Community (Dis)Organization and Racially Motivated Crime Community (Dis)Organization and Racially Motivated Crime Author(s): Christopher J. Lyons Source: American Journal of Sociology, Vol. 113, No. 3 (November 2007), pp. 815-863 Published by: The University of Chicago Press Stable URL: http://www.jstor.org/stable/10.1086/521846 . Accessed: 02/04/2015 21:41 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. . The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to American Journal of Sociology. http://www.jstor.org This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/action/showPublisher?publisherCode=ucpress http://www.jstor.org/stable/10.1086/521846?origin=JSTOR-pdf http://www.jstor.org/page/info/about/policies/terms.jsp http://www.jstor.org/page/info/about/policies/terms.jsp AJS Volume 113 Number 3 (November 2007): 815–63 815 � 2007 by The University of Chicago. All rights reserved. 0002-9602/2007/11303-0006$10.00 Community (Dis)Organization and Racially Motivated Crime1 Christopher J. Lyons University of New Mexico This article examines the relationship between community structural conditions and racially motivated crimes against blacks and whites. Drawing on six years of police reports, census data, and survey data of Chicago communities, the study evaluates alternative hypotheses about the social organization of racial hate crime derived from social disorganization, resource competition, and defended communities perspectives. Multivariate analyses controlling for spatial autocor- relation reveal that antiblack hate crimes, in contrast to general forms of crime, are more likely in relatively organized communities with high levels of informal social control. Conversely, antiwhite incidents appear more numerous in traditionally disorganized com- munities, especially those characterized by residential instability. Acts of violence and discrimination motivated by racial animus are “in- terwoven with the fabric of our culture” (Newton and Newton 1991, p. ix). In the United States alone, history abounds with examples of racial and ethnic violence, from the well-known extremes of slavery and lynch- ing of African-Americans, to the genocide of Native Americans, to more mundane yet pervasive forms of discrimination and harassment. Given the indelible marks of interracial violence on our history, and on our current conceptions of racial and ethnic identity, social scientists have sought for centuries to understand the nature of intergroup conflict. Interest in the study of intergroup criminal victimization has grown 1 A dissertation fellowship from the Harry Frank Guggenheim Foundation supported this research. I thank Ross L. Matsueda, Becky Pettit, Stew Tolnay, Bob Crutchfield, and the AJS reviewers for helpful suggestions. I also thank the Chicago Police De- partment, the Chicago Commission on Human Relations, and the Project on Human Development in Chicago Neighborhoods for sharing data. An earlier version of this article was presented at the 2005 American Sociological Association meetings in Phil- adelphia. Direct correspondence to Christopher J. Lyons, Department of Sociology, MSC 05 3080, 1 University of New Mexico, Albuquerque, New Mexico 87131-0001. E-mail: clyons@unm.edu This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 816 with the emergence of hate crime as a new criminal category (Green, McFalls, and Smith 2001). Modern hate crime policy represents an attempt to engage the legal system in combating criminal acts motivated by prej- udice (Grattet, Jenness, and Curry 1998; Levin 2002). Since the appear- ance of hate crime legislation in the early 1980s (Grattet et al. 1998), hate crimes are now widely understood as criminal offenses motivated at least partially by offender prejudice against the victim’s putative group mem- bership (FBI 1999a). As a legal concept, then, hate crime distinguishes between crime motivated by bias and otherwise similar crime. We know little about the spatial distribution of racially motivated crime, especially compared to other forms of crime. For instance, what are the characteristics of communities that facilitate racial hate crime? Beginning with the seminal work of Coleman (1988), sociologists have argued that social capital facilitates various forms of instrumental action—including community social organization against crime (Sampson, Morenoff, and Earls 1999). Yet social capital may also have a “darker side” insofar as it can be appropriated for other, perhaps illegal actions, such as subversive behavior or the exclusion of racial outsiders (Coleman 1988; Portes 1998; Putnam 2000; Sampson 1999). Although criminologists have long invoked the concept of social disorganization to explain why communities low in economic and social capital have higher crime rates, it is unclear how community organization explains the special case of crime motivated by racial animus. Are racial hate crimes the product of socially disorganized communities low in economic and social capital? Or are racially motivated crimes more likely in communities with substantial resources to exclude outsiders? Given the impact of hate crimes on communities, questions related to community-level variation are particularly pertinent. Proponents of hate crime legislation argue that hate crimes are symbolic crimes that target whole communities as well as individual victims. Often the particular victim matters less to the offender than the victim’s social categorization (Berk, Boyd, and Hamner 1992; Levin and McDevitt 1993). Thus, an attack against one individual may have a “ripple effect” that reverberates to the wider community. Even relatively minor acts, in terms of penal law, can raise levels of fear and mistrust between groups and communities, exacerbating already tense relations between racial groups and elevating the potential for retaliation (Craig 1999; Martin 1995; Wisconsin v. Mitch- ell 1993). Indeed, some city departments have focused on the community context in the hopes of alleviating conflicts before they escalate into larger problems. To intervene successfully at the community level, we need to understand the community factors that lead to variation in hate crimes. The paucity of research on the spatial distribution of hate crime par- tially reflects concerns about the reliability of hate crime data. As Black This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 817 (1983, p. 40) suggested over two decades ago, moralistic crimes of “col- lective liability” such as racially motivated hate crime may “receive a degree of immunity from the law.” As with all official crime data, hate crime data represent only those incidents that come to the attention of the police department. Many argue that hate crime victims may be re- luctant to report incidents to the police out of fear of secondary victim- ization, humiliation, or the ambivalence of the police or the larger com- munity (McDevitt et al. 2000; Herek and Berrill 1992; Comstock 1991; Perry 2001). Recent victimization research, however, suggests the opposite: victims may be more likely to report hate crimes than nonbias crimes, possibly because hate crimes are perceived as more serious (Statistics Canada 1999). Even if potential incidents are more likely to come to the attention of the police, an unknown quantity of “true” hate crimes may not be classified as such by authorities, as some incidents do not provide necessary detail to meet classification criteria. Additional concerns about the reliability of official data stem from different reporting and classifi- cation standards across jurisdictions (Nolan, Akiyama, and Berhanu 2002; McDevitt et al. 2000). This latter issue in particular renders problematic analyses of the spatial distribution of hate crimes across agencies, counties, states, and countries. Perhaps for these reasons, research has tended to focus on important concerns other than etiology, including the political emergence of hate crime as a social problem (e.g., Jacobs and Potter 1998; Jenness and Broad 1997; Jenness and Grattet 2001), the diffusion and institutionalization of hate crime laws (e.g., Grattet et al. 1998; Phillips and Grattet 2000; Sav- elsberg and King 2005), the social production of hate crime statistics (Martin 1995; Boyd, Berk, and Hamner 1996), and hate crime reporting as an indicator of social movement success (McVeigh, Welch, and Bjar- nason 2003). With the exception of Green, Strolovitch, and Wong’s 1998 study of hate crimes in New York City, questions related to the antecedents of hate-motivated behavior remain relatively unexplored. This article attempts to address this gap by exploring the characteristics of communities that explain variation in hate crime. In doing so, it joins a long tradition of “Chicago-style” research (Sampson 2002b) by focusing on the role of community social organization in shaping crime—in this case, racially motivated crime in Chicago. By focusing on hate crime variation within a single jurisdiction (Chicago), the study circumvents some of the concerns of reporting bias, particularly those related to mea- surement inconsistencies across agencies. As others have argued (Green et al. 1998), reporting biases should be relatively similar across commu- nities because data are collected by a single department with an estab- lished tradition of hate crime reporting. Although focusing on community- level variation may not eliminate fundamental concerns about decisions This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 818 to report or classify hate crimes (Boyd, Berk, and Hamner 1996), it may present our most reliable insight into the spatial patterning of racially motivated crime. This study explores the extent to which ecological theories of crime and interracial conflict explain the distribution of hate crime against blacks and whites. Despite evidence that antiwhite hate crimes are the most frequently reported category of hate crime after antiblack incidents (FBI 1999a; Strom 2001), researchers have rarely examined antiminority and antimajority victimization together. Three general perspectives on crime and interracial conflict offer divergent expectations about the relationship between racial hate crime and the economic and social capital of com- munities. Assuming that hate crime is similar to crime generally, social disorganization theory predicts more hate crimes in disadvantaged areas with low levels of social capital. Traditional resource competition theories specify that hate crimes are most likely when competition between racial groups increases, especially during economic downturns when resources are scarce. In contrast, a defended community perspective implies that interracial antagonism is most likely in economically and socially orga- nized communities able to use these resources to exclude racial outsiders. I explore each of these perspectives further below. Then, drawing on six years of Chicago police records, census data, and survey data on social cohesion and informal social control in Chicago communities, I evaluate their implications for the community-level correlates of antiblack and antiwhite hate crime. THE ECOLOGY OF INTERRACIAL CONFLICT Social Disorganization The most prominent explanation for community-level variation in general crime focuses on the deterioration of community social controls brought about by certain structural conditions. In their pioneering study of urban communities, Shaw and McKay (1942) argued that ecological features, such as economic status, residential mobility, and racial heterogeneity, influenced delinquency rates largely indirectly by upsetting community social organization. Later revised by Kornhauser (1974) and others (e.g., Bursik 1988; Sampson and Groves 1989), Shaw and McKay’s concept of social disorganization has come to refer to a condition of weak controls in which communities are unable to achieve collective goals, such as low crime rates. According to Kornhauser (1978, pp. 64–65), the most impor- tant determinant of social organization is economic status, since poorer communities have fewer resources to combat crime. To a lesser extent, population mobility and racial heterogeneity also impede a community’s This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 819 ability to mobilize effectively against crime. A rich research tradition generally affirms the ecological correlation between these structural fea- tures and a variety of criminal outcomes, including juvenile delinquency, homicide, and violent crime. Applying a disorganization perspective to explain racial hate crime assumes that the antecedents of hate crime are therefore similar to those of other types of crime. That is, as with crime in general, economic dep- rivation, along with racial heterogeneity and residential instability, should influence both antiblack and antiwhite hate crime indirectly through com- munity controls. Communities facing economic downturns and other so- cial changes may be unable to exercise control over the dislocated, who may act on their impulses and racial prejudices. In addition, formal re- sources may also be important for combating racially motivated crime. Impoverished communities may experience more racial hate crime partly because they are less able to invest in social programs for police designed to promote tolerance or directly prevent hate crimes (van Dyke, Soule, and Widom 2001, p. 40). Whether hate crimes are the result of informal social control, however, cannot easily be resolved by ecological analyses based on census data alone. Investigation into the potential mechanisms that intervene between structural conditions and hate crime requires attention to informal insti- tutional processes at the community level, such as social cohesion and trust and norms of informal social control. Informed by a “systemic” or network model of community organization, recent elaborations of social disorganization theory have begun to assess more directly the informal social control mechanisms implied by social disorganization theory. The systemic model elaborated by Kasarda and Janowitz (1974) conceives of the community as a complex network of social relationships, ranging from close friendship and kinship ties to more diffuse associations with others and with institutions. A network conception of community directs atten- tion to the constraining and enabling potential of “social relations among persons and the structural connections among positions” within a given ecological unit (Sampson 1999, p. 255). In other words, the systemic model focuses on social capital, or the