The Penology of Racial Innocence: The Erasure of Racism in the Study and Practice of Punishment


  • The authors wish to thank Alexes Harris, Michael McCann, Christopher Parker, Dara Strolovitch, Jack Turner, and four anonymous reviewers for their insightful and constructive feedback. The authors are listed in reverse alphabetical order; each contributed importantly to the development of this article. Please direct correspondence to Naomi Murakawa, Department of Political Science, University of Washington, Seattle, WA 98195-3530; e-mail:


In post–civil rights America, the ascendance of “law-and-order” politics and “postracial” ideology have given rise to what we call the penology of racial innocence. The penology of racial innocence is a framework for assessing the role of race in penal policies and institutions, one that begins with the presumption that criminal justice is innocent of racial power until proven otherwise. Countervailing sociolegal changes render this framework particularly problematic. On the one hand, the definition of racism has contracted in antidiscrimination law and in many social scientific studies of criminal justice, so that racism is defined narrowly as intentional and causally discrete harm. On the other hand, criminal justice institutions have expanded to affect historically unprecedented numbers of people of color, with penal policies broadening in ways that render the identification of racial intent and causation especially difficult. Analyses employing the penology of racial innocence examine the ever-expanding criminal justice system with limited definitions of racism, ultimately contributing to the erasure of racial power. Both racism and criminal justice operate in systemic and serpentine ways; our conceptual tools and methods, therefore, need to be equally systemic and capacious.

In the last third of the twentieth century, the definition of racism and the reach of the criminal justice system moved in opposite directions. On the one hand, the definition of legally actionable racism contracted as antidiscrimination law came to narrowly redress only racial inequality produced by intentional harms with discrete and identifiable causes (Crenshaw 1988; Crenshaw & Peller 1993; Freeman 1978; Haney López 2000; Selmi 1997). This contraction of antidiscrimination law is part of a larger shift in American politics in which the victories of the civil rights movement in the 1960s gave way to widespread belief in a “postracial” and “color-blind” America (Bobo et al. 1997; Bonilla-Silva 2001, 2006; Brown et al. 2003). Over the same time period, however, the scope of criminal justice expanded rapidly, as the rise of “law-and-order” politics produced broader police powers, more mandatory minimum sentences, the reinstatement of the death penalty, and, ultimately, the highest incarceration rate in the world.

These divergent trends produce an alarming paradox. In antidiscrimination law and the conventional wisdom of many whites, racism is waning, aberrant, and located in the bad intentions of individual actors. Yet in the lives of many people of color, criminal justice is expanding, commonplace, and located in systemwide penal policies and practices that are irreducible to bad individuals with evil intent. This post–civil rights paradox is sustained by what we call the penology of racial innocence: the study of punishment that obscures the operation of racial power in penal practices and institutions. Criminal justice research has too often adopted the narrow standards of contemporary antidiscrimination law. Like antidiscrimination law, conventional social science prioritizes identification of causal mechanisms, yet the statistical methods often used in this effort yield little insight into intentions and causation. Simultaneously, the practice of punishment has expanded in ways that make identification of racial intent and isolation of cause even more difficult. In short, the penology of racial innocence begins with presumptions of race-neutrality and adopts narrow definitions of racism, as well as data and methods often ill-suited to its analysis, even as the policies and practices of criminal justice expand in ever-more race-laden ways.

The penology of racial innocence is constituted by two divergences:

1. Divergent trends in racial “intent”: As antidiscrimination law has developed to demand proof of intent to discriminate on the part of specific and identifiable persons, criminal justice has expanded in ways that further obscure individual intentions.

Intentions are generally difficult to identify, but recent penal system expansion diffuses “intent” all the more. Proactive policing tactics based on officer assessment of “the totality of circumstances,” for instance, render officer intent largely unassailable; the proliferation of mandatory minimum statutes and sentencing guidelines displaces discretion from judges to prosecutors' offices, where it is much harder for researchers to obtain data; and expanded administrative control means that decisions to revoke parole and incarcerate violators are increasingly common and largely inscrutable. In short, there is a disconnect between narrowed standards of racial “intent” and the diffusion and expansion of discretionary power. Racial innocence is sustained by searching for racial intent, narrowly defined, in too few places, with too high a bar, and, most important, with a false distinction between the individuals guilty of racist intent and the presumably racially innocent institutions in which they are located.

2. Divergent trends in racial “causation”: As antidiscrimination law has developed to demand the disaggregation of decisionmaking points in order to identify “biased” actions occurring in a single moment, definitions of crime, criminal justice institutions, and discretionary authority have expanded, producing complex causal webs of racial inequality that compound over longer time horizons.

Racial antidiscrimination law has embraced a standard of aggressive disaggregation, eschewing outcomes of inequality as inconclusive and requiring evidence of bias at a discrete decisionmaking point. By contrast, criminal justice pathways have proliferated—with more “entry points” with protracted and indefinite “exit points”—so that seemingly small infractions and racial disparities accrete into lasting criminal justice entanglement characterized by large and often growing racial inequalities. In short, the legal evidentiary requirement of disaggregated racial “causation” is ill-suited to the reality of cumulative and compounding racial inequality through criminal justice expansion. Racial innocence is sustained by searching for racial “causation”—disaggregating the criminal justice system into discrete stages of criminal processing, isolating statistically significant disparities in one criminal justice institution from longer chains of cumulative racial inequality, and, most important, falsely separating isolated racial causes from presumably race-neutral social conditions and institutions.

Taken together, racial intent and racial causation are the high standards that must be met to disconfirm the default presumption of the penology of racial innocence—the presumption that there is no racism. In antidiscrimination law, the burden of proof is on the plaintiff to prove discrimination; similarly, in social science research, the null hypothesis is no race effect, no racial disparity, no racism. As was made clear in McCleskey v. Kemp (1987), the Supreme Court case upholding a capital sentence despite evidence of significant racial disparity, legal “analysis begins with the basic principle that a defendant who alleges an equal protection violation has the burden of proving ‘the existence of purposeful discrimination,’ ” as well as proving that “that the purposeful discrimination ‘had a discriminatory effect’ on him” (McCleskey v. Kemp 1987:293). This basic legal formulation—presumption of no racism, burden on plaintiff to prove both purposeful and directly consequential discrimination—is thus too often mirrored by social science research in criminal justice. The penology of racial innocence has become particularly entrenched, we believe, because the narrow legal standards of racism and the “gold standard” of social science have so much in common: Both emphasize narrow identification of causal mechanisms, and both employ conceptual and methodological tools that take a “snapshot” view rather than a moving picture of race's influence in criminal justice institutions (Pierson 2004).

In what follows, we describe the context and consequences of the penology of racial innocence, and we argue that the search for racial intent and racial causation ensures that racism will be underestimated. We begin by analyzing how the penology of racial innocence emerged under particular conditions of the post–civil rights context: namely, contracted definitions of racism and expanding criminal justice institutions. We then identify the distinguishing features of the penology of racial innocence, focusing on how the search for racial intent and racial causation constrains the study of punishment by isolating individuals from institutions, single incidents from social conditions, and discrete moments from longer time horizons of inequality. The final sections show how this framework fails to recognize the many complex ways racial power operates in and through the ever-expanding criminal justice system.

The Post–Civil Rights Context

The penology of racial innocence emerged under conditions specific to the post–civil rights context. One condition is that legal, political, and social definitions of racism have come to focus narrowly on individual and isolated racist transgressions. In legal terms, the years after Brown v. Board of Education (1954) have seen a narrowing of Supreme Court interpretations of the Equal Protection Clause (Crenshaw 1988; Freeman 1978; Selmi 1997).1 In political and social terms, racism has come to be defined as individual prejudice, often understood as personal sickness or irrationality (Frymer 2007; Frymer et al. 2006). American political and social sensibility has moved away from overt racism, as most Americans now reject the idea of biologically determined racial superiority (Bobo et al. 1997). As a result of the decline of overt racial animus, many assume that racism is now aberrant and exceptional. Post–civil rights racism—also defined as “laissez-faire racism” (Bobo et al. 1997) and “color-blind racism” (Bonilla-Silva 2001)2—is thus marked by the disavowal of systemic racism and the persistence of racial power and inequities; it retains the institutional and material manifestations of racial inequality even as overt expressions of white supremacy are deemed socially unacceptable.

