspicion that “criminal activity may be afoot.” 9 Terry v. Ohio, 392 U.S. 1, 16, 19-30 (1968). For criticisms of Terry, see, for example, Paul Butler, “A Long Step Down the Totalitarian Path”: Justice Douglas’s Great Dissent in Terry v. Ohio, 79 Miss. L.J. 9, 26-29 (2009). Beyond stops, police may make full custodial arrests for conduct as minor as a seatbelt violation that is not punishable by jail time, as the Court held in Atwater v. City of Lago Vista. 10 Atwater v. City of Lago Vista, 532 U.S. 323, 354 (2001). The Court in Atwater justified its decision by pointing to historic practice, as well as to the presence of other institutional constraints on the arrest process, reasoning that if police made manifestly unfair or gratuitous arrests, “political accountability” or “good sense” would ultimately serve as checks on police behavior. 11 Id. at 352-54 (“The upshot of all these influences, combined with the good sense (and, failing that, the political accountability) of most local lawmakers and law-enforcement officials, is a dearth of horribles demanding redress.”).

All too often, “good sense” and “political accountability” fail to check police misconduct, including racial profiling. Black and Latino men are disproportionately subject to criminal arrest. 12 As a general matter, one out of three people will be arrested by age twenty-three. See Robert Brame, Michael G. Turner, Raymond Paternoster & Shawn D. Bushway, Cumulative Prevalence of Arrest from Ages 8 to 23 in a National Sample, 129 Pediatrics 21, 25 (2012). According to one estimate, approximately one in two Black and Hispanic men will be arrested by age twenty-three. Robert Brame, Shawn D. Bushway, Ray Paternoster & Michael G. Turner, Demographic Patterns of Cumulative Arrest Prevalence by Ages 18 and 23, 60 Crime & Delinq. 471, 478 (2014). Racial disparities in arrest have been well-documented in the context of marijuana arrests. See, e.g., Benjamin Mueller, Using Data to Make Sense of a Racial Disparity in NYC Marijuana Arrests, N.Y. Times (May 13, 2018), (“In the first three months of [2018], 89 percent of the roughly 4,000 people arrested for marijuana possession in New York City were black or Hispanic.”). Racial disparities also continue in carceral treatment. Dorothy E. Roberts, The Supreme Court, 2018 Term—Foreword: Abolition Constitutionalism, 133 Harv. L. Rev. 1, 13 (2019) (“Most people sentenced to prison in the United States today are from politically marginalized groups—poor, black, and brown. Not only are black people five times as likely to be incarcerated as white people, but also the lifetime probability of incarceration for black boys born in 2001 is estimated to be thirty-two percent compared to six percent for white boys.” (footnotes omitted)). Overwhelmingly, arrests are for low-level offenses. 13 Low-level offenses tend to dominate criminal caseloads. See Alexandra Natapoff, Misdemeanor Decriminalization, 68 Vand. L. Rev. 1055, 1063 (2015) (“Rarely recognized as such, the misdemeanor is in fact the paradigmatic U.S. criminal case: most cases are misdemeanors, most of what the system does is generate minor convictions, and most Americans who experience the criminal system do so via the petty offense process.”). Low-level arrests in particular reflect socioracial disparities in policing practices more than a reasoned response to moral culpability. As Jamelia Morgan has discussed, hundreds of thousands of disorderly-conduct arrests each year function as a way of “reinforc[ing] social hierarchies based on race, gender, sexual orientation, and disability.” 14 Jamelia N. Morgan, Rethinking Disorderly Conduct, 109 Calif. L. Rev. (forthcoming 2021) (manuscript at 5-6), Marijuana arrests are another frequently cited example. It is not just that Black people are arrested far more often than white people for low-level marijuana violations, despite using marijuana at similar rates. 15 ACLU, The War on Marijuana in Black and White 4, 17-21 (2013), (“[O]n average, a Black person is 3.73 times more likely to be arrested for marijuana possession than a white person, even though Blacks and whites use marijuana at similar rates.”). As Bennett Capers has observed, it is also that marijuana prosecutions can “swing like a pendulum,” with arrest practices wildly fluctuating in a short timespan. 16 Bennett Capers, The Under-policed, 51 Wake Forest L. Rev. 589, 595-96 (2016) (noting that in New York City in 1994 the number of marijuana possession arrests was under two thousand but “[w]ithin six years, the number of misdemeanor marijuana arrests per year had increased to more than 50,000, a 2760% increase”); see also Barbara Fedders, Opioid Policing, 94 Ind. L.J. 389, 406 (2019) (discussing racial disparities in drug arrests). These arrest practices reflect socially constructed metrics about who appears out of place or is stereotyped as posing a danger, and they define the boundaries of who belongs along racial lines. 17 For a discussion of how policing practices relate to residential racial segregation, see Monica C. Bell, Anti-segregation Policing, 95 N.Y.U. L. Rev. 650, 655 (2020) (arguing that “there is a mutually constitutive relationship between daily practices of urban policing and residential segregation, a relationship of mutual reproduction”); Angela Onwuachi-Willig, Policing the Boundaries of Whiteness: The Tragedy of Being “Out of Place” from Emmett Till to Trayvon Martin, 102 Iowa L. Rev. 1113, 1119 (2017); I. Bennett Capers, Policing, Race, and Place, 44 Harv. C.R.-C.L. L. Rev. 43, 60-72 (2009); and Jeffrey Fagan & Elliott Ash, New Policing, New Segregation: From Ferguson to New York, 106 Geo. L.J. Online 33, 87-108 (2017).