structures of relationships among people that facilitate action (Coleman 1988; Sampson et al 1999). Sampson and colleagues (Morenoff, Sampson, and Raudenbush 2001; Sampson, Raudenbush, and Earls 1997) conceptualize “collective efficacy” as a key component of community social capital that mediates between ecological conditions and crime. Generated from surveys of residents’ perceptions of their communities, collective efficacy consists of two un- derlying constructs: (1) social cohesion and trust and (2) norms of informal social control in relation to widely held goals of public safety and crime prevention. Thus defined, efficacious neighborhoods are cohesive neigh- This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 820 borhoods characterized by expectations for informal surveillance, espe- cially of youths, and intervention in problems leading to the control of potential criminal elements. In numerous studies, Sampson and colleagues demonstrate that collective efficacy works to prevent delinquency, vio- lence, and property crime. In addition, collective efficacy largely mediates the effect of macrostructural features, as social disorganization theory would predict. Collective efficacy is oriented toward crime and safety in general (Samp- son 2002a, p. 101) and is not necessarily meant to explain interracial conflict. The question remains, however, whether the pursuit of social goods such as public safety facilitates the control of interracial conflict as well. If hate crimes are etiologically similar to other forms of crime, areas high in social cohesion/trust and informal social control should be able to prevent hate crime just as they are able to prevent other types of crime. Because youths are more likely than those of most other age groups to be victims and offenders of hate crime (Strom 2001), close supervision of youths may discourage potential hate crime incidents. In much the same manner that German families with sufficient social capital are able to prevent children from engaging in antiforeigner violence (Hagan, Mer- kens, and Boehnke 1995), communities with substantial cohesion and informal social control may be able to regulate youths’ behavior, dis- suading them from similarly racist hate crimes. In sum, social disorganization theory predicts more hate crime in ec- onomically disadvantaged and residentially unstable communities. Fur- thermore, communities with high levels of social cohesion and norms of informal social control with respect to crime in general should exhibit lower amounts of racially motivated crime (see table 1). Resource Competition Rather than focusing on the deterioration of social controls, competition theories predict that racial conflicts covary with the degree of intergroup competition for limited resources in a shared environment (Olzak 1990; Soule and van Dyke 1999; van Dyke et al. 2001). Because groups occupy different positions or “niches” in the environment, competition for scarce resources renders group identities and boundaries salient, increasing the likelihood that individuals will align their actions with the interests of their ethnic group (Olzak 1992). In modern economies, labor markets are often segmented by race and ethnicity. Therefore, niches often do not overlap, and groups may not directly compete for the same occupations. During stable niche conditions such as these, interracial competition and resultant intergroup conflict are minimized. But when conditions of equi- librium are disrupted during times of economic contraction, niches begin This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 821 TABLE 1 Theoretical Perspectives and Hypothesized Direction of Relationships with Racial Hate Crimes Theoretical Perspective Economic Disadvantage Social Cohesion/ Informal Social Control Social disorganization . . . . . . . . . . � � Resource competition . . . . . . . . . . Black unemployment: � antiwhite White unemployment: � antiblack NA Defended communities . . . . . . . . . � � to overlap, and direct competition for resources increases. When groups perceive their share of resources to be threatened, ethnic and/or racial conflict ensues. According to Bonacich (1972), this material competition explains all forms of antagonism between ethnic groups, ranging from outgroup vi- olence to institutional discrimination. It is partly through antagonism that more powerful groups are able to maintain their position vis-à-vis other groups (Olzak 1992). For example, Tolnay and Beck (1995, p. 59) interpret the connection between lynching and economic downturns as evidence that “whites attacked when they believed that blacks were threatening their privileged access to . . . society’s scarce resources.” Likewise, Jenness and Grattet (1996) suggest that during times of economic recession, racial hate crimes may be a strategy for the white middle class to eliminate minority group threat. This theory is consistent with a frustration-ag- gression thesis at the microlevel (Hovland and Sears 1940), which connects macroeconomic downturns to psychological strain that motivates indi- viduals to aggress against vulnerable social groups. Evidence from eth- nographic research suggests that offenders do lash out at scapegoats whom they perceive as responsible for their economic displacement. Some hate crime offenders blame their economic insecurity or job instability on af- firmative action policies that increase competition with minorities (Pin- derhughes 1993). The literature on economic competition has focused primarily on an- timinority (and particularly antiblack) victimization by whites. Whether similar mechanisms apply to crimes targeted at majority members, such as antiwhite hate crime committed by blacks, remains to be seen. Whites may be more likely to mobilize against blacks for economic reasons be- cause of racist ideals of entitlement. However, research on related forms of black-on-white victimization, such as interracial homicide, suggests that This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 822 economic stressors may also motivate blacks to attack whites (Blackwell 1990; Jacobs and Wood 1999). Thus, like social disorganization theory, traditional resource competi- tion theory predicts negative associations between racial hate crime and community economic status, though for different reasons. However, com- petition theories imply that race-specific economic status captures the strain felt by a particular group more directly than overall (race-neutral) economic status. Therefore, the theory predicts that antiblack incidents should be most prominent in areas of lower white economic status, whereas antiwhite crimes should be more likely where blacks face poorer economic prospects (see table 1). Although much research reveals a direct relationship between economic recession and various forms of interracial conflict such as lynching (Olzak 1992; Soule 1992; Tolnay and Beck 1995) and black church arson (Soule and van Dyke 1999), some recent studies suggest that the relationship between economic conditions and hate crime may be more complex. In a monthly time series analysis of hate crime in New York City, Green, Glaser, and Rich (1998) find no association between hate crime and eco- nomic recession. In another study, Green, Strolovitch, and Wong (1998) report no relationship between antiblack crimes and race-specific em- ployment conditions once controlling for racial composition. Green and colleagues, however, caution that economic conditions may still be im- portant indirect determinants of hate crime. They suggest, for instance, that “the relationship between economic discontent and intergroup ag- gression may hinge . . . on the ways in which political leaders and or- ganizations frame and mobilize such grievances” (Green, Glaser, and Rich 1998, p. 89).2 Defended Communities A variant of traditional competition theories, the defended communities perspective focuses less on competition for jobs than on the defense of valued community identities. The defended community perspective is in- formed by the ethnographic work of Suttles (1972) and others (e.g., De Sena 1990; Rieder 1985) on the social construction of urban communities. 2 Based on their analysis of antiminority hate crime patterns in New York City and a reanalysis of lynching data, Green, Glaser, and Rich (1998) conclude that the con- nection between economic conditions and racially motivated crime is “elusive.” Caution is warranted, however, when comparing the patterns of contemporary hate crime with post Civil War–era lynching of African-Americans. The historical contexts and mean- ings of these crimes differ in important ways, and they also refer to different criminal actions. In contrast to lynching, contemporary hate crimes include a variety of be- haviors ranging from criminal harassment to more serious physical assault. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 823 According to Suttles (1972), the defended neighborhood occurs when some residents take action against a perceived threat to community identity. Although defensive posturing may take many forms, racially motivated crime is one means, albeit extreme, for “defending” a valued community image and way of life from the threat posed by racial outsiders. Suttles (1972, p. 58) suggests that defensive identity maintenance is often (al- though not only) triggered by fears of racial invasion from adjacent com- munities. In one of the few quantitative community-level analyses of hate crime to date, Green, Strolovitch, and Wong (1998) find support for this hypothesis. Specifically, they report that hate crimes against racial mi- norities in New York City are most common in traditionally white com- munities experiencing recent in-migration of racial outsiders. Such com- munities may be motivated to victimize racial minorities to defend identities based on racial (white) homogeneity from the threat posed by racial invasion. Beyond racial composition and change, the defended communities per- spective implicates other factors that may affect interracial tensions. In this article, I focus specifically on the model’s implications for the rela- tionship between racial hate crime and the economic and social capital of communities, net of racial demographics. Whereas social disorgani- zation and resource competition perspectives predict the same ecological correlation between economic status and racial hate crime, a defended community perspective implies the opposite: racial hate crimes will be more numerous in economically prosperous and collectively efficacious areas. First, under the assumption that racial outsiders pose more of a threat to affluent communities whose identities are based on ideals of homogeneity and economic stability, communities with greater economic capital may be more likely to resort to racial exclusion to preserve com- munity boundaries. Economic resources constitute an important part of the symbolic identity of the community that is perceived worthy of pro- tection. In addition, economic capital may provide resources for com- munities to exclude others and maintain boundaries.3 Second, the defended community model suggests that hate crimes are more likely in internally organized communities with high levels of social 3 McVeigh et al. (2003) offer a competing explanation in their analysis of hate crime reports in counties across the United States. They find that per capita income is positively related to reported hate crimes and ague that “relatively prosperous counties have more resources that could be directed toward the enforcement of hate crime laws” (McVeigh et al. 2003, p. 855). I discuss this alternative “reporting” hypothesis in the conclusion. McVeigh and colleagues, however, focus on a measure that aggregates different kinds of hate crime (not only racial hate crime). Furthermore, for the purposes of the current study, it is unclear whether resources for hate crime reporting vary appreciably at the community level within a single jurisdiction. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 824 cohesion and norms of informal social control. In some cases defensive hate crimes may be initiated with the perceived support of the larger community (Pinderhughes 1993; Suttles 1972, p. 201). If so, racial hate crimes would “not represent internal social disorganization but organized responses to perceived external threats” (Heitgard and Bursik 1987, p. 786). To the extent that hate crimes are the outcome of collective action, they may be more likely in communities with substantial social cohesion/ trust and norms of informal social control. Social closure and trust, for instance, provide the capacity to defend against perceived threats posed by outgroups (Portes 1998; Waldinger 1995). Indeed, as Suttles (1972, p. 35) notes, defended communities “generally call for some level of concerted action and thus a certain degree of cohesion.” Furthermore, communities organized in favor of goals of public safety and order, as indicated by prevailing norms of informal social control of crime, may create a climate conducive to interracial hate crime. Goals of public safety and order are often informed by widespread racial stereo- types, and the association of minority status with disorder may lead to the scrutiny of racial outsiders—even among those who are not explicitly racist. In a recent study, Sampson and Raudenbush (2004) find that racial composition influences the level of perceived disorder in Chicago neigh- borhoods, above and beyond the effect of actual or “objective” disorder. Specifically, residents of all races perceive more disorder in communities with higher proportions of minority populations, even after controlling for observable conditions. Sampson and Raudenbush (2004) suggest that racial stereotypes linking minorities to crime and poverty are powerful determinants of perceived disorder (see also Quillian and Pager 2001). Given widespread racial stereotypes, communities with strong norms of informal social control of crime may be particularly likely to equate safety and order with whiteness, and fear and declining order with minority status. Taken to their logical, albeit extreme end, racialized notions of order and safety suggest that the exclusion of racial outgroups may be perceived by some to be in concert with the control of crime in general. Thus, while norms of informal social control may hinder conventional crime, they may mix with racial stereotypes to increase the likelihood of hate crime against minorities. Social cohesion and informal social control may