The other condition is the expansion of criminal law and its administration. As is now well known, the U.S. incarceration rate has increased roughly six-fold since the 1970s to become the highest in the world (Western 2006). More than one in every 100 adult residents of the United States now lives behind bars (Pew Center on the States 2008:5). Blacks and Latinos each comprise less than 15 percent of the U.S. population, but were 40 and 20 percent (respectively) of the jail and prison population in 2008 (West & Sabol 2009: Table 16). The 2008 imprisonment rate for Latino men (1,200 per 100,000) was more than double that of white men (487 per 100,000), and the imprisonment rate for black men (3,161 per 100,000) was six times higher than that of white men Similarly, the 2008 incarceration rate for Latino women (75 per 100,000) was one-and-a-half times that of white women (50 per 100,000), and black women's incarceration rate (149 per 100,000) was three times that of white women (Sabol et al. 2009: Table 2).3 While racial disparities in incarceration rates have been fairly constant over time (Pettit & Western 2004), the ratio of black to white prison admissions increased from 2.1 in 1930 to 7.0 in 2000 (Oliver 2001:28). By 2004, nearly 60 percent of young black men without a high school degree had spent time behind prison bars (Pettit & Western 2004:161). Thus, blacks have become relatively more likely to experience repeated spells of incarceration; it is precisely this cycling in and out of prison that is thought to be most destructive for both individuals and communities (Clear 2007; Clear et al. 2001; Travis 2005; Urban Institute 2006).

Although the rapid growth of the prison population has understandably received the lion's share of social scientific attention, the entire criminal justice apparatus has expanded dramatically. Between 1980 and 2008, the total number of people under criminal justice supervision—which includes the incarcerated and those on probation and parole—jumped from roughly 2 million to more than 7 million people (Bureau of Justice Statistics n.d.a). Moreover, people of color are overrepresented among arrestees, jail inmates, and probationers as well as among prison inmates and parolees.

At the same time, the logic of penal control and surveillance has spread to noncriminal justice institutions. As Simon argues, the “technologies, discourses, and metaphors of crime and criminal justice” have permeated policies and institutions that are seemingly unrelated to crime-fighting (Simon 2007:4). Schools, for example, are “prisonized” by school-employed police officers, drug sweeps and K-9 units, metal detectors, zero-tolerance rules, and detention and expulsion (Simon 2007:222–6; see also Lyons & Drew 2006). Immigration practices and policies are also increasingly criminalized, as penal apparatuses developed through wars on crime, drugs, and terror are deployed against immigrants (Bohrman & Murakawa 2005; Dow 2004; Simon 2007). People of color are disproportionately impacted by the diffusion of these practices outside of penal institutions.

In short, “racism” and “discrimination” are nowhere in the post–civil rights era, yet punishment and surveillance are increasingly everywhere for blacks and Latinos. In this political context, the study of race in criminal justice institutions has become all the more pressing, and a great deal of scholarship is devoted to the question. Much of this scholarship sheds light on the many ways that race informs criminal justice expansion. Indeed, we draw upon these studies to buttress our empirical claims. At the same time, the study of criminal justice institutions—particularly its representation in the top criminological and sociological journals—has too often embraced narrow definitions of racism, even as criminal justice expands in ways that circumnavigate conventional social scientific analysis. The result is the penology of racial innocence.

Distinguishing Features of the Penology of Racial Innocence

Below, we highlight two features of the narrowing conception of racism, both of which have been influential in both antidiscrimination law and the study of punishment: intent and causation. The intent standard holds that only intentional bias on the part of an identifiable agent constitutes discrimination; the causation standard holds that the precise decision or behavior that caused the condition of inequality must be isolated and disaggregated from the totality of circumstances. Taken together, the twin standards of intent and causation constitute what has been called the perpetrator model of racism, which defines racism narrowly and sees discrimination from the perspective of the perpetrator rather than from the perspective of the victim (Freeman 1978). The following two sections show how the standards of intent and causation have influenced the study of criminal justice, and how the practices and policies of punishment expansion have grown to evade these narrow standards.

In its search for intent and causation, the penology of racial innocence neglects what we think of as the larger context and dimensions of racial power. Specifically, we view racial power as systemic, institutional, and long-standing; it is premised on ideologies and institutions that preserve white advantage, and it perpetuates ongoing patterns of undeserved enrichment and unjust impoverishment (Feagin 2000; Kim 2000). Racial power in criminal justice is systemic; some behaviors may be intentionally racist but many are not; some practices are easily identified as causes of inequality, but many operate through long and causally complex processes. In our view, then, racial power is not the sole province of white bigots to which people of color are subject, but rather a systemic and institutional phenomenon that reproduces racial inequality and the presumption of black and brown criminality. By contrast, the penology of racial innocence presumes systemwide racial innocence—that is, the criminal justice system is innocent of racial power until proven guilty of racial intent and causation.

The Standard of Racial Intent

Although antidiscrimination law is complex and internally contradictory, it is clear that the intent standard has become a common test of alleged violations of equal protection, and that intent has been narrowly construed.4 Case law reveals significant debate over the meaning of intent, and the differences between “discriminatory purpose” and the “foreseeable consequences” associated with the impact test. In the pivotal Washington v. Davis (1976), the Supreme Court held that the Equal Protection Clause of the Fourteenth Amendment only prohibits intentional discrimination. The Court clarified in subsequent cases that “‘discriminatory purpose’… implies that the decisionmaker, in this case a state legislature, selected or reaffirmed a particular course of action at least in part ‘because of,’ not merely ‘in spite of,’ its adverse effects upon an identifiable group” (Personnel Administrator of Massachusetts v. Feeney 1979 at 279, quoted in Selmi 1997:292). In McCleskey v. Kemp (1987), the Supreme Court cited earlier cases to affirm that “discriminatory purpose” implies more than awareness of consequences; it implies decisionmaking for the purpose of discrimination.5 Justice Lewis Powell, writing for the majority, found that statistics were insufficient proof of “purposeful discrimination” necessary to establish violation of the Equal Protection Clause; they were also insufficient to demonstrate the “constitutionally significant risk of racial bias affecting the … capital-sentencing process” impermissible under the cruel and unusual punishment clause (McCleskey v. Kemp 1987:292–3, 313–14). The legal standard of intent thus demands that discrimination must be purposeful (S. Johnson 1988; Selmi 1997).

The intent standard continues to limit legal challenges to racial inequality in criminal justice, as such challenges are typically framed as equal protection cases and have therefore been circumscribed by the legal requirement of purposeful discrimination. Although some creative attorneys are exploring legal challenges that do not require evidence of discriminatory intent, it is clear that the imperative to demonstrate conscious and willful bias on the part of criminal justice actors has been a significant barrier to efforts to challenge criminal justice practices and policies that produce racially disparate outcomes (Baldus et al. 2007; Provine 2007).

How Preoccupation With Racial Intent Constrains the Study of Punishment

The widespread adoption of the intent standard thus severely constrains legal challenges to racism in criminal justice. We argue below that this standard has also been adopted in too much of the social scientific literature. Indeed, some influential scholarship on racism in criminal justice accepts the legal standard of intent by limiting its focus to intentional racial bias at a discrete moment in criminal justice decisionmaking and case processing. As many other scholars have noted, however, framing racism as intentional harm perpetrated at a discrete moment in time does not capture all of the ways race shapes penal beliefs, practices, and outcomes. An emerging body of scholarship on “implicit bias,” for example, notes that racial beliefs may be unconscious or “commonsensical” rather than intentional, but they nonetheless contribute to racial inequality (Eberhardt & Goff 2005; Eberhardt et al. 2004; S. Johnson 1988; Provine 2007). Others emphasize the systemic, institutional, and intersectional nature of racism (C. Cohen 1999; Feagin 2000, 2006; Frymer 2007; Haney López 2000; Strolovitch 2007).