Racialized policing practices create the risk that police encounters will escalate into deadly violence. Philando Castile—a thirty-two-year-old cafeteria supervisor affectionally known as “Mr. Phil” by hundreds of school children—was killed by a police officer during a routine traffic stop, despite evidence that he did everything requested of him. 18 Angela Onwuachi-Willig & Anthony V. Alfieri, (Re)framing Race in Civil Rights Lawyering, 130 Yale L.J. (forthcoming 2021) (manuscript at 2-5), (reviewing Henry Louis Gates, Jr., Stony the Road: Reconstruction, White Supremacy, and the Rise of Jim Crow (2019)). In their discussion of Mr. Castile’s killing, Angela Onwuachi-Willig and Anthony Alfieri highlighted the central role of racial bias. The officer who killed Castile “could not see Castile as anything more than a racial stereotype,” one that made the officer feel “apprehensive of Castile and [led him to] read [Castile] as dangerous almost from the beginning.” 19 Id. at 3. In the thirteen years before he was killed, Castile had been subjected to traffic stops approximately fifty times. 20 Id. at 2.

Racial disparities are particularly evident in programmatic policing practices that subject communities of color to the repeated risk of being stopped and arrested. 21 Tracey L. Meares, Programming Errors: Understanding the Constitutionality of Stop-and-Frisk as a Program, Not an Incident, 82 U. Chi. L. Rev. 159, 165 (2015) (arguing that while the constitutional framework for regulating stop-and-frisk practices “is based on a one-off investigative incident, many of those who are stopped—the majority of them young men of color—do not experience the stops as one-off incidents” and arguing that the legal doctrine should recognize the significance of programmatic policing). Evidence from New York City’s stop-and-frisk litigation shows how police subjected entire communities to race-based stops, as more than 80% of those subjected to stops were racial minorities. 22 Floyd v. City of New York, 959 F. Supp. 2d 540, 573-74 (S.D.N.Y. 2013) (summarizing “uncontested statistics” indicating that the New York City Police Department made 4.4 million stops in an eight-year period from 2004 to 2012 and that over 80% of these stops were of racial minorities). Similarly, in Ferguson, Missouri, from 2012 to 2014, “African Americans account[ed] for 85% of vehicle stops, 90% of citations, and 93% of arrests made by [Ferguson Police Department] officers, despite comprising only 67% of Ferguson’s population.” 23 Civ. Rts. Div., U.S. Dep’t of Just., Investigation of the Ferguson Police Department 4, 62 (2015) (often called the Ferguson Report). The police disproportionately used force against African Americans, with “[n]early 90% of documented force used by FPD officers” being directed at African Americans. 24 Id. at 5, 62. And in Los Angeles, data from July 2018 to April 2019 showed that Black and Latino drivers were far more likely to be subjected to searches during traffic stops than whites. 25 Ben Poston & Cindy Chang, LAPD Searches Blacks and Latinos More. But They’re Less Likely to Have Contraband Than Whites, L.A. Times (Oct. 8, 2019, 3:52 PM PT), (to locate, click “View the live page”).