facilitate hate crime especially in communities whose identities are rooted in racial homoge- neity. In racially homogeneous contexts, social cohesion and informal so- cial control may take the form of “bonding” (as opposed to “bridging”) social capital (Putnam 2000), which encourages exclusive rather than in- terdependent identities. Although bonding social capital can be positive for communities by encouraging solidarity (as in the case of ethnic enclaves [Portes 1998]), it “may also create strong out-group antagonism” and po- This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 825 tentially produce negative externalities (Putnam 2000, p. 23). Of course, not all ethnically homogenous communities may employ social capital to keep outgroups at bay. Following the defended communities model, we would expect most hate crime to occur where (1) the incentive to protect identities rooted in racial homogeneity from the threat of racial invasion (Green, Strolovitch, and Wong 1998) coexists with (2) the social resources needed to exclude outsiders. Social cohesion and informal social control may be leveraged for hate crime particularly (or perhaps exclusively) in racially homogenous communities that are threatened by racial invasion, perhaps in the form of recent in-migration of racial outgroups. The defended communities perspective conceives racial hate crimes as both phenomenologically and etiologically distinct from other types of crime. Therefore, the perspective is consistent with the idea that differ- ential social organization—that is, the degree to which a community is organized for or against crime—depends on the nature or type of crime (Sutherland 1947; Matsueda 1988). For example, as implied above, it may be possible for a community to have low rates of conventional crime while experiencing relatively high rates of racial hate crime. In fact, it may be that the same structural conditions that facilitate the control of crime in general work in the opposite direction for racially motivated crime. In particular, relatively affluent, organized communities with high levels of social cohesion and informal social control may be simultaneously orga- nized against general crime and in favor of racial hate crime. At one extreme, this could mean that some communities actively band together to exclude unwanted newcomers. In this case, social cohesion and norms of informal social control enable a community to organize for hate crime. However, examples of large segments of a community banding together against outgroups, although certainly not unheard of, are prob- ably rare. Most hate crimes more likely result from individuals acting without the expressed consent of the community at large. These individ- uals nonetheless may perceive outgroups in certain communities to be out of place and may attack with the perception of support or ambivalence. A positive relationship between hate crime and norms of informal social control in particular would suggest that emphasis on the surveillance of criminal elements may, ironically, lead to intolerance of diversity and the attenuation of controls against racist victimization. Organization against general crime may not necessarily imply concomitant respect for diversity and individual rights (Sampson 1999, 2002a). This perspective therefore emphasizes the potential “dangers” of various forms of social capital (Coleman 1988; Portes 1998; Putnam 2000; Samp- son 1999). As a resource for collective action, social capital may be em- ployed for “positive” as well as “negative” ends. From the point of view of the actor, social capital is always positive insofar as it is goal oriented. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 826 That is, social capital facilitates desired action. From the standpoint of others, however, the actions facilitated by social capital may engender undesired consequences. In sum, the defended communities perspective predicts fewer hate crimes in economically disadvantaged communities and more racial hate crimes in communities with high levels of social cohesion and informal social control, in contrast to social disorganization and resource compe- tition theories. Table 1 summarizes the hypothesized relationships between community characteristics and racial hate crime for each of the three perspectives. DATA Dependent Variables: Hate Crimes in Chicago I assess these perspectives using antiblack and antiwhite hate crime in- cidents reported to the Chicago Police Department (CPD) from 1997 to 2002.4 The CPD began collecting hate crime reports in 1986.5 Since then, procedures for reporting and classifying hate crimes have been formalized within the police department. As mentioned earlier, police reports at the community level likely provide the most reliable data to explore the spatial patterning of hate crime. Nevertheless, given the limitations inherent in the reporting and classification of hate crime data (e.g., Boyd et al. 1996), it is possible that some of the patterns presented below reflect different reporting mechanisms rather than the racially motivated behavior. I ad- dress these issues further in the conclusion. The Municipal Code of Chicago defines hate crimes as criminal acts that are committed because of “a specific demographic characteristic of the victim” (CPD 2005, p. 4). Once a potential incident is reported to or detected by an officer, CPD guidelines require a preliminary investigation which entails “obtaining witness statements and writing a thorough report including evidence of bias such as racial slurs, written statements, or gestures made by the offender prior to the incident” (CPD 2005, p. 30). In addition to these factors, preliminary assessments of bias motive gen- erally hinge on the perceptions of the victim(s), the context in which the crime takes place (e.g., an ongoing pattern of harassment), and the absence 4 Incident-level data prior to 1997 were not made available by the CPD. Furthermore, data prior to 1997 cannot be reliably aggregated to community areas because maps are unavailable (see below). 5 Although other advocacy groups in Chicago collect hate crime reports on particular constituencies (e.g., Horizons Community Services collects reports on antigay and lesbian violence), the police department remains the only organization that system- atically collects data on racially motivated crime. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 827 of other motives (e.g., material gain) (CPD 2005; Green, Strolovitch, and Wong 1998, p. 380; Martin 1995). Following guidelines established by the FBI, for an incident to be reported as a hate crime, there must be “suf- ficient objective facts . . . to lead a reasonable and prudent person to conclude that the offender’s actions were motivated, in whole or in part, by bias” (FBI 1999b, pp. 4–6; quoted in Iganski 2002, p. 5). Incidents that meet initial screening criteria are classified as motivated by antiwhite or antiblack bias by Civil Rights Unit investigators. Once a preliminary hate motive is ascertained, each reported incident is investigated further to determine whether sufficient evidence exists to charge the offender with the additional felony charge of hate crime. As part of their follow-up investigations, the Civil Rights Unit eventually classifies the bias component of each incident as “bonafide,” “undeter- mined,” or “unfounded.” Of all racial hate crimes reported between 1997 and 2002, about 25% did not meet standards for hate crime classification (unfounded), 25% did not provide information necessary to determine the veracity of a bias component (undetermined), and the remaining 50% were found to be bonafide hate crimes. The results presented below in- clude only those bonafide incidents verified by the CPD, although sen- sitivity analyses reported in appendix B reveal very similar patterns for all reported incidents regardless of follow-up classification. The final data set pools six years of reported racially motivated hate crimes, resulting in a total of 380 antiblack incidents, 175 of which are deemed bonafide, and 185 antiwhite incidents, 101 of which are classified as bonafide. Incidents motivated by animus toward other racial groups (e.g., Latinos, Asians) are too few to analyze reliably. As table 2 reveals, the criminal content of reported hate crimes varies from less serious prop- erty damage to more serious aggravated assault, though well over half of all reports are for violent (antiperson) crimes. Compared to total reported incidents, bonafide incidents tend to involve a greater percentage of an- tiperson crimes and fewer criminal harassment crimes. For both antiblack and antiwhite hate crimes, results do not change substantively from those reported below if data are restricted to violent incidents (see app. B). It should be noted that the CPD does not include murder under the potential list of hate crimes. Table 2 also lists the known race of hate crime offenders as reported by the victim. In the vast majority of cases, antiblack and antiwhite hate crimes in Chicago are interracial incidents: whites are by far the most likely perpetrators of antiblack crime, and blacks are the typical perpe- trators of antiwhite incidents. In about 13% of antiwhite incidents and 22% of antiblack incidents, however, the race of the offender is unknown. Offender race is more likely to be unknown in offenses where offender- victim contact does not occur directly, as in telephone harassment or This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 828 TABLE 2 Antiwhite and Antiblack Incidents: Crime Type, Offender Race, and Descriptives, 1997–2002 White Victim Black Victim Total Reported Bonafide Total Reported Bonafide No. of incidents . . . . . . . . . 185 101 380 175 Type of crime (%): Assault . . . . . . . . . . . . . . . . 35.2 23.8 30.0 26.2 Battery . . . . . . . . . . . . . . . . 32.2 53.8 24.6 33.5 Property . . . . . . . . . . . . . . 13.6 12.5 23.0 27.0 Threat/harassment . . . 12.3 2.5 18.8 9.7 Robbery . . . . . . . . . . . . . . . 4.7 5.0 1.7 0.5 Burglary . . . . . . . . . . . . . . 0.9 2.5 1.0 1.2 Arson . . . . . . . . . . . . . . . . . . 0.9 0.0 0.6 1.9 Other . . . . . . . . . . . . . . . . . . 0.4 0.0 0.2 0.4 Offender race (%): Black . . . . . . . . . . . . . . . . . . 77.6 83.1 1.7 0.5 White . . . . . . . . . . . . . . . . . 5.5 3.5 72.1 75.7 Asian . . . . . . . . . . . . . . . . . . 0.0 0.0 1.0 1.9 Unknown . . . . . . . . . . . . . 16.0 13.4 25.1 21.9 Descriptives: Meana . . . . . . . . . . . . . . . . . 2.4 1.3 4.9 2.2 SDa . . . . . . . . . . . . . . . . . . . . 2.5 1.6 5.7 3.1 Rangea . . . . . . . . . . . . . . . . 0–11 0–8 0–25 0–15 Rate:b . . . . . . . . . . . . . . . . . . . . Meana . . . . . . . . . . . . . . . . . 1.0 0.5 2.0 0.8 SDa . . . . . . . . . . . . . . . . . . . . 1.0 0.6 2.8 1.3 Rangea . . . . . . . . . . . . . . . . 0–5.3 0–3.1 0–14.6 0–6.8 a Across 77 community areas. b Rates represent the average incidence per 10,000 individuals on 2000 census population estimates. vandalism. I conducted additional analyses to investigate the sensitivity of results to the race of offender. These analyses (not shown, available upon request) reveal that results are robust regardless of offender race. Excluding incidents with nonwhite or unknown offenders for antiblack incidents and excluding cases with nonblack and/or unknown offenders for antiwhite incidents does not substantively alter the findings. The anal- yses presented below do not distinguish between offender races. Geographic unit of analysis.—This study seeks to examine the spatial variation of antiblack and antiwhite hate crime across Chicago com- munities. The smallest level of aggregation for hate crime incidents pro- vided by the CPD is the police beat, an area typically larger than the census tract. There are 279 police beats in Chicago (compared to 865 census tracts), which correspond to 25 larger police districts. However, because census data were not available at the police beat level (police beats do not directly match up with census tracts), each incident was This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 829 mapped instead to one of Chicago’s 77 traditional community areas. Com- munity areas in Chicago correspond to census tracts, and aggregating census tract data to community area level is fairly straightforward. The community areas were first outlined in the late 1920s by researchers at the University of Chicago to delineate areas of reasonable ecological in- tegrity. The areas are generally recognized as discernable communities by local residents as well as by “administrative agencies and local institutions concerned with service delivery” (Sampson 1999, p. 248). Although Chi- cago community areas (average population near 40,000) are admittedly larger and more heterogeneous than local neighborhoods, they are sub- stantially smaller units than those used in previous community-level re- search on hate crime (i.e., Green, Strolovitch, and Wong 1998). Mapping incidents to community areas poses some hurdles. Incident data had to be reorganized using alternative mapping techniques. Maps of community area and police beat boundaries revealed that over 60% of police beats are contained completely within the boundaries of com- munity areas. The rest of the police beats fall into more than one com- munity area. For those incidents in police beats that cross community area boundaries, maps published by the police department showing the general location of incidents within each beat were overlaid with com- munity area maps to determine the community area location of each incident. About 97% of incidents were easily assigned community areas based on these maps. Decisions for the remaining 3% of cases were based on “closer calls” (e.g., the incident fell near the borders of two community areas).6 The bottom of table 2 presents some basic descriptives of incidents aggregated to Chicago’s 77 community areas. Figures 1 and 2 illustrate this variation across communities visually, mapping the distribution of antiblack and antiwhite hate crime rates by community racial composi- tion. Despite six years of data, some communities report no antiblack or antiwhite incidents.7 Additional dependent variables.—The study also compares the corre- 6 To maintain the confidentiality of individuals involved in hate crimes, the CPD did not make available exact locations (e.g., street addresses) of the hate crime incidents. 