By contrast, some researchers adopt the intent standard wholesale. In one example, researchers adopting this approach state, “The question is whether the overrepresentation of blacks in prison admissions is the result of proportionately more blacks than whites committing crimes (‘differential involvement’) or whether it is the result of racial discrimination in the administration of justice (‘racial discrimination’)” (Langan 1985:666). In this framework, evidence of racial discrimination is assessed narrowly, typically by comparing the racial composition of those arrested for particular crimes with the racial composition of those imprisoned for those same crimes (for other examples of this methodology, see Austin & Allen 2000; Blumstein 1982, 1993; Langan 1985; Tonry 1995). Findings suggest that 80–90 percent of race differences in imprisonment for (some) violent offenses are explained by arrest patterns, leading researchers to conclude that “racial differences in arrests alone account for the bulk of the racial differences in incarceration” (Blumstein 1982:1268). This research design assumes that arrests are an accurate measure of criminal behavior. To buttress this assumption, researchers draw on studies that compare victims' descriptions of their assailants with the racial composition of those arrested for violent crimes (see Blumstein 1993; D'Alessio & Stolzenberg 2003; Hindelang 1978; Langan 1985; Tonry 1995).6 Studies adopting this approach can only include crimes that involve direct victims who are likely to have had a close look at the perpetrator, and assume that victims' racial identifications are accurate. The results indicate that the racial composition of those arrested for these offenses closely approximates the racial composition of victims' descriptions of their assailants, leading researchers to conclude that the criminal justice system is not discriminatory.

The focus on (some) violent crime in these studies is an important limitation. Arrests for serious violent crimes make up a tiny fraction of all arrests; the overwhelming majority involve allegations of less serious wrongdoing. In 2008, for example, only 4.2 percent of all arrests involved allegations of violent index crimes (FBI 2009: Table 29). There is evidence that the policing of more minor crimes (including drug offenses) is far more discretionary and hence more likely to be racially skewed than the policing of violent crimes (Beckett, Nyrop, Pfingst, & Bowen 2005, Beckett, Nyrop, & Pfingst 2006; DeFleur 1975; Fagan & Davies 2000). Studies comparing victims' descriptions of their assailants with arrest data thus focus on a very small proportion of arrests and omit those categories in which racialized discretion is likely more pronounced. This selection bias is magnified by the fact that race appears to play a more important role in the post-arrest processing of less serious criminal cases. For example, Blumstein (1995) finds that only half (51 percent) of racial disproportionality in drug-related imprisonment is explained by drug arrests. In short, this methodological approach requires a narrow empirical focus and ignores the majority of criminal cases in which race is more likely to matter.

More broadly, this way of framing the question of racism in criminal justice—forced choice between disproportionate black offending or biased administration of justice—replicates the emphasis on individual bias/intent as seen in post–civil rights antidiscrimination law. In our view, this emphasis on intent erases the complex ways that race constitutes social relations and processes. For example, even if it were true that observed race differences in prison admission for violent crimes solely reflected race differences in rates of offending, the idea that differential rates of interpersonal violence are racially innocent is objectionable. Indeed, this construction masks the racialized nature of the social conditions that give rise to higher levels of violence in poor urban communities of color.

Comparative studies find a strong positive correlation between the homicide rate and various measures of economic inequality and this correlation is especially strong in democracies and wealthier countries (Krahn et al. 1986). Moreover, the association of economic inequality and lethal violence is particularly strong in countries that practice “deliberate, invidious exclusion” on the basis of ascribed characteristics such as race (Messner 1989). In the United States, residents of “high-poverty areas”—areas inhabited by high numbers of poor people—are overwhelmingly black and Latino. Nationwide, nearly seven out of eight persons living in a high-poverty urban area are of color (Beckett & Sasson 2004:36). As Wacquant (2007) notes, American urban poverty is pre-eminently a racial poverty, rooted in the ghetto as a specific social form and mechanism of racial domination. The sharp increase in joblessness among black males that resulted from deindustrialization reduced the pool of “marriageable men,” which in turn contributed to the proliferation of single-parent households and intensification of inner-city poverty (Wilson 1990). The increase in single-parent families and destabilization of communities caused by deindustrialization has had significant consequences for levels of interpersonal violence (Sampson 1987).7 In short, racial inequality and discrimination, both past and present, explain comparatively high levels of interpersonal violence in poor urban communities of color. Juxtaposing ostensibly race-neutral crime rates against racialized criminal justice administration—without referencing the racialized social conditions that fuel interpersonal violence—masks the historical, social, and political processes that explain the racially uneven distribution of interpersonal violence.

In some cases, analysts adopt a version of the intent standard and do find that the evidence meets that standard. Tonry (1995), for example, examines racism in criminal justice from the perspective of mens rea in criminal law, in which purpose and knowledge are equally culpable states of mind. From here, Tonry examines whether architects of the drug war knew that their policies would have a disproportionate impact on communities of color. Based on his findings, Tonry argues that Reagan-era officials initiated the war on drugs when they did in order to capitalize on waning social tolerance for drug use and to win political favor, even when “they knew that the war on drugs would be fought mainly in the minority areas” (Tonry 1995:96, 104). Tonry's adoption of this high bar for establishing the existence of discrimination—knowledge and purpose—means that his conclusions have particular rhetorical power. And although his case is persuasive, his reliance upon the intent standard for evaluating the presence of racial discrimination runs the risk of further entrenching this narrow definition of racism. In this framework, racial wrongdoing is confirmed by evidence of deceptive individuals with bad intentions; blameworthy bias is present when officials know that blacks will be harmed but proceed anyway. This conception neglects the routine ways that crime policy is racialized, even without liars or knowing conspirators.

In sum, the intent standard constrains social science research on racism in criminal justice. To the extent that it, like antidiscrimination law, requires identification of intent, the study of criminal justice ultimately treats “bias” as a deviation from an otherwise fair criminal justice system operating in a race-neutral context. This conceptual approach obscures the role of race in the U.S. stratification system, the construction of particular issues as crime problems, and in shaping the current propensity to rely on coercive social control mechanisms to solve those problems. As Mauer puts it, race shapes “both the means by which we choose to respond to crime problems and the vigor with which we do so” (2004:80). The discrimination paradigm ignores the very real possibility that race matters most by shaping how much researchers attend to a particular social problem, whether we choose to define it as a criminal problem, and the kinds of policies we adopt to “solve” it.

Punishment Expansion and the Diffusion of Intent

We argue above that the study of punishment has been constrained by the emphasis on intent—searching for narrow decisions informed by intentional harm—while criminal justice has expanded in ways that evade easy detection of intent. In particular, the rise of “law-and-order” politics and policies entails greater discretionary power for police and prosecutors, whose decisions are comparatively unchecked and understudied; more administrative punishments, such as revocation of parole, that are also difficult to subject to critical scrutiny; and the widespread influence of race in shaping ever-growing perceptions of danger and policies of criminalization. Each of these developments, described below, renders the social scientific search for intent especially problematic.

Diffusion and Expansion of Discretionary Power

While the intent standard requires identification of intentional and purposeful bias on the part of an individual actor in a single decisionmaking process, criminal justice expansion entails the diffusion of discretion across multiple agents and a loosening of the criteria that are meant to shape its expression. In short, legal and policy shifts have authorized the police and prosecutors to enjoy broad and largely inscrutable discretion.