Beyond the empirical evidence, interview-based accounts also provide a window into how police wield their discretion. The Ferguson Report, for instance, is riddled with interviews that document how officers singled out vulnerable residents for harassment. In one case, a police officer accused a thirty-two year old Black man of being a pedophile for no reason, demanded to see his identification, and arrested him for “making a false declaration” for stating his name was “Mike” instead of “Michael.” 26 Civ. Rts. Div., U.S. Dep’t of Just., supra note 23, at 3, 18 (capitalization altered). The Ferguson Report showed how much power police have to control the narrative that emerges from police–resident encounters. But for the Department of Justice investigation, for instance, “Mike” would almost certainly not have had the opportunity to explain how he was falsely accused and arrested. 27 Cf. Alexandra Natapoff, Speechless: The Silencing of Criminal Defendants, 80 N.Y.U. L. Rev. 1458-75 (2005) (explaining how defendant silencing in built into criminal law enforcement).

The interviews also offer further evidence that the problem of policing goes beyond the most visible instances of violence. Rather, visible instances of violence are reflective of systemic problems in policing and the failure of police to serve the communities that they are supposed to protect.

II. The Need to Connect Race-Based Policing with Criminal Records

Even as recent social movements have raised awareness of unjustified policing practices, many people remain unaware that arrests alone create damaging criminal records. 28 For selected scholarship discussing the impact of criminal records, including arrest records, see, for example, Wayne A. Logan & Andrew Guthrie Ferguson, Policing Criminal Justice Data, 101 Minn. L. Rev. 541, 556 (2016) (discussing how “individual cases are dutifully recorded and memorialized” in a way that “would not have been possible or useful” prior to technological changes); Anna Roberts, Arrests as Guilt, 70 Ala. L. Rev. 987, 989 (2019); and Kate Levine, Discipline and Policing, 68 Duke L.J. 839, 890 (2019). One reason for the disconnect may be the relative invisibility of arrest records. As I have discussed elsewhere, arrest records can trigger a host of consequences, such as immigration detention and deportation, eviction from public housing, or the suspension of professional licenses. 29 See generally Jain, Arrests, supra note 3. These processes are systemic—they are the product of governmental decisions to invest time and resources to create, disseminate, and use criminal records. But the consequences of an arrest may not occur until a later point in time, and they may never be discussed by a prosecutor, defense lawyer, or judge.

Another reason for the disconnect may be the widely employed framework of collateral consequences of criminal conviction. The collateral-consequences framework has done much to advance our collective understanding that penalties like deportation can function as a “secret sentence.” 30 Gabriel J. Chin & Richard W. Holmes, Jr., Effective Assistance of Counsel and the Consequences of Guilty Pleas, 87 Cornell L. Rev. 697, 700 (2002). But the term itself is amorphous. As Jenny Roberts has argued, the doctrinal divide between “direct” consequences and “collateral” consequences is “mythical,” because “even consequences that seem to go to the heart of criminal punishment” such as total prison time served may be deemed collateral. 31 Jenny Roberts, The Mythical Divide Between Collateral and Direct Consequences of Criminal Convictions: Involuntary Commitment of “Sexually Violent Predators, 93 Minn. L. Rev. 670, 679-80 (2008) (arguing that the direct–collateral consequences distinction does not track penal versus non-penal sanctions, because, for instance, “in some circuits a defendant may be sentenced in a federal criminal case without knowing that his federal sentence will not begin until he has finished serving a state sentence” and that “[t]he fact that the defendant will thus serve more prison time on the two cases than he expected when he pleaded guilty is deemed ‘collateral’” while “[i]n other circuits, it is considered ‘direct’”). Some commentators have used the term “informal” consequences or “invisible punishments” rather than the term “collateral,” in order to signal that civil penalties are just as weighty as the formal sentence. 32 Michael Pinard, Criminal Records, Race and Redemption, 16 N.Y.U. J. Legis. & Pub. Pol’y 963, 970 (2013) (using the term “informal, non-legal consequences”); Wayne A. Logan, Informal Collateral Consequences, 88 Wash. L. Rev. 1103, 1104 (2013) (describing informal consequences as “arising independently of specific legal authority” as opposed to formal collateral consequences); McGregor Smyth, Holistic Is Not a Bad Word: A Criminal Defense Attorney’s Guide to Using Invisible Punishments as an Advocacy Strategy, 36 U. Tol. L. Rev. 479, 494-96 (2005) (using the term “invisible punishments”). Commentators also employ the term collateral consequences in different ways. Some refer to government-imposed penalties such as voting bans. Others employ the term more broadly so as to encompass the barriers to finding work that job seekers experience as a result of their criminal records. 33 See, e.g., Benjamin Levin, Criminal Employment Law, 39 Cardozo L. Rev. 2265, 2268, 2273-74 (2018) (explaining how barriers to finding work function as a collateral consequence, including when imposed by private employers).