7 The data are geo-coded to the location where the incident occurred and do not provide information on victim or offender residence. Research on hate crime incidents indicate that at least one-third of hate crime incidents occur near the victim’s home (Strom 2001) and suggest that hate crimes may be more likely to occur near the victim’s home than nonbias crimes (McDevitt et al. 2001). Even if hate crime victims are victimized outside the community in which they reside, ecological theories emphasize criminogenic factors in the environment in which the hate crimes occur. The lack of geographical information on offender residence, however, may be more consequential, especially if offenders cross community boundaries to seek out particular kinds of victims. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 830 Fig. 1.—Antiblack hate crime rate by %white 1990, Chicago 1997–2002 lates of racially motivated crime to other crimes known to the police. Specifically, rates of overall crime and violent crime (homicide, rape, rob- bery, and aggravated assault) are constructed from the CPD’s annual reports of index crimes. In 1998, the department began to report index crime totals by community area. Rates for each community area are cal- culated based on the average incidence of each offense between 1998 and 2002 divided by the community area population in 2000. Independent Variables Census data.—To explore the ecological correlates of antiblack and an- tiwhite hate crime, tract-level census data are aggregated to Chicago com- munity areas. The 1990 and 2000 census data provide information on total population, racial/ethnic composition (%Hispanic and non-Hispanic white and black), overall economic status, race-specific economic status, economic inequality, and population mobility (see table 3). The models This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 831 Fig. 2.—Antiwhite hate crime rate by %black 1990, Chicago 1997–2002 for hate crime presented below use 1990 census data because these mea- sures are causally prior to hate crime reports. Results do not change substantively if 2000 census measures are used instead. An index con- centration at extremes or “ICE” (Massey 2001) measures economic in- equality, or the concentration of family poverty versus affluence, and may range from �1 (extreme poverty) to �1 (extreme affluence) (Morenoff et al. 2001, p. 529). Social cohesion and informal social control.—Data on social cohesion and the informal dynamics of social control are taken from the Community Survey of the Project on Human Development in Chicago Neighborhoods (PHDCN). Conducted in 1995, the survey asked 8,782 Chicago residents questions about their neighborhoods. Following Sampson et al. (1997), a social cohesion and trust scale consists of five items asking how strongly respondents agreed that (1) “People around here are willing to help their neighbors”; (2) “This is a close-knit neighborhood”; (3) “People in this neighborhood can be trusted”; (4) “People in this neighborhood generally This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp TABLE 3 List of Community Area Measures Measure Description Census measures (1990 and 2000): Total population Total no. of all persons Racial/ethnic composition Percentage of all persons Hispanic, per- centage non-Hispanic black, percentage non-Hispanic white Change in percentage black/ white Percentage black/white 2000 minus per- centage black/white 1990 Unemployment Percentage of the civilian labor force ages 16 and over who are unemployed Race-specific unemployment Percentage of the black, white, and His- panic civilian labor force ages 16 and over who are unemployed Poverty Percentage all persons below poverty line Index of concentration at ex- tremes (ICE) No. of affluent families (income above $50k) minus the no. of impoverished families, divided by the total no. of families Public assistance Percentage of all families receiving public assistance Single mothers Percentage of families with single mothers (with own children under 18) Renter vs. owner occupancy Percentage of occupied housing units owned vs. rented Moved in last five years Percentage of population ages five and over who have moved in the five years prior to census date PHDCN measures (1995): Social cohesion/trust Scaled from five items: (1) “People around here are willing to help their neighbors”; (2) “This is a close-knit neighborhood”; (3) “People in this neighborhood can be trusted”; (4) “People in this neighbor- hood generally don’t get along” (reverse coded); and (5) “People in this neighbor- hood do not share the same values” (re- verse coded). Informal social control Scaled from five items: (1) youths were skipping school and hanging out on a local street corner; (2) youths were de- facing a local building with graffiti; (3) youths were showing disrespect to an adult; (4) a fight broke out in front of their house; or (5) the fire station closest to home was threatened with closure. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 833 don’t get along” (reverse coded); and (5) “People in this neighborhood do not share the same values” (reverse coded). A scale for informal social control is constructed by combining five items that tap the likelihood that neighbors could be counted on to intervene if: (1) youths were skipping school and hanging out on a local street corner; (2) youths were defacing a local building with graffiti; (3) youths were showing disrespect to an adult; (4) a fight broke out in front of their house; or (5) the fire station closest to home was threatened with closure. Together, the informal control and social cohesion scales capture “the linkage of mutual trust and the willingness to intervene for the common good” (Sampson 2002a, p. 103), or what Sampson et al. (1997) call collective efficacy.8 The community survey data were originally aggregated to 343 “neigh- borhood clusters” (NCs), which are contiguous groupings of Chicago’s 865 census tracts. NCs are designed to be “relatively homogenous . . . with respect to distributions of racial-ethnic mix, SES, housing density, and family structure” (Morenoff et al. 2001, p. 526). Although the majority of the NCs fit completely within a single community, 16 of 343 NCs cross into more than one community area. I employ the following strategy for reorganizing the survey data to the community level. First, to address potential biases in the community-level survey, I employ a two-level random intercept model to estimate scores for informal social control and social cohesion for each neighborhood cluster. This two-level model adjusts for relevant respondent character- istics: race-ethnicity, gender, SES, age, marital status, home ownership, number of years living in the neighborhood, and number of residential moves in the last five years (Sampson et al. 1997).9 Next, I determine the proportion of a community area’s population represented by each NC. For the 16 NCs that straddle communities, I estimate the proportion of the community area population represented by the cluster tracts that fall within the community area boundary. I then weight each cluster’s con- tribution to a community area’s social cohesion and social control scores by multiplying each cluster’s hierarchical linear model–adjusted score by 8 In additional analyses, I also explored a measure of the density of local social ties. The measure of ties is based on the combined average of two questions concerning the number of relatives reported to live in the respondent’s neighborhood (coded 1, 1–2, 3–5, 6–9, 10 or more) (Morenoff et al. 2001). Considered with informal social control and social cohesion, however, social ties are not significantly related to hate crime; therefore, this variable is not included in models discussed below. 9 As the items in the social cohesion and informal social control scales are subject to missing values, scales are constructed based on the average value of completed items per individual. When aggregated to the NC level, the minimum number of completed responses to an item is 8 (maximum is 62), with an average number of completed responses to an item ranging from 21 to 24. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 834 the proportion of the community area population represented by the cluster. Formally, this aggregation procedure can be expressed as: CE p CE # P ,�ca nc nc where CEnc represents the HLM-adjusted score for NCi, which falls at least partially within the community area boundary, and P is the pro- portion of the community area (CA) population represented by the NC tracts that fall within the community area boundaries, for each NC within the community area. NEGATIVE BINOMIAL MODELS AND SPATIAL DEPENDENCE The principal outcomes of the study, antiblack and antiwhite hate crimes, are nonnegative count variables. Event count data typically approximate a Poisson distribution: large counts toward zero and a tail that skews in the positive direction. However, Poisson regression presupposes the sta- tistical independence of events, an assumption that might be violated if hate crimes in one community lead to retaliatory crimes in another (Green, Strolovitch, and Wong 1998, p. 384). The Poisson model also assumes equidispersion, or that the mean equals the variance. As table 2 suggests, the unconditional means and variances for antiblack and antiwhite in- cidents are in fact overdispersed. A negative binomial variant of the Pois- son model that allows for overdispersion is more appropriate given these conditions. For the multivariate analyses presented below examining ra- cial hate crimes, I estimate negative binomial models using community area incident counts pooled over seven years. The number of observations in these analyses is equal to the number of community areas (Np77). All models of general crime rates are estimated in ordinary least squares.10 I also control for the possibility of spatial clustering or dependence in the hate crime data. Even after controlling for the structural features of community areas, spatial dependence may be particularly likely for a number of reasons. First, the difficulty in identifying “communities” spa- tially means that we may introduce error when defining communities as aggregates of census tracts (Morenoff et al. 2001). Second, hate crimes may be spatially linked in important ways. For example, hate crimes in one community may lead to retaliatory hate crimes in neighboring com- 10 For comparative purposes, I estimated models for hate crime rates using a Tobit specification to correct for left censoring. Overall, the Tobit models reveal patterns similar to the negative binomial count models presented below (results available upon request). This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 835 munities. Failing at least to control for spatial autocorrelation could bias estimates for key independent variables (Anselin 1988; Baller et al. 2001). Using GeoDa software (Bruch and Mare 2006), I include in the hate crime models spatial lag variables (Baller et al. 2001) that essentially represent the average antiblack and antiwhite count in contiguous com- munities. I use a first order contiguity spatial weight matrix that identifies contiguous communities that share common boundaries (often referred to as queen contiguity). Formally, the spatial lag can be expressed as: w s , (1)� ij j j where wij is an element of spatial weights matrix (row standardized), and sj is the count of antiblack or antiwhite incidents of each community’s neighbor as defined by the weights matrix (Anselin 1988, 1995; Baller and Richardson 2002). In the multivariate models that follow, I incorporate a spatial lag term to assess the independent effects of spatial proximity on community levels of racially motivated crime.11 RESULTS Before moving to multivariate models, table 4 presents bivariate corre- lations between measures of key community characteristics and rates of racial hate crime and general crime. The associations between economic characteristics and antiblack incidents run counter to expectations based on economic competition or social disorganization. In fact, in sharp con- trast to conventional crimes, antiblack incidents are correlated positively with %white (see also fig. 1) and community affluence (note the positive sign for ICE) and associated negatively with traditional indicators of disadvantage. Furthermore, antiblack hate crimes are associated posi- tively with community social cohesion and informal social control. Con- versely, correlations with antiwhite incidents seldom reach statistical sig- nificance. An important exception is that antiwhite incidents correlate positively with both total and violent crime rates. In contrast to correlations with hate crime, table 4 confirms the well- 11 Spatial dependence can be accounted for using either a spatial lag or a spatial error model (Baller et al. 2001). There are a number of reasons for preferring the spatial lag to the spatial error specification. The error model is actually nested within the spatial lag model (Baller et al. 2001, p. 566). Furthermore, if the “real” specification of the dependence is a spatial lag, then not including the lag term would bias coeffi- cients. In contrast, if the spatial dependence is due to the clustering of unmeasured variables (spatial error), then failing to control for the error dependence would result in larger standard errors but unbiased coefficients (less efficiency). Therefore, I opt for the more conservative choice and specify a spatial lag model. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 836 TABLE 4 Bivariate Correlations with Hate Crime and General Crime Rages Antiwhite Antiblack Total Crime Violent Crime Antiwhite rate . . . . . . . . . . . . . . . . 1.00 .71** .41* .22* Antiblack rate . . . . . . . . . . . . . . . . .71** 1.00 .21 �.05 Population . . . . . . . . . . . . . . . . . . . . �.23** �.20* �.15 �.14 %black . . . . . . . . . . . . . . . . . . . . . . . . .16 �.13 .40** .82** %white . . . . . . . . . . . . . . . . . . . . . . . . .01 .25** �.27** �.72** %Hispanic . . . . . . . . . . . . . . . . . . . . �.28** �.11 �.31** �.37** Unemployment . . . . . . . . . . . . . . . .04 �.24** .31** .83** Black unemployment . . . . . . . . .12 �.09 .48** .84** White unemployment . . . . . . . . .02 �.16 .15 .47** Poverty . . . . . . . . . . . . . . . . . . . . . . . . .07 �.21* .39** .84** Public assistance . . . . . . . . . . . . . .08 �.21* .33** .87** Single mothers . . . . . . . . . . . . . . . . .02 �.25* .27** .81** Owners (vs. renters) . . . . . . . . . .12 .26** �.36** �.54** Mobility (past five years) . . . . �.08 �.14 .24** �.12 ICE . . . . . . . . . . . . . . . . . . . . . . . . . . . . .07 .28** �.22* �.78** Collective efficacy . . . . . . . . . . . . .08 .35** �.30** �.60** Informal social control . . . . . . .19 .41** �.21* �.50** Social cohesion . . . . . . . . . . . . . . . .01 .27** �.32** �.60** Note.