Recent decades have witnessed a concerted effort to enhance police discretion and legitimate their focus on less serious offenses. Although policing necessarily and inevitably involves discretion, the extent to which this is the case varies. In the 1960s and 1970s, legal authorities expressed an increased deal of concern about unfettered police discretion. For example, in one key Supreme Court case, Papachristou v. City of Jacksonville (1972), a unanimous court ruled that a Florida loitering law “furnishes a convenient tool for ‘harsh and discriminatory enforcement by local prosecuting officials, against particular groups deemed to merit their displeasure.’ It results in a regime in which the poor and the unpopular are permitted to ‘stand on a public sidewalk … only at the whim of any police officer’” (Papachristou v. City of Jacksonville 1972:170; quoting Shuttlesworth v. City of Birmingham 1965:90 and Thornhill v. Alabama 1940:97–8; see also Stewart 1998 for thoughtful analysis).

In the intervening decades, however, the Court's efforts to extend rights protections and circumscribe police discretion have been undermined by a number of developments, perhaps most profoundly by the war on drugs. By shifting attention and resources to the antidrug campaign, police organizations came to focus on behavior that is far more common than, say, armed robbery. Indeed, the enforcement of laws prohibiting the possession and distribution of illicit drugs is necessarily highly discretionary: Only a very small percentage of the roughly 25 million Americans who report using illegal drugs every year will be arrested for doing so. As with other crimes involving consensual parties, proactive tactics are required to secure arrests. The police enjoy nearly unlimited discretion in deciding where to look for drug offenders and against whom to use proactive tactics (Duster 1997). Reasonable suspicion can be based upon “the totality of circumstances,” as the Supreme Court held in United States v. Cortez (1981; Davis 1997). And as is now well known, the war on drugs is one of the most important causes of rising levels of racial inequality in arrests and prison admissions (Beckett, Nyrop, Pfingst, & Bowen 2005; Beckett, Nyrop, & Pfingst 2006; Provine 2007).

More recently, advocates of “broken windows policing” have offered another justification for enhanced police discretion and the focus on nonviolent crimes. Proponents of this philosophy argue that neighborhoods that fail to fix broken windows or address other manifestations of “disorder” display a lack of informal social control, thus inviting serious criminals into the neighborhood (Kelling & Coles 1996; Wilson & Kelling 1982). Advocates of broken windows policing therefore call for a fundamental reorientation of policing, one that offers city governments a broad and flexible means of regulating public spaces and removing those deemed “disorderly.” Despite compelling empirical challenges to the theory of crime that underpins broken windows (see Eck & Maguire 2000; Harcourt 2001; Harcourt & Ludwig 2006; Karmen 2000; Sampson & Raudenbush 1999; Taylor 2000), broken windows policing has become “common sense” and is now widely embraced by police departments across the United States (Herbert 2001; Herbert & Brown 2006) and in many other countries as well (Mitchell & Beckett 2008; N. Smith 2001; Wacquant 2003). Broken windows policing legitimates the police focus on “disorder” and enhances law enforcement's legal authority to do so. Moreover, the civility codes and their offspring provide the police with an important set of tools for general “order maintenance” and enable the police to make stops and conduct searches that they otherwise would not have legal authority to make (Beckett & Herbert 2008; Fagan & Davies 2000; Harcourt 2001).

These developments have been accompanied by court rulings that weaken Fourth Amendment protections. As Maclin (1993:202) notes, the purpose of the Fourth Amendment is to provide judicial checks on intrusive governmental investigations; it is anchored in “distrust of police power and discretion.” Yet these protections have been undermined as police authority and discretion have expanded. As one analyst of recent history of Supreme Court decisions concludes, there is only one unifying principles of those cases: “the police win” (Cloud 1993; see also Maclin 1993, Sklansky 1997). In these decisions, deference is granted to police power such that police discretion to investigate and arrest is significantly wider than during the Papachristou era (Dubber 2005).

In short, the expansion and legitimation of nearly unfettered police discretion renders the search for discriminatory intent especially difficult. Ironically, the adoption of sentencing guidelines and other sentencing schemes that structure judicial decisionmaking and limit judicial discretion also shifts discretionary power to an institutional location that is comparatively impervious to oversight (Davis 1998). Specifically, the adoption of sentencing guidelines shifts discretionary power from judges to prosecutors; social scientists call this the “hydraulic displacement” of discretion (Miethe 1987; see also Harris 2007; Tonry 1995). This institutional dynamic renders the search for racial intent difficult, as prosecutorial decisionmaking processes are highly influenced by case-specific factors and comparatively difficult to study.

Although understudied, prosecutorial discretion is enormously consequential (Davis 1998). For example, prosecutorial discretion plays an important role in determining which drug defendants will be tried in the federal system and which in the state system, a decision that has significant implications for sentencing outcomes. Similarly, prosecutors increasingly determine which juveniles to send to the adult system, which cases to “plead down” from felonies to misdemeanors, and for which capital defendants to seek death rather than life. These discretionary processes are difficult to study in part because they occur within (multiple) prosecutors' offices. The more general focus on decisionmaking within a particular segment of the criminal justice apparatus also leads researchers to ignore these decisionmaking processes. Indeed, most research strategies require an empirical focus on those clearly situated in a particular criminal justice subsystem such as the juvenile or superior courts. As a result, these studies ignore the key role of criminal justice actors in locating criminal defendants within the criminal justice system (but see Harris 2007).

Administrative Expansion

Above we argue that court rulings enabling police discretion based on “the totality of circumstances” and sentencing guidelines that displace decisionmaking to comparatively unchecked and understudied agents render the search for racial intent particularly fruitless. At the same time, mass incarceration has led to a remarkable expansion of probation and parole institutions whose work increasingly consists of detecting violations of the administrative rules that govern parolees and probationers (Simon 1993). These technical rule violations perpetuate prison and jail expansion. From 2000 to 2006, for example, 36–40 percent of state prison admissions involved those who were alleged to have violated parole conditions, up from 17 percent in 1980 (Beckett & Sasson 2004:189; Sabol & Couture 2008:5). Thus, criminal justice expansion tends to beget criminal justice expansion: As more people are swept into the increasingly capacious social control net, they are subject to enhanced regulation and surveillance, which produces and legitimates yet more penal intervention and racial inequality therein. Notably, the processes by which rule violators are administratively processed and sometimes reincarcerated are subject to little oversight or analysis.

Defining Danger and Disorder

The intent standard distinguishes the racial intentions of particular actors from larger and presumably race-neutral processes and preferences. This standard is homologous with the social scientific search for bias while it presumes that the larger context of penal expectations—what constitutes disorder, which behaviors are considered dangerous, and how government should respond—is race-neutral. The search for racial discrimination in the criminal justice system does not shed light on the many ways that race and ethnicity (along with gender and class) shape what we see and fear, how much we worry, and whether punitive policies are implemented in response to our concerns. The role of race in these complex social and political processes implicates actors both inside and outside the criminal justice system and is not captured by studies that search for evidence of intentional discrimination solely within the criminal justice system.

Over the course of the post–civil rights era, public perceptions of danger and policy constructions of the crime problem have been inextricably linked to perceptions and constructions of racial order (Beckett 1997; Fleury-Steiner et al. 2009; Murakawa 2008). For example, race shapes assessments of the severity of crime-related problems and perceptions of dangerousness. The percentage of young black men living in a neighborhood has a strong positive effect on perceptions of the severity of the crime problem in that neighborhood, even after crime rates and other relevant factors are taken into account (Quillian & Pager 2001). Resident perceptions of the level of neighborhood disorder are significantly affected by that neighborhood's racial, ethnic, and class composition (Sampson & Raudenbush 2004). Similarly, experimental researchers have found that respondents are more likely to incorrectly perceive that (virtual) blacks are holding guns and, as a result, to shoot (virtual) blacks than whites (see Correll et al. 2002; Greenwald et al. 2003). These and other studies indicate that race shapes perceptions of the severity and dangerousness of crime-related situations and problems.