Precisely because the problem of criminal records is multifaceted, reforms that aim to address one aspect of the problem may obscure others. Focusing on collateral consequences of convictions risks obscuring that the problem is not limited to conviction. If we overlook the role of arrests, we perpetuate an idealized but ultimately inaccurate narrative about how the criminal process operates. In this idealized narrative, lawyers can identify particular laws—such as those that disenfranchise felons or that require noncitizens to be deported after a criminal conviction—and determine whether a criminal conviction will trigger a particular consequence. But in practice, penalties stemming from a criminal record are ubiquitous. Those with criminal records may never realize the full impact of their contact with the criminal justice system, because they have no ability to track how their record is used.

Relatedly, a focus on forgiveness or redemption in policy reform may obscure the role of race-based policing as the primary reason why some criminal records are created in the first place. During the Obama Administration, Housing and Urban Development Secretary Shaun Donovan put it this way in an appeal to housing providers to employ discretion in favor of renting to tenants with criminal records: “[T]his is an Administration that believes in the importance of second chances—that people who have paid their debt to society deserve the opportunity to become productive citizens and caring parents, to set the past aside and embrace the future.” 34 Letter from Shaun Donovan, Sec’y, U.S. Dep’t of Hous. & Urb. Dev.; and Sandra B. Henriquez, Assistant Sec’y for Pub. & Indian Hous., U.S. Dep’t of Hous. & Urban. Dev., to PHA Executive Director 2 (June 17, 2011), The language of “second chances” for those who have already “paid a debt to society” does not account for those for whom the criminal record itself is unjustified. The issue is not simply that those with criminal records deserve a second chance. It is that, due to unjustified underlying policing practices, some arrested individuals never had a first chance.

A focus on the impact of a criminal record also invites a particular set of comparisons between those with criminal records and those without them. But it is important to recognize that even among equivalent record-holders, racial minorities may experience worse outcomes. It is not just that overpoliced racial minorities are disproportionately likely to have criminal records. Rather, the mark of a criminal record itself may be more harmful for racial minorities.

In the context of employment, studies have shown how criminal records can amplify underlying race-based discrimination. In a seminal 2001 study, sociologist Devah Pager found that criminal records disproportionately blocked applicants from obtaining access to entry-level jobs across the board, but that they had more of an impact on Black applicants. 35 Pager, Marked, supra note 4, at 91. Pager matched Black and white testers with comparable qualifications for entry-level jobs, and she examined how both race and a criminal record—a nonviolent drug conviction and eighteen months of incarceration—affected who received a callback interview. 36 Pager sought to minimize the impact of soft characteristics like personality by measuring the likelihood of receiving a callback, rather than the likelihood of actually getting the job. Id. at 61 (explaining the “focus only on this initial stage of the employment process”). The study showed cascading levels of race-based discrimination. First, when designing the study, Black testers had to complete 200 audits to receive a number of callbacks comparable to that white testers received after 150 audits. 37 Id. at 60. And while all applicants experienced worse job prospects because of a criminal record, white testers fared better than similarly situated Black testers. 38 Id. at 69, 90-91. Ultimately, Black applicants without a criminal record had the same employment prospects as white applicants with a criminal record. 39 Id. at 90-91.