—Pearson correlations for rates; community areas.N p 77 * P!.10. ** P!.05. documented positive associations between violent and overall crime and typical indicators of community disorganization (economic disadvantage and residential mobility) and negative associations with collective efficacy. This preliminary look suggests that the ecological correlates of racial hate crime, especially antiblack hate crime, differ from those of crime in general. Social disorganization theory suggests that two primary underlying con- structs—concentrated disadvantage and residential mobility—contribute to a community’s ability to control crime. Guided by previous research (e.g., Morenoff et al. 2001), principal component analyses were performed on a set of measures thought to represent these constructs: percentage of individuals in a community living in poverty, percentage of families re- ceiving public assistance, unemployment rates, percentage of single- mother families, percentage of housing units owned (vs. rented), and the percentage who have moved within five years of the census date. The results of the principal components analysis, presented in table 5, confirm these two factors. For the multivariate analyses that follow, disadvantage is factored with- out %black. Given the centrality of racial composition to theories of in- This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 837 TABLE 5 Principal Component Analysis: Chicago Census, 1990 Components Disadvantage Residential Stability Eigenvalue . . . . . . . . . . . . . . . . . . . . . . 4.17 1.45 %variance . . . . . . . . . . . . . . . . . . . . . . . 69.57 24.10 Loadings: Single mother . . . . . . . . . . . . . . . . .95 .23 Unemployed . . . . . . . . . . . . . . . . . . .94 .20 Poverty . . . . . . . . . . . . . . . . . . . . . . . .98 �.07 Public assistance . . . . . . . . . . . . . .97 .16 Mobility . . . . . . . . . . . . . . . . . . . . . . . .05 �.96 Owner . . . . . . . . . . . . . . . . . . . . . . . . �.70 .63 terracial conflict, it is important to try to disentangle the effects of race from economic conditions and other community characteristics.12 I consider next multivariate models examining the relationship between racial hate crimes and community characteristics.13 Looking first at an- tiblack crimes in table 6, model 2 verifies the negative association between community disadvantage and antiblack incidents, net of spatial proximity and log population. Residential mobility is not significant in any model, nor is there evidence that antiblack crimes are more likely in areas of high white unemployment (model 3). The effect of economic disadvantage holds when controlling for race-specific economic conditions but is not robust to racial composition (%white and %Hispanic in model 4, and %black and %Hispanic in model 5). When the ICE index is substituted for concentrated disadvantage, substantive conclusions remain the same: antiblack hate crime reports are most likely in relatively affluent com- 12 Given moderate correlations between racial composition, disadvantage, informal social control, and social cohesion (see app. A), collinearity is a potential concern in the analyses below. Diagnostics, however, reveal little evidence of collinearity: no VIF score is above 3, except disadvantage, which has a VIF score of 4.1. 13 As one reviewer noted, a key concern in the analyses that follow is the lack of a suitable lagged measure of interracial crime or hate crime. Measures of interracial crimes at the community level were unavailable, and interracial homicides, while available, are too rare at the community level to provide a reliable control. Instead, in additional analyses (results available upon request), I examined the relationship between 1997–2002 hate crimes and key predictors controlling for 1989 hate crimes obtained from the Chicago Commission on Human Relations. On the whole, controlling for 1989 hate crime incidents produces remarkably similar patterns to those reported below. The exceptions are for antiwhite hate crimes only. After controlling for antiwhite incidents in 1989, the effects of residential stability are attenuated, although still in the directions presented in table 8. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp 838 T A B L E 6 N e g a t iv e B in o m ia l R e g r e ss io n s: C o m m u n it y C h a r a c t e r is t ic s a n d “B o n a fi d e ” A n t ib l a c k H a t e C r im e s, 1 9 9 7 -2 0 0 2 M o d el 1 M o d el 2 M o d el 3 M o d el 4 M o d el 5 M o d el 6 M o d el 7 M o d el 8 M o d el 9 M o d el 1 0 M o d el 1 1 C o n st a n t .. .. .. .. .. .. .. .. .. � 4 .0 1 * � 6 .3 6 * * � 6 .2 9 * * � 6 .3 8 * * � 5 .6 4 * * � 1 4 .7 8 * * * � 1 1 .1 5 * * � 1 0 .1 1 * * � 1 1 .1 0 * * � 1 0 .1 8 * * � 1 2 .6 7 * * (2 .3 2 ) (2 .3 9 ) (2 .3 9 ) (2 .6 6 ) (2 .6 0 ) (3 .7 1 ) (4 .0 5 ) (4 .1 2 ) (4 .7 6 ) (4 .7 9 ) (5 .7 5 ) L n p o p u la ti o n 1 9 9 0 .. .. .. . .4 6 * * .6 7 * * .6 7 * * .6 0 * * .6 7 * * .8 8 * * .6 7 * * .7 0 * * .6 8 * * .7 5 * * .7 8 * * (. 2 2 ) (. 2 3 ) (. 2 3 ) (. 2 5 ) (. 2 6 ) (. 2 5 ) (. 2 6 ) (. 2 6 ) (. 2 6 ) (. 2 6 ) (. 2 4 ) S p a ti a l p ro x im it y .. .. .. .. . .6 7 * * .3 1 .3 0 .2 5 .2 6 .2 4 .2 0 .2 2 .1 9 .2 1 � .1 5 (. 3 4 ) (. 3 1 ) (. 3 1 ) (. 3 0 ) (. 3 0 ) (. 3 2 ) (. 2 9 ) (. 2 9 ) (. 2 9 ) (. 2 9 ) (. 2 8 ) D is a d v a n ta g e .. .. .. .. .. .. . � .6 5 * * * � .6 0 * * � .2 2 � .0 5 .0 3 .1 9 .1 7 (. 1 8 ) (. 2 0 ) (. 3 4 ) (. 3 2 ) (. 3 7 ) (. 3 7 ) (. 3 5 ) S ta b il it y .. .. .. .. .. .. .. .. .. .1 1 .1 3 .1 7 .2 4 .0 3 .1 1 .1 6 (. 1 5 ) (. 1 5 ) (. 1 5 ) (. 1 6 ) (. 1 8 ) (. 1 8 ) (. 1 7 ) W h it e u n em p lo y m en t .. .. . � .0 2 (. 0 2 ) % w h it e 1 9 9 0 .. .. .. .. .. .. .. .0 1 3 .0 1 * .0 1 3 .0 3 (. 0 1 ) (. 0 0 6 ) (. 0 1 ) (. 0 6 ) % b la ck 1 9 9 0 .. .. .. .. .. .. .. � .0 1 8 * * � .0 1 * * � .0 1 8 * * (. 0 1 ) (. 0 0 6 ) (. 0 1 ) % H is p a n ic 1 9 9 0 .. .. .. .. .. .0 1 � .0 0 7 .0 1 � .0 0 2 � .0 1 2 � .0 0 5 .0 2 * * (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) In fo rm a l so ci a l co n tr o l .. .. 1 .5 0 * * 1 .3 4 * * 1 .3 5 * * 1 .3 1 * * 1 .2 9 * * 1 .4 8 * * (. 6 0 ) (. 5 9 ) (. 5 8 ) (. 6 7 ) (. 6 6 ) (. 9 0 ) S o ci a l co h es io n .. .. .. .. .. . � .2 2 � .2 3 � .2 6 � .2 4 � .2 6 � .4 3 (. 4 7 ) (. 5 6 ) (. 5 3 ) (. 5 9 ) (. 5 7 ) (. 5 0 ) This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp 839 % ch a n g e in b la ck p o p u la - ti o n , 1 9 9 0 – 2 0 0 0 .. .. .. .. . � .5 2 (. 4 4 ) In fo rm a l so ci a l co n tr o l # % w h it e .. .. .. .. .. .. .. .. . � .0 0 2 (. 0 2 ) In fo rm a l so ci a l co n tr o l # % ch a n g e in b la ck p o p u - la ti o n .. .. .. .. .. .. .. .. .. . .1 4 (. 1 3 ) % w h it e 1 9 9 0 # % ch a n g e in b la ck p o p u la ti o n .. .. . � .0 2 * (. 0 1 ) In fo rm a l so ci a l co n tr o l # % w h it e # % ch a n g e in b la ck .. .. .. .. .. .. .. .. .. . .0 0 5 * * (. 0 0 2 ) O v er d is p er si o n .. .. .. .. .. .. 1 .2 3 .8 7 .8 7 .8 0 .7 6 .8 4 .7 1 .7 0 .7 1 .6 7 .3 1 (3 1 .0 0 ) (. 2 5 ) (. 2 5 ) (. 2 4 ) (. 2 3 ) (. 2 5 ) (. 2 3 ) (. 2 2 ) (. 2 3 ) (. 2 2 ) (� .1 6 ) L o g li k el ih o o d .. .. .. .. .. .. 1 4 6 .1 7 1 3 9 .1 0 � 1 3 8 .8 7 1 3 7 .7 3 � 1 3 6 .4 6 � 1 3 9 .2 7 � 1 3 6 .1 9 � 1 3 5 .2 1 � 3 6 .1 7 � 1 3 4 .9 0 � 1 2 6 .2 3 N o t e .— N p 7 7 C h ic a g o co m m u n it y a re a s; u n st a n d a rd iz ed co ef fi ci en ts ; S E s a re in p a re n th es es . * P ! .1 0 . * * P ! .0 5 . * * * P ! .0 0 1 . This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 840 munities, yet the effects of economic conditions are attenuated net of racial composition (results not shown). Models 6–11 consider the relationship between antiblack hate crime and social cohesion and norms of informal social control. Controlling for log population and spatial proximity, model 6 shows a large and positive effect of informal social control on antiblack incidents and a negative and nonsignificant effect of social cohesion. The positive association between informal controls and antiblack hate crimes is robust to controls for racial composition (models 7 and 8) as well as economic disadvantage and res- idential stability (models 9 and 10). Interestingly, the coefficient for com- munity social cohesion is opposite that for informal social control, but never reaches traditional levels of significance.14 Thus, norms of informal social control, rather than social cohesion, economic conditions, or racial composition, facilitate hate crimes against blacks. These results generally provide support for a defended community thesis. Net of other charac- teristics, communities with heightened norms of informal surveillance have more antiblack hate crime. I also explore whether the effects of social cohesion and informal social control vary across various community characteristics, such as racial com- position and change. In racially integrated communities, social capital may be more likely to bridge racial divides. Under these conditions, in- formal social control and social cohesion may promote tolerance and dis- courage racially motivated conflict. Alternatively, in relatively homoge- nous white communities or those that are proximate to black communities, social cohesion and informal social control may facilitate exclusive iden- tities that encourage antagonism toward racial minorities (Putnam 2000). White homogeneity may interact with social cohesion and/or informal social control, especially in communities facing the threat of racial in- vasion. The positive effect of informal social control in models 9 and 10 may therefore hold only for specific communities with particular identities based on racial homogeneity. Although no interactions are detected for social cohesion, model 11 presents evidence of a three-way interaction between informal social con- trol, %white 1990, and change in proportion black between 1990 and 2000. This interaction is meant to capture communities whose identities may be rooted especially in racial homogeneity: traditionally white commu- nities experiencing the threat of racial invasion (Green et al 1998). As 14 A single collective efficacy scale that combines informal control and social cohesion is positive and significant in early models but is not robust to racial composition or economic conditions. It appears that combining the components of collective efficacy into one scale masks countervailing effects of informal social control and social cohesion on antiblack hate crime at the community level. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 841 TABLE 7 Predicted Antiblack Hate Crime, Chicago Communities, 1997–2002 White Communities (85% White) Nonwhite Communities (10% White) Threata No Threatb Threata No Threatb High informal social con- trol . . . . . . . . . . . . . . . . . . . . . . . . 34.7 3.9 3.1 2.1 Low informal social con- trol . . . . . . . . . . . . . . . . . . . . . . . . .1 .9 .1 .3 Note.—See table 6, model 11. Except for %white, informal social control, and change in %black, all variables held at mean values. High informal social control: 1 SD above mean; low informal social control: 1 SD below mean. a Black in-migration 15%. b Black in-migration 0%. model 11 reveals, even in the presence of all lower-level two-way inter- actions, the three-way interaction is positive and significant: the effect of informal social control on antiblack hate crime increases in racially ho- mogenous white communities that are experiencing recent in-migration of black newcomers. Table 7 presents predicted values for antiblack hate crimes based on the interactions between informal social control, %white, and the rate at which racial homogeneity is threatened by black in-migration. The table illustrates two important findings. First, antiblack hate crimes are more likely in white communities, and especially white communities with sub- stantial levels of informal social control. Second, the effect of informal social control is far greatest, however, in those white communities facing the threat of racial invasion (upper left cell). Consistent with a defended communities perspective, the confluence of ability (informal social control) and incentive (threat to racial homogeneity) produces the greatest risk of racially motivated crime. Interestingly, however, even in communities that are not ethnically homogenous and/or undergoing the threat of racial invasion, the effect of informal social control on hate crime is positive.15 Models of community characteristics and antiwhite hate crime, pre- sented in table 8, reveal different patterns than for antiblack incidents. When the sample includes all communities in Chicago (model 2), con- centrated disadvantage and residential mobility do not appear to affect antiwhite hate crimes. These results, however, do not consider the poten- 15 I also test a similar three-way interaction between informal social control, %white, and proximity to black communities (measured by a spatial lag for %black 1990). This interaction is also positive, indicating that the effect of informal social control on antiblack hate crime is heightened in white communities that are spatially proximate to black communities. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp 842 T A B L E 8 N e g a t iv e B in o m ia l R e g r e ss io n s: C o m m u n it y C h a r a c t e r is t ic s a n d “B o n a fi d e ” A n t iw h it e H a t e C r im e , 1 9 9 7 – 2 0 0 2 A l l C h ic a g o a E x c l u d in g O u t l ie r sb M o d el 1 M o d el 2 M o d el 2 M o d el 3 M o d el 4 M o d el 5 M o d el 6 M o d el 7 M o d el 8 M o d el 9 M o d el 1 0 C o n st a n t .. .. .. .. .. .. .. .. .. � 6 .1 2 * * � 5 .3 4 * * � 5 .8 2 * * � 5 .8 5 � 6 .2 4 * * � 5 .9 7 * * � 2 .3 1 � 3 .9 5 � 3 .6 3 � 7 .8 3 * * � 5 .5 9 * (1 .8 4 ) (1 .8 7 ) (1 .8 3 ) (1 .8 4 ) (1 .8 9 ) (1 .9 7 ) (2 .6 8 ) (2 .9 6 ) (3 .1 7 ) (3 .7 7 ) (3 .1 3 ) L n p o p u la ti o n 1 9 9 0 .. .. .. . .6 4 * * * .5 6 * * .5 9 * * .5 9 .6 4 * * .6 3 * * .5 8 * * .6 8 * * * .6 6 * * * .6 9 * * .6 9 * * (. 1 8 ) (. 1 8 ) (. 1 7 ) (. 1 7 ) (. 1 9 ) (. 1 8 ) (. 1 8 ) (. 1 9 ) (. 1 9 ) (. 2 0 ) (. 2 0 ) S p a ti a l p ro x im it y .. .. .. .. . .6 9 * * .7 3 * * .5 7 * * .5 8 .5 4 * * .5 3 * * .5 9 * * .4 6 * * .4 9 * * .4 6 * .4 5 * (. 2 2 ) (. 2 4 ) (. 2 4 ) (. 2 4 ) (. 2 4 ) (. 2 4 ) (. 2 1 ) (. 2 3 ) (. 2 4 ) (. 2 4 ) (. 