Race also plays an important role in shaping support for coercive and “tough” solutions to crime. This appears to be true at both the individual and policy level. At the micro level, experimental studies indicate that the cultural association of blacks with crime (and welfare) has enhanced white support for more punitive anticrime (and antipoverty) measures.8 That is, when exposed to otherwise identical crime or poverty-related stories or scenarios, members of the public generally prefer “tougher” policy and legal responses to those social problems when perpetrators/recipients are depicted as black (Chiricos et al. 2004; Gilliam & Iyengar 2000; Gilliam et al. 2002; Iyengar 1995; Gilens 1995, 1996; Roberts & Stalans 1997). Indeed, one recent study found that the mere insertion of a racially charged phrase—“inner-city”—in a survey question about crime significantly enhanced support for prison expansion while decreasing support for more preventive anticrime measures (Hurwitz & Peffley 2005). The extent to which the crime problem is perceived to be a black one is thus a significant predictor of popular support for punitive policies.

In short, framing the question of racism in criminal justice as one of individual bias ignores how racism shapes the very process of identifying disorder and defining criminality. The history of drug control provides another telling example. Historians and social scientists have shown that drugs that are associated with racial and ethnic minorities are more likely to be defined as dangerous and to be the target of aggressive antidrug efforts (see Duster 1997; Lusane 1991; Manderson 1997; Morgan 1982; Musto 1973; Reinarman & Levine 1997). For example, in the early twentieth century, opium smoking—associated with Chinese immigrants—became the subject of an intense antidrug effort, while oral consumption of opium—widespread among Anglo Americans—was not considered to be a significant social problem (Morgan 1982; Musto 1973; Reinarman & Levine 1997). More recently, critics have argued that sentencing laws that single out crack offenders for harsher penalties rather than race-neutral policy considerations similarly reflect the association of crack with poor urban blacks (Beckett, Nyrop, Pfingst, & Bowen 2005; Beckett, Nyrop, & Pfingst 2006; Duster 1997; Manderson 1997; Musto 1973; Reinarman & Levine 1997; Steiner 2001; Tonry 1995; U.S. Sentencing Commission 2002). This literature thus indicates that drug policies and enforcement practices are influenced by the cultural construction and racial coding of drugs and those who ingest them; ostensibly race-neutral practices often reflect the association of certain substances or modes of ingestion with racially or ethnically stigmatized groups rather than public health or safety considerations.

In sum, the intent standard requires separating blameworthy, intentional bias from the “background” of race-neutral criminal justice. The irony is that criminal justice expansion itself has been constituted by, and is predicated on, the intensification of racially guided preferences and racially influenced policies—more discretionary actors, more self-reinforcing administrative punishments, perceptions, and policies linking racial order to “law and order”—so that, in effect, the “deviation” of intentional racial harm becomes less discernible as the “background” of race-laden penality becomes more widespread. The tendency to frame the question in terms of individual intent obscures the systemic, institutional, political, and cultural processes that have led blacks and Latinos to experience incarceration at unprecedented rates, and it contributes as well to the penology of racial innocence.

Searching for racial power rather than just racist “intent” would require systemic critique. The Supreme Court rejected McCleskey's claim of discrimination in part in order to preserve criminal justice as we know it; indeed, the Supreme Court claimed the challenge in its broadest form extends “to every actor in the Georgia capital sentencing process, from the prosecutor who sought the death penalty and the jury that imposed the sentence to the State itself that enacted the capital punishment statute” (McCleskey v. Kemp 1987:293). We agree; a broader definition of discrimination would require re-evaluation of the criminal justice system as we know it.

The Standard of Racial Causation

A second distinguishing feature of the penology of racial innocence is that the framework for identifying racism requires disaggregating racism into short and discrete causal moments, whereas recent penal expansion entails protracted and cumulative processes of racial inequality. While the standard of intent separates the handful of blameworthy racist actors from the innocent masses, the standard of causation separates discrete moments of transgression from the totality of conditions that perpetuate inequality (Freeman 1978). As such, the causation standard requires disaggregating long-term cumulative processes into short and isolated moments to identify the precise decision or behavior that “caused” the condition of inequality.

Richmond v. J. A. Croson Co. (1989) provides a notable example of the causation standard in antidiscrimination law. In this case, the Supreme Court struck down as unconstitutional the city of Richmond, Virginia's municipal policy of awarding 30 percent of all construction contracts to minority subcontractors. According to the Court, the evidence of discrimination presented—including the city's history of racial segregation, evidence that local contractors' associations have few minority members, and that Richmond is 50 percent black while less than 1 percent of Richmond's prime contracting dollars went to racial minorities—did not show “violation by anyone in the city's construction industry” (Richmond v. J. A. Croson Co. 1989:498–504). Instead, the Court suggested that evidence of discrimination must be isolated and distinct from the aggregate conditions of historical and present-day disparities, such as evidence that a particular minority business entrepreneur has suffered from a biased action from particular city or prime contractors. By contrast, Justice Thurgood Marshall's dissent in Richmond offers a powerful critique of disaggregation. Marshall criticized the majority for taking “the disingenuous approach of disaggregating Richmond's local evidence, attacking it piecemeal, and thereby concluding that no single piece of evidence adduced by the city, ‘standing alone,’… suffices to prove past discrimination.” Justice Marshall insisted that items of evidence do not “‘stand alone,’ or exist in alien opposition” (Richmond v. J. A. Croson Co. 1989:542; see also Crenshaw & Peller 1993; Rosenfeld 1989).

Indeed, disaggregation and decontextualization obscure the exercise of racial power. Consider, for example, that defense attorneys in the Rodney King trial showed frame-by-frame still photos of the beating, asking experts on prisoner restraint to identify the exact moment when King ceased to pose a threat, and the exact moment when force became excessive. Critical race scholars argue that this kind of disaggregation erases racial power by divorcing its effects from their social context and historical meaning (see Crenshaw & Peller 1993). Similarly, the widespread use of research techniques that disaggregated and decontextualized criminal justice decisionmaking—and the relatively paucity of efforts to reaggregate and recontextualize—means that social scientific research often contributes to the penology of racial innocence.

How the Effort to Isolate Racial Causation Constrains the Study of Punishment

Studies of judicial decisionmaking show how disaggregation and the interpretation of statistical evidence can contribute—unintentionally—to the penology of racial innocence. Studies of judicial discretion predominate in the criminological and sociological literatures on race and criminal justice processing; most of these studies are situated in jurisdictions governed by sentencing guidelines. Although many of these studies find evidence of “race effects” in judicial sentencing, there is also evidence that sentencing guidelines tend to reduce (but not eliminate) these effects. The social scientific focus on judicial sentencing in states with sentencing guidelines is thus the equivalent to the proverbial search for lost keys under the convenience of the light cast by the streetlamp.

Nonetheless, many studies do find evidence that race/ethnicity influence sentencing outcomes (Albonetti 1997; Bontrager et al. 2005; Bushway & Piehl 2001; Engen et al. 2003; B. Johnson 2006; Kramer & Steffensmeier 1993; Spohn & Holleran 2000; Steffensmeier & Demuth 2001; Ulmer & Kramer 1996; Wooldredge et al. 2005). In these studies, the emphasis is on identifying the factors that influence judicial decisionmaking; “legal” factors are distinguished from “nonlegal” factors such as the race of the defendant. Often, defendants' criminal record is included as a control variable and therefore (implicitly) treated as nonracial. But treating criminal history in this fashion obscures the possibly racialized dynamics that produce variation in this ostensibly race-neutral legal factor. Some argue that observed race differences in criminal records are a function of differential offending propensities and are therefore appropriately treated as race-neutral. Yet there is ample evidence that race profoundly shapes the likelihood of arrest and prosecution for many offenses, particularly drug crimes (Beckett, Nyrop, Pfingst, & Bowen 2005; Beckett, Nyrop, & Pfingst 2006; Blumstein 1993; Duster 1997; Goode 2002; W. Johnson et al. 1977; Provine 2007; Riley 1997; Sterling 1997; Tonry 1995). Thus, at least some of the observed racial differences in criminal histories arguably reflect discrimination rather than “differential rates of offending.” By disaggregating the moment of sentencing from the context of penal expansion, and by treating criminal record as a race-neutral independent variable, sentencing studies of judicial sentencing understate the role of race and its accumulated significance.