Pager’s study also showed that employers did not explain how the presence of a criminal record affected their decisions. For instance, employers would claim that they would be in touch after doing a reference check, but they would never actually check the reference. 40 Id. at 66. Applicants who took employers at their word might infer that they were not given the job because of a poor reference check, rather than recognizing that the problem was the criminal record.

An important 2014 study of “ban the box” reforms by Amanda Agan and Sonja Starr also showed how racial discrimination, combined with the use of criminal records, forecloses job opportunities. 41 Amanda Agan & Sonja Starr, Ban the Box, Criminal Records, and Racial Discrimination: A Field Experiment, 133 Q. J. Econ. 191 (2018). The study examined racial discrimination in entry-level jobs before and after legislation was rolled out in New York and New Jersey that barred employers from asking about criminal history in an initial employment application. Prior to ban-the-box reforms, “white applicants received 7% more callbacks than similar black applicants, but after [those reforms] this gap grew to 43%.” 42 Id. at 195. The results showed that ban-the-box reforms appeared to help white applicants with criminal records, who “saw a substantial increase in callbacks.” 43 Id. at 195. But they had the opposite effect on Black applicants, who experienced a “substantial drop” in callbacks. 44 Id. As Agan and Starr discuss, this pattern suggests that employers stereotyped Black applicants as having criminal records when those employers were barred from asking about records on the initial application form. 45 Id. (“This pattern suggests that when employers lack individualized information, they tend to generalize that black applicants, but not white applicants, are likely to have records.”).

These studies show how racial discrimination magnifies the impact of a criminal record. The Fair Housing Act and Title VII of the Civil Rights Act of 1964 protect against racial discrimination in housing and employment, respectively, including disparate impact and disparate treatment. 46 Fair Housing Act, 42 U.S.C. §§2604-3606; Title VII of the Civil Rights Act of 1964, 42 U.S.C. §§2000e to 2000e-17. Courts have held that that a blanket prohibition on renting to or employing those with criminal records can constitute unlawful discrimination. 47 See, e.g., Smith v. Fussenich, 440 F. Supp. 1077, 1078, 1080-81 (D. Conn. 1977) (finding an equal protection violation where a Connecticut statute barred those with felony convictions from working as licensed detectives, because the statute was overly broad, made no distinction between different felons, and made an irrational distinction between convicted felons and convicted misdemeanants); see also Dallan F. Flake, When Any Sentence Is a Life Sentence: Employment Discrimination Against Ex-offenders, 93 Wash. U. L. Rev. 45, 70-71 (2015) (discussing constitutional and statutory doctrines related to employment bans based on criminal records). And where employers or housing providers have stated policies regarding criminal records, those policies need to serve a legitimate purpose. As the Third Circuit held in a widely cited decision, El v. Southeastern Pennsylvania Transportation Authority, Title VII “require[s] that the [hiring] policy under review accurately distinguish between applicants that pose an unacceptable level of risk [because of their criminal record] and those that do not.” 48 El v. Se. Pa. Transp. Auth. (SEPTA), 479 F.3d 232, 244-45 (3d Cir. 2007). At issue in El was whether a formal policy that “disallowed hiring anyone with, among other things, a violent criminal conviction,” discriminated against an African American employee with a forty-seven-year-old homicide conviction that had been entered when he was fifteen. 49 Id. at 235-37. The plaintiff had been terminated shortly after his employer did a background check; the timing of the termination decision, and the employer’s explanation that the termination was based on the background check, gave him the information necessary to file a complaint alleging unlawful discrimination. 50 Id. at 235-36 (noting that the offer was contingent on a successful criminal background check). But all too often, prospective tenants or applicants are kept in the dark about the impact of their records, especially when they are denied access to a job or rental housing in the first place. Employers and housing providers who conduct criminal record checks are generally under no obligation to explain how a criminal record affected their decisionmaking, nor are they required to adopt and publish criminal-record policies. In the absence of visibility into how criminal records are used, we lack even the most basic tools to surface, and ultimately to redress, race-based discrimination.