2 4 ) E co n o m ic d is a d v a n ta g e .. . .1 3 .3 0 * * .2 9 .2 6 .2 0 .3 3 .2 8 (. 1 2 ) (. 1 2 ) (. 1 4 ) (. 2 1 ) (. 2 1 ) (. 2 5 ) (. 2 5 ) R es id en ti a l st a b il it y .. .. .. . � .1 5 � .2 6 * * � .2 5 � .2 6 * * � .2 7 * * � .2 8 * * � .3 0 * * (. 1 2 ) (. 1 2 ) (. 1 2 ) (. 1 3 ) (. 1 3 ) (. 1 3 ) (. 1 4 ) B la ck u n em p lo y m en t 1 9 9 0 .. .. .. .. .. .. .. .. .. .. .0 1 (. 0 1 ) % b la ck 1 9 9 0 .. .. .. .. .. .. .. .0 0 1 .0 0 2 .0 0 4 (. 0 1 ) (. 0 0 ) (. 0 1 ) % w h it e 1 9 9 0 .. .. .. .. .. .. .. � .0 0 3 � .0 0 5 � .0 1 (. 0 0 5 ) (. 0 1 ) (. 0 1 ) This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp 843 % H is p a n ic 1 9 9 0 .. .. .. .. .. � .0 0 4 � .0 0 5 � .0 0 7 � .0 0 5 .0 0 0 � .0 0 4 (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) In fo rm a l so ci a l co n tr o l .. .. � 1 .1 0 * � 1 .1 1 * � 1 .1 0 * � .5 4 � .5 6 (. 5 8 ) (. 5 8 ) (. 5 8 ) (. 6 1 ) (. 6 1 ) S o ci a l co h es io n .. .. .. .. .. .. .1 8 .4 4 .3 0 .7 7 .8 5 (. 4 0 ) (. 5 0 ) (. 5 0 ) (. 5 1 ) (. 5 2 ) O v er d is p er si o n .. .. .. .. .. .. .3 6 .3 4 .2 4 .2 4 .2 3 .2 4 .2 4 .2 1 .2 1 .1 8 .1 8 (. 1 6 ) (. 1 6 ) (. 1 4 ) (. 1 4 ) (. 1 4 ) (. 1 4 ) (. 1 5 ) (. 1 4 ) (. 1 5 ) (. 1 3 ) (. 1 3 ) L o g li k el ih o o d .. .. .. .. .. .. � 1 3 0 .3 9 � 1 2 9 .0 5 � 1 1 6 .6 8 � 1 1 6 .6 7 � 1 1 6 .3 1 � 1 1 6 .2 1 � 1 1 8 .8 2 � 1 1 7 .6 3 � 1 1 7 .9 1 � 1 1 5 .1 3 � 1 1 4 .8 0 N o t e .— U n st a n d a rd iz ed co ef fi ci en ts ; S E s a re in p a re n th es es . a N p 7 7 C h ic a g o co m m u n it y a re a s. b N p 7 4 C h ic a g o co m m u n it y a re a s. * P ! .1 0 . * * P ! .0 5 . * * * P ! .0 0 1 . This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 844 tial influences of statistical outliers. Although no outliers were detected for antiblack hate crimes, interior analyses reveal three communities in the southwest corner of Chicago that unduly influence the patterns for antiwhite incidents. These three communities have unusually high levels of antiwhite hate crime given their relative affluence and racial (white) homogeneity.16 Once these outliers are omitted, both concentrated dis- advantage (positive) and residential stability (negative) are significantly related to antiwhite hate crimes. It appears that antiwhite incidents in the vast majority of Chicago are more likely in traditionally economically disadvantaged communities with high rates of population turnover. Only in the relatively affluent southwest corner of the city do antiwhite incidents follow a different pattern. Although thorough explanation for these dif- ferent patterns in the southwest warrants qualitative exploration beyond the scope of the present study, I address potential reasons in the conclusion. In models excluding the southwest corner, the effects of disadvantage hold after considering black unemployment, which is not significant (model 3), but is not robust to controls for racial composition (models 4 and 5). However, consistent with a systemic model of social organization, residential stability remains a significant and negative predictor of anti- white incidents net of racial composition or race-specific economic conditions. Models 6–10 explore the effects of cohesion and informal social control on antiwhite hate crime. In contrast to antiblack incidents, the coefficient for informal social control is negative and marginally significant in model 6. According to social disorganization theory, this would suggest that informal social control works to prevent antiwhite incidents just as it prevents general forms of crime, at least in sections of Chicago other than the southwest corner. The effect is robust to controls for racial composition (models 7 and 8), but not economic disadvantage or residential stability (models 9 and 10). In models 9 and 10, the only significant predictor of antiwhite incidents, other than log population and spatial proximity, is residential stability. Like conventional crime in general, antiwhite inci- dents are more likely in relatively unstable communities with frequent population turnover. I also explored whether the effects of social cohesion and informal social control vary by racial composition and/or the degree of threat posed to racial homogeneity in the form of changes in the white population. Unlike the models for antiblack hate crimes, there is no evidence that racial composition or the rate of increase in the white population proportion 16 Outlier analyses include examination of residuals, leverage, dfits, and Cook’s dis- tance. The three communities in the southwest corner of Chicago have dfits and Cook’s distance scores two to three times above recommended cut-off values. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp T A B L E 9 O L S R e g r e ss io n M o d e l s: C o m m u n it y E c o n o m ic a n d S o c ia l C a p it a l a n d G e n e r a l C r im e R a t e s, 1 9 9 8 – 2 0 0 2 T o t a l C r im e V io l e n t C r im e M o d el 1 M o d el 2 M o d el 3 M o d el 4 M o d el 5 M o d el 1 M o d el 2 M o d el 3 M o d el 4 M o d el 5 C o n st a n t .. .. .. .. .. .. .. .. .. . 7 9 7 .8 2 5 1 7 .3 7 * * * 1 ,6 6 9 .6 9 1 ,5 4 3 .4 4 � 1 ,0 8 0 .9 5 1 8 3 .6 2 * * * 1 2 3 .1 2 * * * 7 9 0 .9 9 4 0 5 .6 5 * � 9 4 .9 8 (5 8 .6 5 ) (1 5 4 .5 6 ) (7 2 9 .6 6 ) (9 5 6 .9 7 ) (1 1 9 7 .4 9 ) (8 .5 0 ) (2 1 .5 5 ) (1 4 2 .4 7 ) (1 4 1 .3 9 ) (1 7 4 .5 9 ) S p a ti a l p ro x im it y .. .. .. .. . 1 9 9 .4 0 * 1 2 1 .4 1 2 6 1 .6 2 * 1 7 1 .4 5 7 7 .8 4 3 0 .8 0 * * 1 4 .8 6 9 5 .2 5 * * * 3 8 .8 9 * 8 .6 9 (1 1 3 .2 4 ) (1 0 7 .4 9 ) (1 1 0 .7 2 ) (1 1 4 .7 6 ) (1 1 0 .4 5 ) (1 5 .0 9 ) (1 3 .5 5 ) (1 7 .6 5 ) (1 5 .0 7 ) (1 4 .4 6 ) D is a d v a n ta g e .. .. .. .. .. .. . 1 4 7 .5 2 * � 6 8 .7 2 5 .6 8 1 1 7 .6 4 * * * 7 7 .9 3 * * * 8 7 .5 8 * * * (6 8 .9 2 ) (9 9 .5 7 ) (1 1 5 .8 6 ) (1 1 .1 2 ) (1 3 .4 9 ) (1 6 .2 5 ) R es id en ti a l st a b il it y .. .. .. � 1 5 6 .6 5 * � 2 3 7 .0 4 * * * � 2 8 9 .1 8 * * * 7 .8 3 � 7 .7 5 � 1 5 .0 4 (5 9 .2 0 ) (5 9 .6 4 ) (7 0 .9 9 ) (8 .7 4 ) (8 .2 2 ) (1 0 .0 5 ) % b la ck .. .. .. .. .. .. .. .. .. .. . 7 .3 8 * 2 .3 8 8 .1 5 * * * 1 .4 5 * * * 2 .2 3 * 1 .6 8 * * * (2 .6 8 ) (2 .3 7 ) (2 .8 7 ) (. 3 6 ) (. 3 6 ) (. 4 1 ) % H is p a n ic .. .. .. .. .. .. .. .. . � 1 .3 5 � 4 .7 5 .4 0 .0 1 .2 0 .4 2 (3 .4 3 ) (4 .1 4 ) (3 .9 8 ) (. 4 7 ) (. 6 1 ) (. 5 6 ) In fo rm a l so ci a l co n tr o l .. . 3 3 .1 3 2 4 .0 5 2 6 7 .0 1 � 6 6 .4 5 � 5 8 .7 7 2 6 .9 8 (2 9 0 .5 8 ) (2 8 3 .6 4 ) (2 9 6 .3 1 ) (5 5 .8 3 ) (4 2 .5 8 ) (4 2 .2 7 ) S o ci a l co h es io n .. .. .. .. .. .. � 2 7 0 .7 3 * � 2 3 4 .6 4 � 4 2 .3 0 � 9 9 .0 7 * � 2 9 .1 7 2 7 .6 4 (1 6 8 .4 0 ) (2 2 1 .6 2 ) (2 1 5 .2 9 ) (3 8 .5 1 ) (3 1 .1 4 ) (3 0 .2 8 ) A d ju st ed R 2 .. .. .. .. .. .. .. . .2 1 .3 2 .1 3 .1 9 .3 3 .7 3 .8 2 .5 3 .7 5 .8 2 N o t e .— N p 7 7 co m m u n it y a re a s; u n st a n d a rd iz ed co ef fi ci en ts ; S E s a re in p a re n th es es . * P ! .1 0 . * * P ! .0 5 . * * * P ! .0 0 1 . This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 846 between 1990 and 2000 condition the influence of informal social control or social cohesion on antiwhite hate crime (results not shown). Tables 6 and 8 also test for spatial clustering in antiblack and antiwhite hate crimes. There is little evidence of spatial lag effects for antiblack incidents net of key structural variables. As revealed in table 6, any spatial clustering in antiblack hate crimes appears to be explained by the clus- tering of economic conditions, social capital, or racial composition across communities. In contrast, the structural conditions considered in table 8 do not completely explain the clustering of antiwhite incidents. Although the spatial lag effect is attenuated after controlling for economic conditions and residential stability, the coefficient remains significant and positive. The lag effect also persists after controlling for social cohesion and in- formal social control. These results suggest diffusion or contagion pro- cesses for antiwhite hate crimes: hate crimes against whites in one com- munity may lead to similar crimes in neighboring communities. How do these ecological patterns for racial hate crime compare with those for crime in general? The patterns for total and violent crime rates, presented in table 9, validate what previous social disorganization re- search has demonstrated: disadvantage, residential instability, and %black are strongly associated with conventional crime rates, with disadvantage and %black stronger predictors of violent crime and residential instability a stronger predictor of total crime. Social cohesion, but not informal social control, is negatively related to crime rates, yet this effect is not robust to other structural variables at this level of aggregation. The patterns for conventional crime (especially total crime) are somewhat similar to those for antiwhite incidents in the majority of Chicago (excluding the southwest corner). Both total crime and antiwhite hate crime occur in relatively ecologically disorganized communities characterized by residential insta- bility. In contrast, antiblack incidents appear etiologically distinct and are most numerous in more traditionally “organized” communities character- ized by norms of informal social control. Although the OLS models for crime rates consider different years (1998–2002) than the negative bino- mial regressions for hate crimes (1997–2002), restricting hate crime models to 1998–2002 incidents produces very similar results. CONCLUSIONS Drawing on police records, census data, and Chicago community survey data, this article provides new insight into the etiology of racially moti- vated crime at the community level. Multivariate analyses of antiblack and antiwhite hate crimes reveal a number of important findings. First, antiblack hate crimes are most numerous in relatively organized com- This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 847 munities with higher levels of informal social control, and especially in internally organized white communities undergoing the threat of racial invasion. Antiblack incidents are also more common in economically af- fluent communities, although the effect of economic conditions is not robust to racial composition and norms of informal social control. As much research has shown, communities rich in economic resources and dynamics of informal social control are able to manage forms of unwanted conventional crime and secure public goods such as safety. However, such communities appear simultaneously organized in favor of antiblack crimes. The correlates of antiblack hate crimes are distinguishable from those of crime in general. That is, crimes motivated by animus against blacks appear to be distinct phenomena, and theories traditionally em- ployed to explain the spatial organization of general crime must be revised to explain the special case of antiblack hate crime. Although researchers have tended to focus on the positive consequences of social capital for communities, some have conceded that social resources can be used for purposes that may be considered “antisocial” by some (Portes 1998; Putnam 2000; Sampson 1999, 2002a; see also Thomson 2005). This study provides empirical support for such concerns. The positive effect of a particular form of social capital, informal social control, on antiblack hate crimes warns of the potentially negative consequences of collectively pursuing safety and public order without concomitant toler- ance for racial diversity and individual rights. Racial stereotypes linking minorities—and especially blacks—to social ills such as disorder, violence, and poverty are widespread. Even members of white communities who are not “explicitly” racist may view the presence of blacks as a sign of decline and impending disorder (Sampson and Raudenbush 2004). Com- munities with heightened norms of informal social control that are mo- tivated to maintain safe and orderly ways of life may also be disposed to monitor the racial composition within their borders. In the name of or- ganizing against crime and disorder, these communities may run the risk of alienating certain groups and encouraging defensive (and dangerous) sentiments. The defended communities thesis assumes that communities are mo- tivated to protect valued identities from threats posed by outside elements. Insofar as blacks represent symbolic threats to the identities of certain communities, racially motivated crime may be a means for maintaining lines of class and color. Indeed, informal social control of general crime is more strongly related to antiblack hate crime in relatively homogenous communities that are undergoing changes in racial composition. That is, in communities that may be particularly motivated to maintain bound- aries in the face of external threats, informal social control appears to facilitate defensive posturing. However, the results indicate that informal This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 848 social control operates across a wider range of communities. The positive effect of informal social control on antiblack hate crime persists even among those communities not facing prospects of in-migration of racial outgoups. The results indicate that the antecedents of antiwhite and antiblack crimes differ in important ways. In contrast to antiblack hate crime, an- tiwhite hate crimes are somewhat more likely in disadvantaged com- munities, especially those with higher levels of residential instability. Ac- cording to a systemic model of social organization, residential mobility disrupts the social relations that enable organization against crime. Thus, like other forms of crime, antiwhite hate crimes appear to be a product of social disorganization brought about by population turnover. The dif- ferent patterns for antiblack and antiwhite hate crime underscore the importance of disaggregating hate crimes by bias motivation. Although scholars often disregard bias motivation and consider hate crimes as an aggregate category, this study suggests that antiwhite and antiblack crimes may be etiologically distinct and should be analyzed separately. The different patterns for antiblack and antiwhite hate crime warrant additional research. One set of potential explanations for these differences focuses on the degree to which general criminal propensity versus animus directly motivates antiblack versus antiwhite offenders. Offenders of de- fensive antiblack crimes may be more likely to display motives consistent with a specialization model (Messner, McHugh, and Felson 2004) whereby offenders “specialize” in hate