Surprisingly, even evidence that race shapes judicial sentencing is sometimes interpreted in a way that contributes to the penology of racial innocence. In the sentencing literature, the role of race in classificatory and perceptual processes is often subsumed under the idea that judicial decisionmaking is guided more generally by judges' “focal concerns,” i.e., perceptions of dangerousness and assessments of blameworthiness (Albonetti 1991, 1997; Engen & Gainey 2000; Spohn & Holleran 2000). That is, statistical evidence that black and Latino defendants are punished more severely is treated as indirect evidence that judges' decisions are shaped by their focal concerns. Conceptualizing evidence that blacks and/or Latinos are significantly more likely to receive harsher penalties (controlling for all other relevant factors) as evidence of judges' focal concerns obscures the presence of racial discrimination in sentencing—even in states that have adopted sentencing guidelines. In part, this conceptual move reflects researchers' attempt to make sense of the impact of various defendant characteristics—such as age and gender, as well as race/ethnicity—on sentencing decisions. It may also reflect the belief that statistically significant race effects may result from some unmeasured process rather than from racial discrimination.9 Nonetheless, researchers' unwillingness to treat statistical evidence of racial effects as evidence of discrimination raises the distinct possibility that no social scientific evidence is thought to be sufficient to establish racial discrimination.

In short, criminological and sociological studies of race and criminal justice processing tend to focus on judicial discretion in jurisdictions in which that discretion is somewhat constrained. Yet these studies tend not to contextualize their narrow institutional and temporal focus in the larger context of penal expansion, or to highlight the very real possibility that sentencing guidelines enhance (unscrutinized) prosecutorial discretion; they also tend not to call statistical evidence that black and Latino defendants are punished more severely than their white counterparts “racism” or “discrimination.”10 This erasure is especially paradoxical given widespread evidence of cumulative and compounding racism in the criminal justice system.

Punishment Expansion and the Longer Chains of Compounding Causation

Criminal justice expansion itself entails more complexity and recursivity than conventional studies of arrest, conviction, and sentencing can capture. Focusing on a narrow band of criminal processing ignores the proliferation of “entry points” and protracted “exit points” from criminal justice; moreover, this longer chain of entanglement with criminal justice becomes self-replicating. Seeing “racism” as limited to discrete causes and “criminal justice” as limited to isolated processing stages generates the penology of racial innocence that both underestimates and excludes from consideration the vast, interrelated penal processes that produce racial inequality. In what follows we provide evidence that criminal justice expansion has expanded and deepened the reach and impact of penal institutions, and has exacerbated racial inequality, in ways that render disaggregation particularly problematic.

More “Entry Points”

The legitimation of the war on drugs and broken windows policing has produced an upsurge in arrests for comparatively minor crimes. Emphasis on nonviolent crimes has entailed the deployment of more discretionary police tactics and an associated increase in racial disparity in arrests. More than 1.8 million Americans are arrested on drug charges each year, roughly four times the number of annual arrests for aggravated assault and 15 times the number for robbery (Bureau of Justice Statistics n.d.b; FBI 2009: Table 29). As many have observed, people of color are significantly overrepresented among those arrested for drug law violations (see Beckett, Nyrop, Pfingst, & Bowen 2005; Beckett, Nyrop, & Pfingst 2006; Duster 1997; Tonry 1995), and this became increasingly true over the course of the nation's war on drugs. Between 1980 and 2000, the national black drug arrest rate more than quadrupled, from roughly 6.5 to 29.1 per 1,000 people, while the white drug arrest rate increased much more moderately, from approximately 3.5 to 4.6 per 1,000 people (Beckett, Nyrop, & Pfingst 2006:106). Although the degree to which blacks are overrepresented in drug arrests relative to whites varies (see Beckett, Nyrop, & Pfingst 2006), it is clear that blacks are far more likely to be arrested for drug law violations than their involvement with illicit drugs would predict.11

Order-maintenance policing similarly generates more “entry points” into the criminal justice system. Broken windows policing reposes on addressing small infractions on “order,” and, as many scholars have confirmed, policing “disorder” brings greater arrests of people of color. With the rise of order-maintenance policing in New York City, for example, the number of misdemeanor arrests increased about 80 percent, from approximately 129,403 in 1993 to 224,665 in 2000. Similarly, the number of drug arrests more than doubled, from 66,744 in 1993 to 140,122 in 2000 (Greene 1999: Table 1; New York State Division of Criminal Justice Services, adult arrests for 2000 as of 27 Jan. 2010, [accessed 20 July 2010]). Of these, a growing percentage involved only marijuana: New York City pot arrests rose from fewer than 10,000 a year in 1993 to more than 60,000 in 2001 (Golub et al. 2007:139). Studies of the New York Police Department's stop-and-frisk patterns in the 1990s indicate that blacks and Latinos were much more likely than whites to be stopped, and that stops of people of color were less likely to be legally justified and to lead to an arrest than were stops of white people (Fagan & Davies 2000).12

More Small Penalties and Protracted “Exit Points”

Those who are arrested for minor crimes are subject to an increasingly wide array of criminal justice sanctions, including arrest, short-term jail stays, probation, spatial exclusions, legal financial obligations, and the loss of political, social, and legal rights. Indeed, there is evidence that one important consequence of penal expansion has been the sharp increase in the number of arrest warrants issued. Being “wanted” by the police has important social and economic consequences for people with warrants and their families. On the basis of six years of fieldwork in a poor, black Philadelphia neighborhood, Goffman (2009:353) concludes that:

Young men who are wanted by the police find that activities, relations, and localities that others rely on to maintain a decent and respectable identity are transformed into a system that the authorities make use of to arrest and confine them. The police and the courts become dangerous to interact with, as does showing up to work or going to places like hospitals.

Moreover, federal welfare legislation adopted in 1996 prohibits states from providing Temporary Assistance for Needy Families, Supplemental Security Income (SSI), general assistance, public and federally assisted housing, and food stamps to individuals who are “fleeing felons” (i.e., have a bench warrant stemming from a felony conviction) or are in violation of any condition of probation or parole.13 The Social Security Administration database is now linked to state warrant databases, so that the cessation of benefits occurs automatically upon issuance of an arrest warrant (provided that warrant appears in the state database). Although not as severe as imprisonment, the comparatively minor sanctions we highlight here may nonetheless adversely affect their targets and contribute to social and racial inequality. They also make the process of exiting the criminal justice system more difficult. Below, we briefly highlight the examples of monetary penalties and alternatives to incarceration.

Monetary Penalties

State and local governments are increasingly attempting to recoup criminal justice expenditures by imposing a number of new fees and fines on those passing through juvenile, municipal, state, and federal courts (Harris et al. 2010). Nationally, 84.2 percent of felons sentenced to probation were ordered by the courts to pay fees, fines, and court costs in 1995; 39.7 percent were also required to pay restitution to victims (Bonczar 1997: Table 8). This survey also found that 85 percent of misdemeanants sentenced to probation were assessed fees, fines, and court costs; 17.6 percent were also assessed restitution (Bonczar 1997: Table 8; see also Gordon & Glaser 1991). Although there is likely a good deal of variation across states and localities in the imposition of legal financial obligations (LFOs), the debt that may result from a criminal conviction can be quite significant and may include fees associated with the costs of public defense, drug investigations, compensation for victims, DNA analysis, filing fees, and many others. In addition, debt can accumulate from incarceration-related expenses including processing fees, room and board, medical and dental visits, telephone use, drug testing, and participation in alternative programs such as work release and electronic monitoring (Rosenthal & Weissman 2007). Failure to pay monetary sanctions leads to a nontrivial number of warrants, arrests, probation revocations, jail stays, and even prison admissions (Harris et al. 2010). Thus there is reason to suspect that the use of criminal justice sanctions to deter and penalize nonpayment of LFOs may help explain the continued growth of the criminal justice system and racial inequality therein, as well as the difficulty many people with criminal records have disentangling themselves from the criminal justice system.