III. Recognizing the Negative Credential of a Criminal Record

A racial reckoning in policing requires connecting unjustified arrests to the creation of lasting criminal records. A marking or credentialing framework offers a way to conceptualize how criminal records both magnify and conceal race-based discrimination. One way of recognizing the work that the negative credential of an arrest is doing is to disaggregate the criminal arrest and record-creation process into four stages: (1) the physical restraint on freedom (the arrest for Fourth Amendment purposes), (2) the creation of a criminal record (the credentialing process), (3) the dissemination of criminal records (dissemination of the credential), and (4) adverse decisions based on criminal records (the impact of the credential). Reckoning with race in the criminal justice system requires recognizing that the problem is not just the police: It is with a legal regime that entrenches racial subordination through criminal records.

We need a new shared vocabulary about what an arrest is today. An arrest is something more permanent and more consequential than it has been in the past. For Fourth Amendment purposes, courts have at times relied on historic arrest practices to justify contemporary arrest practices. In Atwater, the Court drew an analogy between a custodial arrest for a misdemeanor seatbelt violation and historical practices of making low-level warrantless arrests. The Court likened an arrest for a seatbelt violation, for instance, to a sixteenth-century statute that “authorized peace officers to arrest persons playing ‘unlawful game[s]’ like bowling, tennis, dice, and cards.” 51 Atwater v. City of Lago Vista, 532 U.S. 318, 334-35 (2001) (alteration in original) (quoting An Acte for Mayntenance of Artyllarie and Debarringe of Unlawful Games, 33 Hen. 8 c. 9, §§ 11-16 (1541)). The Court observed that “early- and mid- 19th-century decisions expressly sustain[ed] . . . laws authorizing peace officers to make warrantless arrests for misdemeanors not involving any breach of the peace.” 52 Id. at 342. The Court’s reasoning implies that low-level arrests today are equivalent to low-level arrests that were historically made for crimes such as “bowling, tennis, dice and cards.”

This analogy is not sustainable if criminal records are taken into account. In the past, a night watchman’s decision to make an arrest for a low-level activity involved a government-imposed restraint on freedom and had the potential for stigma. But today, that stigma is formalized through the creation and dissemination of a criminal record. Even if the restraint on freedom is the same, the overall impact of the arrest is weightier because of the record.

Common sense and political accountability cannot work as checks on how criminal records are created if police, prosecutors, and lawmakers do not look at the consequences of a criminal record—or if they lack a way to identify what the full consequences of a record might be. In cases where an arrest record has more to do with biases in policing than underlying moral blameworthiness, the question should be why this credential is being created in the first place.

The simplest way to avoid creating unjustified marks is to reduce reliance on criminal arrests themselves, such as by decriminalizing conduct that does not reflect underlying moral culpability or by reducing the scope of policing practices. 53 Rachel A. Harmon, Why Arrest?, 115 Mich. L. Rev. 307, 320 (2016) (“Given that arrests are costly deprivations of liberty, they should, as a normative matter, be imposed only when they serve a significant state interest, when the risk of harm is not grossly disproportionate to that interest, and when they are used fairly.”). This approach dovetails with other proposals to fundamentally change policing. 54 This argument appears in the literature on over-criminalization, as well as in literature on defunding the police. See Jessica M. Eaglin, To “Defund” the Police, 73 Stan. L. Rev. Online 120 (2021). Reducing criminal arrest practices altogether offers a way to be more nuanced about the types of credentials the criminal justice system creates.