crime and do not engage in other types of crime. Such offenders are distinguished more by racial prejudice than general criminal propensity. Defensive hate crimes may involve special- ized offenders because these incidents serve particular, future-oriented functions: restricting the spatial movement of racial outsiders or protecting against threats to a valued community (and collective) identity. This im- plies that racial territoriality and animus induced by social change are a direct cause of antiblack hate crimes. In contrast, the patterns for anti- white hate crimes are more similar to the patterns for conventional types of crime. This may suggest that offenders of antiwhite hate crime may be more criminally versatile than antiblack offenders. That is, they may be motivated by a variety of criminal goals, especially those for financial gain. In many urban environments, blacks disproportionately live in struc- turally disadvantaged communities where motivations for crime are en- demic (Sampson and Wilson 1995), and Chicago is no exception. This reasoning does not necessarily mean that antiwhite offenders are less prejudiced than antiblack offenders, or vice versa. However, in the case of antiwhite hate crimes, other criminal motives brought about by struc- tural conditions or stable criminal propensities may complicate the re- lationship between racial prejudice and behavior. Of course, such con- This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 849 clusions about the differences between antiblack and antiwhite hate crime remain speculative and await additional research on the ecology of racially motivated crime. The models for antiwhite hate crimes are complicated by the presence of communities in Chicago that appear to follow divergent patterns. What explains these deviant cases in the southwest corner of Chicago? Although a full explanation is beyond the scope of the present study, a number of possibilities merit further investigation. First, antiwhite incidents in these communities could be retaliatory—that is, made in response to antiblack incidents. The anomalous communities have higher-than-average rates of antiblack hate crime, although they are well predicted by the defended community model. In fact, these areas appear prime examples of relatively organized communities defending against a perceived threat posed by blacks. If blacks victimize whites in retaliation to antiblack incidents, then the relationship between antiwhite incidents and community eco- nomic and social capital may be indirect. These factors increase the like- lihood of antiblack incidents, which may increase the likelihood that blacks will retaliate. Second, Perry (2002, p. 73) has suggested that whites may be more willing than minorities to report racially motivated crimes or more readily define acts as an affront to their privileged racial status. Although this argument has not been subject to empirical test, it is consistent with research showing whites’ greater trust and confidence in the police com- pared to minorities (e.g., Weitzer and Tuch 2005). It is also possible that the reporting of antiwhite crime in some communities is a strategy for identity maintenance—another resource for defending the color line from perceived threats to valued identities. In addition to these possibilities, the collective memory of blacks and whites might influence hate crime rates in Chicago’s southwest. Both whites and blacks may be more sensitive to potentially racially motivated conflict given the history of tension in Chicago’s southwest dating back to Civil Rights protests and KKK gatherings (Berlet 2001). This historical framework may also help explain why this relatively stable and affluent section of the city reports unusually high numbers of antiwhite hate crime. These speculative explanations require more extensive investigation. This study explored the extent to which theoretical perspectives on the etiology of interracial crime explain the spatial distribution of officially reported hate crimes. Although the patterns for antiblack hate crimes are supportive of the defended communities perspective, definitive tests must be able to rule out alternative explanations for the positive association between antiblack incidents and community economic and social capital. One competing explanation concerns the issue of reporting versus the actual commission of hate crimes. Specifically, relatively affluent and col- This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 850 lectively efficacious communities may be more disposed than less “orga- nized” communities to report antiblack crimes when they do occur. In an analysis of hate crime reporting across U.S. counties, McVeigh et al. (2003) argue that counties with higher per capita income have higher numbers of reported hate crimes because they have more resources to encourage enforcement of hate crime laws. Although McVeigh et al. (2003) focus on a larger unit of analysis (counties vs. communities) and aggregate all hate crimes into one outcome, this line of reasoning counters a defended com- munities perspective. Likewise, it could be argued that the components of collective efficacy work to encourage the reporting of hate crimes by fostering trust and other informal mechanisms that support victims. Browning (2002), for instance, finds that women feel more comfortable disclosing domestic disturbances to neighbors and authorities if they reside in collectively efficacious neighborhoods (although overall, collective ef- ficacy is negatively related to neighborhood levels of domestic violence). Social cohesion and trust and norms of informal social control may also encourage reporting of crimes by improving relationships between citizens and formal agents of control. Attempts at increasing neighborhood efficacy through community policing emphasize that “citizen calls to the police . . . are one form of social control from the bottom up” (Sampson 2002a, p. 103). The reporting argument implies that efficacious or relatively affluent communities are more sensitive to antiblack hate crimes and encourage victims or witnesses to report incidents to authorities. In communities with fewer economic resources and norms of informal social control and less cohesion, the facility of reporting incidents may be impeded. However, reporting bias is typically inversely related to incident seriousness (Myers 1980; Skogan 1984). That is, serious crimes involving threat of or actual bodily harm are more likely reported than “less serious” crimes. If so, we might expect the reporting benefits of economic and social capital to be greatest for incidents that might, in other contexts, go unreported, such as incidents not involving bodily harm. The data provide some opportunity to assess this hypothesis by ex- ploring whether the effects of community characteristics vary by incident seriousness. Dividing antiblack hate crime counts into violent (antiperson) crimes and nonviolent (primarily property) crimes, supplementary anal- yses reveal little support for the reporting hypothesis. For antiblack in- cidents, informal social control is positively related to both violent inci- dents and nonviolent incidents, and tests for differences in the magnitude of effects across models (Clogg, Petkova, and Haritou 1995) reveal no evidence that informal social control is more strongly related to less serious antiblack incidents. The same results hold for economic disadvantage, This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 851 which is negatively related to both categories of antiblack hate crime (results available upon request). These additional analyses do not provide definitive tests of whether community characteristics facilitate the commission versus the reporting of antiblack hate crime, nor do they address the discretion involved in defining incidents as racially motivated (see Boyd et al. 1996). Nonetheless, they do lend further credibility to the defended communities model. Fur- thermore, if social and economic resources facilitate the reporting of an- tiblack hate crime, one might expect them also to affect the reporting of antiwhite hate crime in similar ways. Yet, for antiwhite hate crime, in- formal social controls produce fewer reports. In the absence of reliable victimization data that allows more rigorous analysis of reporting behav- ior, the defended community perspective remains a theoretically compel- ling explanation for the social organization of antiminority hate crime. Finally, the current study examined hate crimes across relatively large and heterogeneous community areas, and the mechanisms of collective organization may operate more appropriately at the neighborhood level. Although original tests of the defended community thesis (Green, Stro- lovitch, and Wong 1998) focused on substantially larger units (average population near 130,000), future work should endeavor to explore the patterns for antiblack and antiwhite hate crimes at smaller levels of geo- graphical analysis. This article represents an initial step forward in our understanding of the newly redefined but long-standing social problem of racially motivated crime. The topic of social organization and crime has long occupied so- ciologists interested in explaining the spatial variation of crime, though it has been unclear how these theories can be applied to the specific case of crimes motivated by racial animus. This study indicates that disor- ganization may well explain reported hate crimes against whites, but antiblack hate crimes are not the product of traditionally crime-ridden communities. In contrast, in the process of collectively organizing for valued community ideals such as safety and order, communities may simultaneously encourage, either explicitly or implicitly, antiblack incidents. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp A P P E N D IX A T A B L E A 1 B iv a r ia t e C o r r e l a t io n s 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 . A n ti b la ck ra te .. .. .. .. .. .. .. .. 1 .0 0 2 . A n ti w h it e ra te .. .. .. .. .. .. .. .. .7 1 * * * 1 .0 0 3 . C ri m e ra te (1 9 9 8 – 2 0 0 2 ) .. .. .. . .2 1 * .4 1 * * 1 .0 0 4 . V io le n t cr im e ra te .. .. .. .. .. .. . � .0 5 .2 2 * * .6 1 * * * 1 .0 0 5 . L o g p o p u la ti o n 1 9 9 0 .. .. .. .. .. � .2 6 * * � .2 5 * * � .1 9 * � .0 8 1 .0 0 6 . % w h it e 1 9 9 0 .. .. .. .. .. .. .. .. .. .2 4 * * � .0 4 � .3 5 * * � .7 9 * * * � .0 5 1 .0 0 7 . % b la ck 1 9 9 0 .. .. .. .. .. .. .. .. .. � .1 5 .1 4 .4 4 * * .8 3 * * * � .1 2 � .8 6 * * * 1 .0 0 8 . % H is p a n ic 1 9 9 0 .. .. .. .. .. .. .. � .1 1 � .2 3 * * � .2 4 * * � .2 5 * * .3 3 � .0 1 � .4 7 * * * 1 .0 0 9 . A n ti b la ck sp a ti a l la g .. .. .. .. .. .1 2 .0 0 � .0 3 � .1 1 .1 5 .2 0 * * � .1 6 .0 1 1 .0 0 1 0 . A n ti w h it e sp a ti a l la g .. .. .. .. .0 8 .1 5 .1 3 .2 7 * * .1 4 � .1 9 * .2 0 * � .0 4 .6 8 * * * 1 .0 0 1 1 . U n em p lo y m en t .. .. .. .. .. .. .. � .2 0 * .0 6 .3 6 * * .8 7 * * * � .1 2 � .7 9 * .7 8 * * * � .1 6 � .1 3 .2 4 * * 1 .0 0 1 2 . B la ck u n em p lo y m en t .. .. .. .. � .0 5 .1 2 .1 8 .3 7 * * .0 6 � .3 3 * * .2 7 * * � .0 3 � .1 1 � .0 5 .4 6 * * * 1 .0 0 1 3 . W h it e u n em p lo y m en t .. .. .. .. � .1 8 .0 6 .0 4 .3 0 * * � .0 4 � .2 8 * * .2 4 * * .0 2 � .1 1 .0 0 .3 2 * * .1 3 1 4 . P o v er ty .. .. .. .. .. .. .. .. .. .. .. � .2 1 * .0 7 .3 9 * * .8 4 * * * � .0 2 � .7 9 * * * .7 0 * * * � .0 5 � .1 6 .1 8 .9 3 * * * .4 9 * * * 1 5 . P u b li c a ss is ta n ce .. .. .. .. .. .. . � .2 1 * .0 8 .3 3 * * .8 7 * * * � .1 1 � .8 0 * * * .8 0 * * * � .2 0 * � .1 5 .1 9 .9 5 * * * .4 3 * * 1 6 . S in g le m o th er s .. .. .. .. .. .. .. . � .2 1 * .0 5 .3 5 * * .8 3 * * * � .0 7 � .8 0 * * * .7 9 * * * � .1 5 � .1 3 .2 5 * * .9 5 * * * .4 5 * * * 1 7 . O w n er (v s. re n te r) .. .. .. .. .. . .2 6 * * .1 2 � .3 6 * * � .5 4 * * * � .2 3 * * .5 3 * * * � .3 3 * * � .1 6 .1 5 � .0 2 � .5 6 * * * � .4 1 * * 1 8 . M o b il it y in p a st fi v e y ea rs .. . � .1 4 � .0 8 .2 4 * * � .1 2 .2 9 * * .2 0 * � .3 1 * * .1 9 * .0 0 � .1 1 � .1 7 .0 4 1 9 . D is a d v a n ta g e .. .. .. .. .. .. .. .. � .2 5 * * .0 3 .3 6 * * .8 6 * * * � .0 1 � .8 1 * * * .7 6 * * * � .1 1 � .1 5 .2 0 * .9 4 * * * .4 5 * * * 2 0 . R es id en ti a l st a b il it y .. .. .. .. .. .1 1 .0 6 � .2 8 * * .0 5 � .2 9 * * � .1 3 .2 3 * * � .1 6 .0 2 .1 1 .1 3 � .0 7 2 1 . IC E .. .. .. .. .. .. .. .. .. .. .. .. . .2 8 * * .0 7 � .2 2 * * � .7 8 * * * � .0 4 .7 9 * * * � .6 2 * * * � .8 9 * * * .1 6 � .1 2 .1 6 � .1 9 2 2 . C o ll ec ti v e ef fi ca cy .. .. .. .. .. . .3 5 * * .0 9 � .3 0 * * � .6 0 * * * � .2 7 * * .7 1 * * * � .5 1 * * * � .6 9 * * * .1 7 � .1 9 .1 7 � .0 1 2 3 . In fo rm a l so ci a l co n tr o l .. .. .. . .4 1 * * .1 9 � .2 1 * � .5 0 * * * � .2 5 * * .5 7 * * * � .4 0 * * � .5 8 * * * .1 4 � .1 8 .1 4 .0 1 2 4 . S o ci a l co h es io n .. .. .. .. .. .. .. .2 7 * * .0 1 � .3 2 * * � .6 0 * * * � .2 6 * * .7 2 * * * � .5 3 * * � .6 9 * * * .1 7 � .1 8 .1 7 � .0 2 This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp T A B L E A 1 (C o n ti n u ed ) 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 2 3 2 4 1 . A n ti b la ck ra te .. .. .. .. .. .. .. .. 2 . A n ti w h it e ra te .. .. .. .. .. .. .. .. 3 . C ri m e ra te (1 9 9 8 – 2 0 0 2 ) .. .. .. . 4 . V io le n t cr im e ra te .. .. .. .. .. .. . 5 . L o g p o p u la ti o n 1 9 9 0 .. .. .. .. .. 6 . % w h it e 1 9 9 0 .. .. .. .. .. .. .. .. .. 7 . % b la ck 1 9 9 0 .. .. .. .. .. .. .. .. .. 8 . % H is p a n ic 1 9 9 0 .. .. .. .. .. .. .. 9 . A n ti b la ck sp a ti a l la g .. .. .. .. .. 1 0 . A n ti w h it e sp a ti a l la g .. .. .. .. 1 1 . U n em p lo y m en t .. .. .. .. .. .. .. 1 2 . B la ck u n em p lo y m en t .. .. .. .. 1 3 . W h it e u n em p lo y m en t .. .. .. .. 1 .0 0 1 4 . P o v er ty .. .. .. .. .. .. .. .. .. .. .. .2 9 * * 1 .0 0 1 5 . P u b li c a ss is ta n ce .. .. .. .. .. .. . .3 5 * * .9 4 * * * 1 .0 0 1 6 . S in g le m o th er s .. .. .. .. .. .. .. . .3 0 * * .9 2 * * * .9 3 * * * 1 .0 0 1 7 . O w n er (v s. re n te r) .. .. .. .. .. . � .1 4 � .7 4 * * * � .5 5 * * * � .5 8 * * * 1 .0 0 1 8 . M o b il it y in p a st fi v e y ea rs .. . � .1 1 .0 3 � .1 9 � .1 6 � .5 7 * * * 1 .0 0 1 9 . D is a d v a n ta g e .. .. .. .. .. .. .. .. .3 1 * * .9 8 * * * .9 7 * * * .9 5 * * * � .7 0 * * * � .0 4 1 .0 0 2 0 . R es id en ti a l st a b il it y .. .. .. .. .. .1 2 � .0 7 .1 6 .1 0 .6 3 * * * � .9 7 * * * .0 0 1 .0 0 2 1 . IC E .. .. .. .. .. .. .. .. .. .. .. .. . � .8 7 * * * � .4 4 * * � .3 1 * * � .9 4 * * * .6 9 * * * .0 5 � .9 3 * * * � .0 4 1 .0 0 2 2 . C o ll ec ti v e ef fi ca cy .. .. .. .. .. . � .6 7 * * * � .4 0 * * � .3 3 * * � .7 7 * * * .7 3 * * * � .2 6 * * � .7 5 * * * .3 0 * * * .7 2 * * * 1 .0 0 2 3 . In fo rm a l so ci a l co n tr o l .. .. .. . � .5 6 * * * � .3 4 * * � .3 4 * * � .6 7 * * * .7 0 * * * � .3 5 * * * � .6 5 * * * .3 6 * * * .6 6 * * * .9 0 * * * 1 .0 0 2 4 . S o ci a l co h es io n .. .. .. .. .. .. .. � .6 7 * * * � .3 9 * * * � .2 9 * * � .7 6 * * * .6 7 * * * � .1 8 � .7 4 * * * .2 3 * * * .6 9 * * * .9 6 * * * .7 4 * * * 1 .0 0 N o t e .