Alternatives to Incarceration

Many of those concerned about prison expansion encourage the adoption of sentencing alternatives, including probation, electronic home monitoring, and daily reporting programs (see Tonry 1998). Surprisingly, these alternatives have expanded even more dramatically than prisons. Although probation is popularly understood as an alternative to incarceration, technical violations of probation and parole clearly contribute to the growth of the prison and jail populations. Moreover, these institutions have been altered in ways that increase the likelihood that parolees and probationers will return to jail or prison. Declining revenues for social services for those under community supervision combined with rising caseloads mean that these programs are far less able to promote reintegration and reduce recidivism (Simon 1993).

But limited resources and rising caseloads are compatible with a revised version of the nature and purpose of community corrections. According to the new vision, community corrections should aim to increase surveillance and manage risk rather than rehabilitate (Simon 1993). This vision is espoused in a 1999 Manhattan Institute report entitled “Broken Windows Probation: The Next Step in Fighting Crime.” Noting that the probation population has grown rapidly and that many probationers are readmitted to prison or jail, this report advocates a fundamental reorientation of probation. Probation, the report urges, must be primarily seen as a mechanism for achieving public safety rather than rehabilitation (Manhattan Institute 1999:5). In order to undermine probationers' expectation that they get two or more “free” violations, the report urges that “this permissive practice must be abandoned. All conditions of a probation sentence must be enforced, and all violations must be responded to in a timely fashion” (1999:7). Many probation departments have embraced this logic and the practices associated with it.

In short, although probation is touted as an alternative to incarceration, there is ample evidence that technical violations of probation and parole fuel prison and jail expansion in many locales. There is also reason to suspect that rising caseloads are encouraging a more surveillance-oriented and restrictive approach to community supervision that imposes significant costs on those experiencing this form of criminal justice. Other alternatives to incarceration are also increasing in popularity: The number of people “under jail supervision” but not confined to jail nearly doubled from 1995 to 2005 (Harrison & Beck 2006: Table 8). These alternatives include electronic monitoring, home detention, day reporting, community service, pretrial supervision, and work release and treatment programs. These programs are often aimed at the avoidance of incarceration, as well as therapeutic goals. However, there is reason to be concerned that these too either are or will become oriented primarily toward surveillance and may fuel the expansion of the “hard end” of the criminal justice system. As Stanley Cohen (1985) warned long ago, even the best-intentioned alternative programs often end up serving as supplements, rather than alternatives, to the hard end of the system, with the result that more small fish are caught in growing criminal justice nets. The paucity of resources devoted to therapeutic and vocational programs only exacerbates this risk.

In short, as prisons have expanded, other branches of the criminal justice apparatus have grown. Comparatively minor sanctions are imposed with greater frequency than imprisonment; some of these are promoted as alternatives to incarceration. However, given rising caseloads and the underfunding of social service and vocational programs, many of these programs tend to focus on surveillance rather than reintegration. Moreover, the requirements associated with these programs are often intensive; violations of these requirements are an important cause of criminal justice expansion and reinforce pre-existing racial inequalities in criminal justice. Disaggregating and decontextualizing stages in criminal punishment and limiting attention primarily to the courts overlooks the ways in which these minor sanctions make exiting the criminal justice system increasingly fraught and difficult and ignores entirely the way that this institutional dynamic fuels racial inequities in criminal justice.

Reinforcing, Compounding Inequality

Racism in criminal justice is a cumulative phenomenon and must be studied cumulatively rather than in isolated micro-moments of criminal justice intervention. Racial inequality in criminal justice accumulates in three ways. Over individual lifetimes, even minor criminal justice interventions leave marks that alter the life course and life chances (life cycle effects). At the group level, the accumulation of inequality happens across space (neighborhood effects) as well as time (generational effects). Social scientific research that disaggregates ignores evidence that crime and punishment are, at least in the era of mass incarceration and hyper-penality, interconnected and recursive rather than independent phenomena. Similarly, studies that examine evidence of racial bias at a single stage of the criminal justice process overlook the ways in which racialized dynamics shape criminal justice processes and outcomes.

For example, recent studies indicate that mass incarceration itself is criminogenic, and probably more so for blacks than for whites. Mass incarceration negatively affects the mental and physical health of inmates, destabilizes families and communities, and reduces earnings and employability (Braman 2002; Clear 2007; Clear et al. 2001; Farmer 2002; R. Johnson & Raphael 2006; Massoglia 2008; Pager 2003, 2005, 2007; Travis 2005; Western 2006; Western & McLanahan 2000; Western & Pettit 2005). Moreover, these effects may be more severe for blacks. For example, Pager finds that white job applicants with a criminal record are more likely to receive a “call back” than similarly qualified black applicants with no criminal record (2003, 2005, 2007). At the community level, too, the adverse effects of mass incarceration are concentrated in a small number of communities of color from which a significant proportion of inmates are drawn (Clear 2007).

In summary, the criminal justice system and the discretionary power of its agents have expanded dramatically; it has become far easier to become entangled in, and far more difficult to extricate oneself from, the growing reach of the penal system. These developments have very dramatically affected poor communities of color. Yet disaggregation and decontextualization in the study of criminal justice obscures the exercise of racial power in and through these penal institutions and processes.


We have argued that a basic paradox characterizes post–civil rights America: In the law, the conventional wisdom of many whites, and some social science research, racism is waning, aberrant, and individual; yet in the lives of many blacks and Latinos, the long arm of criminal justice is increasing, widespread, and systemic. The penology of racial innocence accommodates rather than confronts this basic paradox, erasing racism in the study and practices of punishment. Alongside victorious political proclamations of a “postracial” and “color-blind” America, the dominant legal framework claims that racism exists only with proof of racial “intent” and racial “causation.” Conventional social scientific studies of race in the criminal justice system have too often replicated these narrow standards of what racism is and how it can be identified, even as the criminal justice system grows in race-laden ways that diffuse “intent” and complicate “causation.”

The penology of racial innocence therefore insulates the criminal justice system from the very critiques it most warrants. If one were to take seriously all the ways that racial power shapes criminal justice, it might initiate serious changes. This possibility was not lost on the McCleskey Court, with Justice Powell writing that the disparate impact standard “throws into serious question the principles that underlie our entire criminal justice system” (McCleskey v. Kemp 1987:314–16). While the McCleskey Court used this slippery slope argument as grounds for rejecting expansive definitions of racism, we believe that the conditions of punishment do indeed warrant reconsideration.

In the post–civil rights era, both racism and criminal justice operate in systemic, interactive, and serpentine ways; epistemologies and methods for investigating racial power should be equally systemic and capacious. While the intent standard requires identification of blameworthy individuals who deviate from the race-neutral institutional and social background, one must also critically examine the background itself. This entails showing how race-neutral processes are indeed racialized. Studying racialization in affirmative action, for example, does more than expose biased moments in hiring; instead, it exposes the racial underpinnings of “merit” and investigates why jobs, wealth, education, and power are distributed as they are (Crenshaw et al. 1995). Similarly, in the criminal justice context, studying race entails more than exposing moments of bias; it also entails examining which behaviors are criminalized, how broad discretionary powers reflect and reinforce racial power even in the absence of malicious intent, and why the rallying call to “law and order” became so pronounced in the post–civil rights era.

Moreover, while the narrow standard of causation demands continual disaggregation, one must examine racial inequality as the re-aggregation of racial inequality in causally complex and recursive chains. This entails showing how seemingly minor criminal justice interventions beget yet more criminal justice intervention to produce significant racial inequality, as ripples of disadvantage spread over the individual life cycle, the neighborhood, and the racial group in cumulative and compounding ways. Increasing numbers of scholars are taking up these challenges, many of whom are cited here. We applaud these efforts. To reject the penology of racial innocence entails answering Baldwin's admonishment: “It is not permissible that the authors of devastation should be innocent. It is the innocence which constitutes the crime” ([1963]1998:292).