In addition, we should consider how to sever the process of criminal arrest from the creation of a criminal record. The progressive prosecution movement is often associated with prosecutors who decline to pursue charges for low-level arrests that disproportionately burden communities of color. Prosecutors who choose not to pursue charges can avoid creating the mark of a conviction, but they cannot prevent the creation of a negative credential altogether. Structural changes in when prosecutors engage in screening might prevent arrests from turning into criminal records. Adam Gershowitz has argued for adopting this model, advocating for early prosecutorial screening, prior to the booking process. 55 Adam M. Gershowitz, Justice on the Line: Prosecutorial Screening Before Arrest, 2019 U. Ill. L. Rev. 833, 837, 869 (pointing to Harris County, Texas as a case study and arguing that this model would be feasible for other jurisdictions). See also Ingrid V. Eagly, Criminal Justice for Noncitizens: An Analysis of Variation in Local Enforcement, 88 N.Y.U. L. Rev. 1126, 1172 (2013) (discussing the “direct-filing” model in Harris County, Texas). This type of intervention offers a way to disaggregate the physical experience of being arrested from the creation of a criminal record. Although much depends on how prosecutors exercise their discretion, taking a prebooking approach to prosecutorial screening could prevent unjustified or overbroad arrests from being formalized into a criminal record.

In addition to restricting when the mark of a criminal record is created, the government should reduce the visibility of such marks. Statutes that prohibit sharing of arrest records, or that provide for automatic expungement of criminal records, have the potential to mitigate the impact of a criminal record. But these types of reforms have been underutilized because of unnecessary eligibility requirements. 56 See 50-State Comparison: Expungement, Sealing & Other Record Relief, Collateral Consequences Res. Ctr. (updated Mar. 2021), (offering a fifty-state comparison of requirements for criminal record relief). When those with criminal records are required to affirmatively seek relief, there is a significant “uptake gap.” 57 J.J. Prescott & Sonja B. Starr, Expungement of Criminal Convictions: An Empirical Study, 133 Harv. L. Rev. 2460, 2466, 2501-10 (2020) (determining that “only 6.5% of all eligible individuals [in Michigan] receive an expungement within five years of the date at which they first qualify for one”). As Colleen Chien has shown, “the informational and bureaucratic hoops that one must jump through to get their second chances are often extensive and costly,” which results in an enormous gap between those who are eligible for relief such as expungement and those who actually apply for and receive such relief. 58 Colleen Chien, America’s Paper Prisons: The Second Chance Gap, 119 Mich. L. Rev. 519, 540-41 (2020). Less than 20% of those who are eligible for expungement of convictions have applied for relief. 59 Id. at 555. Rather than asking the record holder to present grounds for relief, the primary question should be whether there is a good reason for making the mark visible in the first place. 60 See Jain, Proportionality, supra note 3, at 978 (arguing for examining “whether there is a compelling penal rationale for the penalty, rather than whether the defendant can demonstrate hardship or good character”); see also Joy Radice, The Reintegrative State, 66 Emory L.J. 1315, 1386-87 (2017) (discussing how “the criminal justice system’s discretionary nature has been linked historically to its disproportionate impact on poor people of color” and arguing for making relief mechanisms automatic rather than discretionary).

In addition, reckoning with racial discrimination in policing requires acknowledging that the problem is not just with the police. It is with a host of actors that rely on criminal records in discriminatory ways. Employers who support police reform—who believe that policing practices reflect race-based discrimination—should consider how criminal records close the door to work along racial lines. They should also be willing to adopt written policies explaining how they employ records, and they should be transparent about information such as whether they check records, why they do so, and how criminal records affect hiring practices.

Decisions about how and when to rely on criminal records, however, should not be left entirely to the discretion of employers or housing providers. In policing, as well as in a host of other areas, discretionary decisions are rife with the potential for discrimination. In a 2002 decision, Department of Housing & Urban Development v. Rucker, the Supreme Court erred in conceptualizing discretionary decisionmaking as the solution to overbroad uses of criminal records. 61 Dep’t of Hous. & Urb. Dev. v. Rucker, 535 U.S. 125, 130 (2002). In Rucker,the Court upheld the eviction of Perlie Rucker, her daughter, two grandchildren, and great-grandchildren after the daughter’s off-premises drug-related arrest. 62 See Rucker v. Davis, No. C 98-00781, 1998 WL 345403, at *2 (N.D. Cal. June 19, 1998). Rucker argued that her eviction violated her due-process rights given that she regularly searched her daughter’s room for evidence of drugs and had no ability to control or prevent the conduct of her daughter well outside her apartment. 63 Id.; see also Jain, Arrests, supra note 3, at 836-37 (discussing Rucker). The Court unanimously held that there was no due-process violation, explaining that Rucker had been granted proper notice of the eviction and that the housing authority was properly “acting as a landlord of property that it owns, invoking a clause in a lease to which respondents have agreed and which Congress has expressly required.” 64 Rucker, 535 U.S. at 135-36.

A marking framework offers a way of recognizing why the arrests of Rucker’s adult daughter, Gelinda, and her adult son, Michael (who listed Rucker’s address as his own, even though he did not live with her) triggered the eviction. 65 Joint Appendix at 13, Rucker, 535 U.S. 125 (Nos. 00-1770, 00-1781), 2001 WL 34093958. Rucker’s eviction was not a collateral consequence of a criminal conviction—the eviction was discretionary and could be triggered by arrest alone. In Rucker, the key question should not have been whether discretion was available, but how police and the housing authority had employed their discretion. Rucker’s complaint showed the reach of low-level police arrest activity. According to the complaint, Gelinda was arrested three blocks from Rucker’s residence after she was observed committing an open container violation. 66 Id. Michael, Rucker’s son, was arrested approximately six months later when a police officer observed him “loitering at a bus stop with a second man.” 67 Id. Searches of Gelinda and Michael revealed that each was carrying a rock of cocaine. 68 Id. These facts, which were not noted in the Supreme Court’s decision, show how policing practices played a key role in triggering the eviction. In upholding the eviction, the Court referred to the need to give housing authorities the discretion to evict tenants who “cannot control drug crime, or other criminal activities by a household member which threaten [the] health or safety of other residents.” 69 Rucker, 535 U.S. at 133 (internal quotations omitted). The Court did not discuss the lack of safeguards in place to protect tenants, however, including in cases where there was little apparent link to the health or safety of other residents.

Rucker’s eviction was the product of a host of discretionary decisions, including a decision to arrest, to formalize the arrest into a credential, to notify the housing authority of the arrest, and ultimately to pursue eviction. Any discretion public-housing authorities ultimately exercised was responsive to policing practices.

Discretion cannot be a meaningful check against racial bias where there is significant information asymmetry about how criminal records are used. In the context of employment or housing, we know far too little about how criminal records relate to racial discrimination. Michael and Gelinda likely had no idea that their arrests would lead to their mother’s eviction months after the fact. Nor would most other arrested individuals in similar circumstances. Reconceptualizing the role of criminal records—including by restricting the use and dissemination of certain records; by engaging in robust enforcement of antidiscrimination law; and by rendering visible how employers, housing providers, and others rely on criminal records—should be part of the project of reckoning with the ongoing impact of policing.


As we collectively reckon with the role of race in the criminal justice system, it is critical to recognize how police wield enormous power to create the mark of a criminal record. Criminal records—their creation, distribution, and use—play a hidden role in formalizing social stigma. In a world where arrests create markers that last long after the criminal case is complete, the key question is why and how we permit race-based policing to live on and potentially to affect many aspects of an arrested individual’s life. Racial reckoning in criminal justice requires recognizing that the problem of policing is not confined to the police; a host of other actors rely on criminal records in ways that entrench inequality. A key question should be whether and when a police officer’s decision to arrest should automatically serve a credentialing function in the first place.

* Assistant Professor, University of North Carolina School of Law. This Essay was written for the 2021 Policing, Race, and Power Symposium hosted by the Stanford Law Review and the Stanford Black Law Students Association, and for the cross-journal Reckoning and Reformation Symposium. Thank you to Guy-Uriel Charles, Jessica Eaglin, Trevor Gardner, Ben Levin, Tracey Meares, Brian Murray, Anna Roberts, Aaron Tang, to the organizers of the Reckoning and Reformation Symposium, and to participants in a faculty workshop at Seton Hall Law School for helpful comments on earlier drafts. Haley Amster of the Stanford Law Review provided excellent editorial assistance.

The featured image “Courtroom One Gavel” by Joe Gratz is marked with CC0 1.0.