  1. 1We refer to the contraction of antidiscrimination law throughout this article. We do not mean to suggest that there was ever a heyday of expansive antidiscrimination law, although there were early cases such as Griggs v. Duke (1971) that allowed for disparate impact challenges. What we call antidiscrimination law's contraction might also be described as its failure to evolve. Selmi, for example, argues that the Supreme Court has adhered to its 1950s vision of seeing discrimination only in formal barriers and total or near-total exclusion; hence, the Court easily finds unlawful discrimination in affirmative action but rarely finds it in subtle or outcome-based forms (Selmi 1997:334–5).

  2. 2There is variation in definitions of post–civil rights era racism. Some emphasize the emergence of “symbolic racism” as the blend of antiblack affect with traditional American values embodied in the Protestant ethic (Kinder & Sears 1981); others emphasize the rise of “laissez-faire racism” that blames blacks themselves for possessing inferior cultural traits that produce lower economic standing (Bobo & Kluegel 1997); others still emphasize “color-blind racism” as the ideology through which whites rationalize minorities' lower status as the product of market dynamics, naturally occurring phenomena, and cultural limitations (Bonilla-Silva 2006). While there are important differences between these scholarly accounts, for our purposes they describe a basic pattern of denying the existence of widespread, commonplace, and institutional racism.

  3. 3For an excellent critique of references to “the vanishing black male,” see Legette (1999) and Reed (1992).

  4. 4There is debate over whether the intent standard is intrinsically restrictive. Selmi (1997:287), for example, suggests that it is the Supreme Court's limited vision of discrimination that constrains its interpretation of intent. Moreover, Selmi points out that constitutional challenges based on a theory of disparate impact do not necessarily fare better than those based on purposeful discrimination. In claims premised on disparate impact, the Supreme Court frequently accepts the defendant's justification for the challenged practice (see, for example, Wards Cove Packing Co. v. Atonio [1989], which rejected the claim of prima facie discrimination based on disparate impact of a high percentage of nonwhite workers in unskilled cannery jobs and a low percentage of such workers in skilled non-cannery positions). Nonetheless, we demonstrate that social scientific appropriation of the intent standard has restricted conceptions and interpretations of discrimination.

  5. 5In Washington v. Davis (1976), the plaintiffs challenged a written test for police officers on the grounds that the test excluded a higher proportion of black candidates than white candidates. In Personnel Administrator of Massaschusetts v. Feeney (1979), the plaintiff was a civilian woman who challenged a Massachusetts law giving preference to veterans over nonveterans for all state jobs upon passing the civil service examination. The Supreme Court rejected her constitutional challenge, holding that she failed to prove intent.

  6. 6This juxtaposition is also evident in Kennedy's (1994) critique of scholarship and court rulings that treat racial disparities as evidence of racism. Kennedy argues that the perception that criminal justice institutions are pervaded by invidious racial prejudice is not only overblown but also masks the real problem: comparatively high rates of interpersonal violence in black communities. This juxtaposition ignores the possibility that both the differential distribution of violence and racial inequality in criminal justice are profoundly racialized (see also Kennedy 1997:20).

  7. 7Had poor blacks lived in racially integrated neighborhoods throughout metropolitan areas, deindustrialization would have driven up rates of black poverty but would not have produced isolated neighborhoods characterized by highly concentrated poverty. Racial segregation is, in turn, a product of both historical and contemporary racial discrimination: high levels of racial residential segregation cannot be explained by either the preferences of blacks or their socioeconomic situation. Rather, ongoing housing discrimination and prejudice continue to keep many poor and working-class blacks in the ghetto (Massey 1990:354).

  8. 8Indeed, this cultural association appears to be quite strong. One study found that when viewers were exposed to news stories about crime, 60 percent of the viewers who watched a news story with no perpetrator shown falsely recalled seeing one, and 70 percent of these viewers believed the perpetrator to be black (Gilliam & Iyengar 2000). More recent studies suggest that stereotypes can work in the opposite direction as well. That is, exposure to black faces facilitates the perception of crime-relevant objects, while exposure to crime-relevant objects directs visual attention toward black faces (see Eberhardt et al. 2004).

  9. 9However, scholars' reluctance to infer that race shapes judges' perceptions and motivations is not shared by other scholars who seek to make theoretical sense of a statistically significant correlation. For example, political scientists regularly interpret statistically significant “party effects” as evidence of “partisanship” among political actors (see Jacobs & Helms 1996; K. Smith 2004). Yet evidence of race effects tends not to be construed as evidence of discrimination. For a thoughtful critique of race effects, see Bonilla-Silva & Zuberi (2008), Holland (2008), Zuberi (2001).

  10. 10Some studies of judicial sentencing outcomes explore whether the racial/ethnic context in which sentencing occurs affects sentencing outcomes, and whether and how these contextual factors interact with defendant characteristics. These studies generally employ multilevel statistical techniques and draw upon the racial threat perspective, which highlights the possibility that racial inequities will be most common in jurisdictions with medium-sized populations of color that are large enough to trigger (white) fear or hostility but too small to enjoy significant political power (see Blalock 1967; Blumer 1958; Liska, Lawrence, & Sanchirico 1982; Spohn & Holleran 2002; Steffensmeier & Demuth 2001). Implicit in this approach, then, is the idea that the racial (or class) composition of the population influences the response to the crime problem generally and to individuals accused of crimes specifically. Studies evaluating this perspective often do find that the racial composition of the population is a significant predictor of enhanced penality (see Beckett & Western 2001; Bridges & Crutchfield 1988; Carroll & Cornell 1985; Greenberg & West 2001; Huff & Stahura 1980; Jackson 1989; Jackson & Carroll 1981; Jacobs & Helms 1996; Liska, Lawrence, & Benson 1981; Marvell & Moody 1996; Mosher 2001; Stults & Baumer 2007; Western & Beckett 1999), yet these effects are not easily understood as “discrimination.” Other studies suggest that the racial composition of the jurisdiction interacts with defendant race/ethnicity (Helms & Jacobs 2002; B. Johnson 2006; B. Johnson et al. 2008; Sampson & Laub 1993; Ulmer & Bradley 2006; Wooldredge & Thistlethwaite 2004). These studies provide strong evidence that racism coexists and interacts with racial constructs that reproduce inequality; these complex processes are beyond the purview of narrowly defined intent and causation standards.

  11. 11Some argue that although racially disparate outcomes do not reflect the racial composition of those violating drug laws, they are nonetheless explicable in race-neutral terms. For example, some contend that the police focus on outdoor drug markets because it is cheaper and more productive for them to do so (Goode 2002). Few studies have empirically evaluated these claims. However, the available empirical evidence suggests that (1) the focus on outdoor markets contributes to, but does not explain, racially disparate arrest patterns, and (2) that outdoor arrests, especially buy-busts, may be both more labor intensive and less productive than indoor arrests for which search warrants are required (see Beckett, Nyrop, Pfingst, & Bowen 2005; Beckett, Nyrop, & Pfingst 2006).

  12. 12In New York City, 9.5 blacks were stopped for each black person arrested, while only 7.9 white people were stopped for each white person arrested (Office of the Attorney General of New York 1999: Table I.B.2). Stops involving blacks were also less likely to meet legal standards of reasonable suspicion (Fagan & Davies 2000:478–82).

  13. 13The Personal Responsibility and Work Opportunity Act of 1996, 42 U.S.C. § 608 (a)(9)(A)(ii).

Naomi Murakawais Assistant Professor in the Department of Political Science and Adjunct Assistant Professor in the Law, Societies & Justice Program at the University of Washington. Her research interests include race, American political development, and crime policy and the carceral state.

Katherine Beckettis Professor in the Department of Sociology and the Law, Societies & Justice Program at the University of Washington. Her interests include punishment, law, and social control; she is particularly interested in legal struggles over contested social issues, including drug use, crime, midwifery, and genetically engineered foods. Her most recent book explored the development and implementation of new urban control techniques in Seattle.