— P ea rs o n co rr el a ti o n s; ce n su s d a ta fr o m th e 1 9 9 0 S F 3 ; N p 7 7 co m m u n it y a re a s. * P ! .1 0 . * * P ! .0 5 . * * * P ! .0 0 1 . This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp A P P E N D IX B T A B L E B 1 N e g a t iv e B in o m ia l R e g r e ss io n s: C o m m u n it y E c o n o m ic a n d S o c ia l C a p it a l a n d A n t ib l a c k H a t e C r im e s (A ll R ep o rt ed In ci d en ts ), 1 9 9 7 – 2 0 0 2 M o d el 1 M o d el 2 M o d el 3 M o d el 4 M o d el 5 M o d el 6 M o d el 7 M o d el 8 M o d el 9 M o d el 1 0 M o d el 1 1 C o n st a n t .. .. .. .. .. .. .. .. .. .. .. .. .. .. � 4 .3 6 * * � 5 .2 8 * * � 5 .2 9 * * � 4 .8 1 * * * � 4 .5 6 * * � 1 1 .9 9 * * � .8 .5 8 * * � 7 .8 3 * * � 7 .9 9 * * � 7 .6 2 * * � 5 .0 5 (1 .7 3 ) (1 .7 0 ) (1 .6 9 ) (1 .8 6 ) (1 .8 5 ) (2 .6 4 ) (2 .8 2 ) (2 .9 3 ) (3 .2 9 ) (3 .3 9 ) (4 .1 1 ) L n p o p u la ti o n 1 9 9 0 .. .. .. .. .. .. .. .. .. .5 7 * * .6 4 * * .6 6 * * .5 8 * * .5 9 * * .8 5 * * .6 7 * * .6 9 * * .6 4 * * * .6 5 * * * .5 7 * * * (. 1 7 ) (. 1 6 ) (. 1 6 ) (. 1 8 ) (. 1 8 ) (. 1 7 ) (. 1 8 ) (. 1 8 ) (. 1 8 ) (. 1 8 ) (. 1 7 ) S p a ti a l p ro x im it y .. .. .. .. .. .. .. .. .. .. .5 6 * * .2 8 .2 6 .2 8 .2 8 .3 1 .2 5 .2 7 .2 6 .2 6 .0 1 (. 2 6 ) (. 2 2 ) (. 2 3 ) (. 2 2 ) (. 2 2 ) (. 2 5 ) (. 2 2 ) (. 2 2 ) (. 2 1 ) (. 2 1 ) (. 1 9 ) D is a d v a n ta g e .. .. .. .. .. .. .. .. .. .. .. .. � .6 0 * * � .5 4 * * � .5 2 * * � .4 9 * * � .3 6 � .3 4 � .4 2 (. 1 3 ) (. 1 4 ) (. 2 5 ) (. 2 4 ) (. 2 6 ) (. 2 6 ) (. 2 7 ) S ta b il it y .. .. .. .. .. .. .. .. .. .. .. .. .. .. . � .0 5 � .0 3 � .0 4 � .0 3 � .1 2 � .1 1 � .1 1 (. 1 0 ) (. 1 1 ) (. 1 1 ) (. 1 1 ) (. 1 3 ) (. 1 4 ) (. 1 2 ) W h it e u n em p lo y m en t .. .. .. .. .. .. .. .. � .0 2 (. 0 2 ) % w h it e 1 9 9 0 .. .. .. .. .. .. .. .. .. .. .. .. .0 0 3 .0 1 * * .0 0 4 � .0 1 (. 0 1 ) (. 0 0 4 ) (. 0 1 ) (. 0 4 ) % b la ck 1 9 9 0 .. .. .. .. .. .. .. .. .. .. .. .. � .0 0 4 � .0 1 * * � .0 0 4 (. 0 1 ) (. 0 0 4 ) (. 0 1 ) % H is p a n ic 1 9 9 0 .. .. .. .. .. .. .. .. .. .. . .0 1 .0 0 2 .0 1 .0 0 3 .0 0 6 .0 0 2 .0 1 (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) In fo rm a l so ci a l co n tr o l .. .. .. .. .. .. .. 1 .3 4 * * 1 .1 5 * * 1 .1 7 * * 1 .1 2 * * 1 .1 1 * * .7 0 * * (. 4 7 ) (. 4 6 ) (. 4 5 ) (. 5 4 ) (. 5 4 ) (. 8 6 ) S o ci a l co h es io n .. .. .. .. .. .. .. .. .. .. .. � .0 5 � .4 8 � .4 4 � .4 3 � .4 3 � .5 3 (. 3 7 ) (. 4 1 ) (. 4 0 ) (. 4 4 ) (. 4 3 ) (. 3 7 ) This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp T A B L E B 1 (C o n ti n u ed ) M o d el 1 M o d el 2 M o d el 3 M o d el 4 M o d el 5 M o d el 6 M o d el 7 M o d el 8 M o d el 9 M o d el 1 0 M o d el 1 1 % ch a n g e in b la ck p o p u la ti o n , 1 9 9 0 – 2 0 0 0 .. .. .. .. .. .. .. .. .. .. .. .. � .5 7 * (. 3 2 ) In fo rm a l so ci a l co n tr o l # % w h it e .. . .0 0 5 (. 0 1 ) In fo rm a l so ci a l co n tr o l # % ch a n g e in b la ck p o p u la ti o n .. .. .. .. .. .. .. . .1 8 * (. 1 0 ) % w h it e 1 9 9 0 # % ch a n g e in b la ck p o p u la ti o n .. .. .. .. .. .. .. .. .. .. .. .. � .0 1 2 * (. 0 0 7 ) In fo rm a l so ci a l co n tr o l # % w h it e # % ch a n g e in b la ck .. .. .. .. .. .. .. .. . .0 0 3 * * (. 0 0 1 6 ) O v er d is p er si o n .. .. .. .. .. .. .. .. .. .. .. .7 8 .5 5 .5 4 .5 3 .5 4 .6 0 .5 0 .5 0 .4 8 .4 8 .2 6 (. 1 7 ) (. 1 3 ) (. 1 3 ) (. 1 3 ) (. 1 3 ) (. 1 4 ) (. 1 3 ) (. 1 3 ) (. 1 2 ) (. 1 2 ) (. 0 9 ) L o g li k el ih o o d .. .. .. .. .. .. .. .. .. .. .. . � 1 9 8 .6 7 1 8 8 .7 8 � 1 8 7 .8 5 � 1 8 8 .3 6 � 1 8 8 .2 7 � 1 9 1 .8 6 � 1 8 7 .5 7 � 1 8 6 .2 8 � 1 8 6 .1 4 � 1 8 6 .0 6 � 1 7 6 .4 0 N o t e .— N p 7 7 C h ic a g o co m m u n it y a re a s; u n st a n d a rd iz ed co ef fi ci en ts ; S E s in p a re n th es es . * P ! .1 0 . * * P ! .0 5 . * * * P ! .0 0 1 . This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp T A B L E B 2 N e g a t iv e B in o m ia l R e g r e ss io n s: C o m m u n it y C h a r a c t e r is t ic s a n d A n t iw h it e H a t e C r im e (A ll R ep o rt ed In ci d en ts ), 1 9 9 7 -2 0 0 2 A l l C h ic a g o a E x c l u d in g O u t l ie r sb M o d el 1 M o d el 2 M o d el 2 M o d el 3 M o d el 4 M o d el 5 M o d el 6 M o d el 7 M o d el 8 M o d el 9 M o d el 1 0 C o n st a n t .. .. .. .. .. .. .. .. .. .. . � 4 .4 2 * * � 4 .2 2 * * � 4 .7 2 * * � 4 .7 6 * * � 5 .0 1 * * � 4 .2 1 * * � 2 .8 6 � 4 .0 3 � 4 .2 9 � 6 .1 7 * � 5 .5 9 * (1 .6 2 ) (1 .6 8 ) (1 .5 7 ) (1 .5 7 ) (1 .6 1 ) (1 .6 5 ) (2 .3 7 ) (2 .7 5 ) (2 .6 0 ) (3 .2 2 ) (3 .1 3 ) L n p o p u la ti o n 1 9 9 0 .. .. .. .. .. .5 0 * * .4 8 * * .5 2 * * .5 2 * * .5 1 * * .5 1 * * .5 1 * * .5 6 * * .5 8 * * .5 4 * * .5 4 * * (. 1 6 ) (. 1 6 ) (. 1 5 ) (. 1 5 ) (. 1 6 ) (. 1 6 ) (. 1 6 ) (. 1 6 ) (. 1 6 ) (. 1 7 ) (. 1 6 ) S p a ti a l p ro x im it y .. .. .. .. .. .. .5 6 * * .5 8 * * .5 0 * * .5 1 * * .4 7 * * .4 8 * * .5 2 * * .4 4 * * .4 2 * * .4 5 * * .4 4 * * (. 2 1 ) (. 2 2 ) (. 2 1 ) (. 2 1 ) (. 2 1 ) (. 2 1 ) (. 2 0 ) (. 2 1 ) (. 2 1 ) (. 2 2 ) (. 2 2 ) E co n o m ic d is a d v a n ta g e .. .. .. .0 1 .1 9 * .1 8 � .0 1 � .0 8 .0 4 � .0 1 (. 1 1 ) (. 1 1 ) (. 1 2 ) (. 1 8 ) (. 1 9 ) (. 2 2 ) (. 2 2 ) R es id en ti a l st a b il it y .. .. .. .. .. � .0 5 � .1 8 * � .1 8 * � .2 3 * * � 2 4 * * � .2 6 * * � .2 7 * * (. 1 1 ) (. 1 1 ) (. 1 1 ) (. 1 1 ) (. 1 1 ) (. 1 3 ) (. 1 2 ) B la ck u n em p lo y m en t 1 9 9 0 .. . .0 0 3 (. 0 1 ) % w h it e 1 9 9 0 .. .. .. .. .. .. .. .. . � .0 1 * � .0 1 � .0 1 (. 0 0 6 ) (. 0 1 ) (. 0 1 ) % b la ck 1 9 9 0 .. .. .. .. .. .. .. .. . .0 1 .0 0 3 .0 1 (. 0 1 ) (. 0 0 ) (. 0 1 ) % H is p a n ic 1 9 9 0 .. .. .. .. .. .. . .0 0 3 � .0 0 4 -. 0 0 1 � .0 0 4 .0 0 4 � .0 0 3 (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) In fo rm a l so ci a l co n tr o l .. .. .. . � .8 8 * � .8 0 � .8 0 .4 1 .3 2 (. 5 2 ) (. 5 2 ) (. 5 2 ) (. 5 5 ) (. 5 5 ) S o ci a l co h es io n .. .. .. .. .. .. .. . � .2 7 � .1 4 � .0 1 .1 7 .2 7 (. 3 7 ) (. 4 3 ) (. 4 4 ) (. 4 6 ) (. 4 6 ) O v er d is p er si o n .. .. .. .. .. .. .. . .3 6 .3 6 .2 0 .2 0 .1 9 .1 9 .2 3 .2 1 .2 1 .1 9 .1 8 (. 1 4 ) (. 1 4 ) (. 1 1 ) (. 1 1 ) (. 1 1 ) (. 1 1 ) (. 1 2 ) (. 1 2 ) (. 1 2 ) (. 1 1 ) (. 1 1 ) L o g li k el ih o o d .. .. .. .. .. .. .. . � 1 4 7 .8 6 � 1 4 7 .7 6 � 1 3 0 .9 6 � 1 3 0 .9 2 � 1 2 9 .9 7 5 1 2 9 .3 1 � 1 3 2 .5 7 � 1 3 1 .9 9 � 1 3 1 .6 1 � 1 2 9 .8 7 � 1 2 9 .1 0 6 N o t e .— U n st a n d a rd iz ed co ef fi ci en ts ; S E s in p a re n th es es . a N p 7 7 C h ic a g o co m m u n it y a re a s. b N p 7 4 C h ic a g o co m m u n it y a re a s. * P ! .1 0 . * * P ! .0 5 . * * * P ! .0 0 1 . This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp T A B L E B 3 N e g a t iv e B in o m ia l R e g r e ss io n s: C o m m u n it y C h a r a c t e r is t ic s a n d “B o n a fi d e ” V io l e n t A n t ib l a c k H a t e C r im e s, 1 9 9 7 – 2 0 0 2 M o d el 1 M o d el 2 M o d el 3 M o d el 4 M o d el 5 M o d el 6 M o d el 7 M o d el 8 M o d el 9 M o d el 1 0 M o d el 1 1 C o n st a n t .. .. .. .. .. .. .. .. .. .. .. .. .. .. � 4 .7 9 * * � 6 .9 3 * * � 6 .8 5 * * � 6 .5 1 * * � 6 .1 8 * * � 1 3 .7 1 * * * � 1 0 .0 2 � 8 .9 5 * * � 8 .1 9 * � 7 .2 8 � 1 1 .9 5 * * (2 .1 9 ) (2 .3 5 ) (2 .3 5 ) (2 .5 4 ) (2 .4 8 ) (3 .6 7 ) (3 .8 9 ) (3 .9 3 ) (4 .4 3 ) (4 .4 6 ) (5 .8 1 ) L n p o p u la ti o n 1 9 9 0 .. .. .. .. .. .. .. .. .. .5 2 * * .7 1 * * .7 1 * * .6 2 * * .6 9 * * .8 6 * * * .6 4 * * .6 8 * * .6 2 * * .6 9 * * .6 6 * * (. 2 1 ) (. 2 2 ) (. 2 2 ) (. 2 4 ) (. 2 5 ) (. 2 4 ) (. 2 5 ) (. 2 4 ) (. 2 5 ) (. 2 5 ) (. 2 4 ) S p a ti a l p ro x im it y .. .. .. .. .. .. .. .. .. .. .7 3 * * .3 7 .3 6 .3 6 .3 6 .3 9 .3 3 .3 6 .3 5 .3 6 � .0 4 .3 3 (. 2 9 ) (. 2 9 ) (. 2 8 ) (. 2 8 ) (. 3 2 ) (. 2 8 ) (. 2 8 ) (. 2 8 ) (. 2 7 ) (. 2 6 ) D is a d v a n ta g e .. .. .. .. .. .. .. .. .. .. .. .. � 6 9 * * * � .6 5 * * � .4 4 � .1 9 � .3 2 � .1 3 � .0 1 (. 1 8 ) (. 2 0 ) (. 3 5 ) (. 3 3 ) (. 3 7 ) (. 3 6 ) (. 3 6 ) S ta b il it y .. .. .. .. .. .. .. .. .. .. .. .. .. .. . .0 8 .0 9 .1 2 .1 9 .0 5 .1 4 .2 4 (. 1 4 ) (. 1 4 ) (. 1 4 ) (. 1 5 ) (. 1 7 ) (. 1 8 ) (. 1 6 ) W h it e u n em p lo y m en t .. .. .. .. .. .. .. .. � .0 1 (. 0 2 ) % w h it e 1 9 9 0 .. .. .. .. .. .. .. .. .. .. .. .. .0 0 8 .0 1 * * .0 1 .0 8 (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 6 ) % b la ck 1 9 9 0 .. .. .. .. .. .. .. .. .. .. .. .. � .0 1 5 * � .0 2 * * � .0 2 * * (. 0 1 ) (. 0 1 ) (. 0 1 ) % H is p a n ic 1 9 9 0 .. .. .. .. .. .. .. .. .. .. . .0 1 � .0 3 3 .0 1 3 * * � .0 0 2 .0 1 1 � .0 0 4 .0 2 * * (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) In fo rm a l so ci a l co n tr o l .. .. .. .. .. .. .. 1 .3 3 * * 1 .2 1 * * 1 .2 1 * * .9 8 * .9 6 * 1 .9 1 * (. 6 1 ) (. 5 8 ) (. 5 7 ) (. 5 8 ) (. 5 9 ) (1 .0 8 ) S o ci a l co h es io n .. .. .. .. .. .. .. .. .. .. .. .1 2 � .3 9 � .4 1 � .5 1 � .5 7 � .6 9 (. 4 7 ) (. 5 3 ) (. 5 1 ) (. 5 6 ) (. 5 5 ) (. 4 9 ) This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp T A B L E B 3 (C o n ti n u ed ) M o d el 1 M o d el 2 M o d el 3 M o d el 4 M o d el 5 M o d el 6 M o d el 7 M o d el 8 M o d el 9 M o d el 1 0 M o d el 1 1 % ch a n g e in b la ck p o p u la ti o n , 1 9 9 0 – 2 0 0 0 .. .. .. .. .. .. .. .. .. .. .. .. � .1 0 (. 5 7 ) In fo rm a l so ci a l co n tr o l # % w h it e .. . � .0 2 (. 0 2 ) In fo rm a l so ci a l co n tr o l # % ch a n g e in b la ck p o p u la ti o n .. .. .. .. .. .. .. . .0 3 (. 1 7 ) % w h it e 1 9 9 0 # % ch a n g e in b la ck p o p u la ti o n .. .. .. .. .. .. .. .. .. .. .. .. � .0 3 * * (. 0 1 ) In fo rm a l so ci a l co n tr o l # % w h it e # % ch a n g e in b la ck .. .. .. .. .. .. .. .. . .0 1 * * (. 0 0 3 ) O v er d is p er si o n .. .. .. .. .. .. .. .. .. .. .. 1 .0 4 .6 5 .6 6 .6 0 .5 8 .7 7 .5 5 .5 7 .5 3 .5 6 .2 2 (. 3 0 ) (. 2 4 ) (. 2 4 ) (. 2 3 ) (. 2 2 ) (. 2 6 ) (. 2 1 ) (. 2 2 ) (. 2 1 ) (. 2 2 ) (. 1 6 ) L o g li k el ih o o d .. .. .. .. .. .. .. .. .. .. .. . � 1 3 7 .3 3 � 1 3 0 .3 3 � 1 2 9 .8 6 � 1 2 8 .7 8 � 1 2 7 .4 � 1 3 2 .6 1 � 1 2 6 .6 2 � 1 2 8 .0 8 � 1 2 6 .2 6 � 1 2 7 .6 7 � 1 1 9 .2 3 N o t e .— N p 7 7 C h ic a g o co m m u n it y a re a s; u n st a n d a rd iz ed co ef fi ci en ts ; S E s in p a re n th es es . * P ! .1 0 . * * P ! .0 5 . * * * P ! .0 0 1 . This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp T A B L E B 4 N e g a t iv e B in o m ia l R e g r e ss io n s: C o m m u n it y C h a r a c t e r is t ic s a n d “B o n a fi d e ” V io l e n t A n t iw h it e H a t e C r im e , 1 9 9 7 – 2 0 0 2 A l l C h ic a g o a E x c l u d in g O u t l ie r sb M o d el 1 M o d el 2 M o d el 2 M o d el 3 M o d el 4 M o d el 5 M o d el 6 M o d el 7 M o d el 8 M o d el 9 M o d el 1 0 C o n st a n t .. .. .. .. .. .. .. .. .. .. . � 4 .5 7 * * � 3 .7 5 * � 4 .0 0 * � 3 .9 6 * � 4 .0 1 * � 3 .5 4 1 .1 9 � .4 6 � .2 3 � 3 .6 1 � 3 .9 8 (2 .1 0 ) (2 .1 0 ) (2 .1 1 ) (2 .1 2 ) (2 .1 9 ) (2 .2 7 ) (3 .0 6 ) (3 .4 0 ) (3 .7 0 ) (4 .1 5 ) (4 .3 0 ) L n p o p u la ti o n 1 9 9 0 .. .. .. .. .. .4 6 * * .3 7 * .3 8 * .3 9 * .3 7 * .3 6 * .3 3 * .3 9 * .3 8 * .3 8 * .3 8 * (. 2 0 ) (. 2 0 ) (. 2 0 ) (. 2 0 ) (. 2 2 ) (. 2 2 ) (. 2 0 ) (. 2 1 ) (. 2 1 ) (. 2 2 ) (. 2 2 ) S p a ti a l p ro x im it y .. .. .. .. .. .. .7 0 * * .7 1 * * .5 6 * .5 6 * .5 5 * .5 5 * .6 2 * * .5 0 * .5 3 * .5 0 * .5 0 * (. 2 7 ) (. 2 8 ) (. 2 9 ) (. 2 9 ) (. 3 0 ) (. 3 0 ) (. 2 6 ) (. 2 8 ) (. 2 9 ) (. 3 0 ) (. 3 0 ) E co n o m ic d is a d v a n ta g e .. .. .. .2 4 * .3 8 * * .3 9 * * .3 0 .2 3 .2 5 .3 0 (. 1 4 ) (. 1 5 ) (. 1 7 ) (. 2 6 ) (. 2 6 ) (. 2 9 ) (. 2 9 ) R es id en ti a l st a b il it y .. .. .. .. .. � .1 6 � .2 6 * � .2 6 * � .2 8 * � .2 9 * � .3 0 * � .2 9 * (. 1 4 ) (. 1 5 ) (. 1 5 ) (. 1 6 ) (. 1 6 ) (. 1 7 ) (. 1 7 ) B la ck u n em p lo y m en t 1 9 9 0 .. . � .0 0 2 (. 0 1 ) % w h it e 1 9 9 0 .. .. .. .. .. .. .. .. . .0 0 3 .0 0 3 .0 1 (. 0 1 ) (. 0 1 ) (. 0 1 ) % b la ck 1 9 9 0 .. .. .. .. .. .. .. .. . � .0 0 6 � .0 0 6 � .0 1 (. 0 0 8 ) (. 0 1 ) (. 0 1 ) % H is p a n ic 1 9 9 0 .. .. .. .. .. .. . .0 0 3 .0 0 0 � .0 0 3 .0 0 0 .0 0 0 .0 0 5 (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) (. 0 1 ) In fo rm a l so ci a l co n tr o l .. .. .. . � 1 .2 3 * � 1 .2 2 * � 1 .2 2 * � .7 2 � .7 0 (. 7 0 ) (. 7 0 ) (. 7 0 ) (. 7 4 ) (. 7 4 ) S o ci a l co h es io n .. .. .. .. .. .. .. . � .0 3 .2 9 .1 6 .7 2 .6 3 (. 4 9 ) (. 6 0 ) (. 6 0 ) (. 6 2 ) (. 6 2 ) O v er d is p er si o n .. .. .. .. .. .. .. . .5 0 .4 4 .3 9 .3 9 .3 9 .3 9 .3 9 .3 5 .3 6 .3 2 .3 2 (. 2 5 ) (. 2 4 ) (. 2 4 ) (. 2 4 ) (. 2 4 ) (. 2 4 ) (. 2 5 ) (. 2 4 ) (. 2 4 ) (. 2 3 ) (. 2 3 ) L o g li k el ih o o d .. .. .. .. .. .. .. . � 1 1 3 .5 0 � 1 1 1 .3 9 � 1 0 2 .4 5 � 1 0 2 .4 4 � 1 0 2 .3 7 � 1 0 2 .2 1 � 1 0 3 .7 0 � 1 0 3 .1 7 � 1 0 3 .4 2 � 1 0 1 .6 5 � 1 0 1 .3 5 N o t e .— U n st a n d a rd iz ed co ef fi ci en ts ; S E s in p a re n th es es . a N p 7 7 C h ic a g o co m m u n it y a re a s. b N p 7 4 C h ic a g o co m m u n it y a re a s. * P ! .1 0 . * * P ! .0 5 . * * * P ! .0 0 1 . This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 860 REFERENCES Anselin, Luc. 1988. Spatial Econometrics: Methods and Models. Dordrecht: Kluwer Academic Publishers. ———. 1995. “Local Indicators of Spatial Association—LISA.” Geographical Analysis 27:93–115. Baller, Robert D., Luc Anselin, Steven Messner, Glenn Deane, and Darnell Hawkins. 2001. “Structural Covariates of U.S. County Homicide Rates: Incorporating Spatial Effects.” Criminology 39:561–90. Baller, Robert D., and Kelly K. Richardson. 2002. “Social Integration, Imitation, and the Geographic Patterning of Suicide.” American Sociological Review 67:873–88. Berk, Richard A., Elizabeth A. Boyd, and Karl M. Hamner. 1992. “Thinking More Clearly about Hate-Motivated Crimes.” Pp. 123–43 in Hate Crimes: Confronting Violence against Lesbians and Gay Men, edited by Gregory M. Herek and Kevin Berill. Newbury Park, Calif.: Sage. Berlet, Chip. 2001. “Hate Groups, Racial Tension and Ethnoviolence in an Integrating Neighborhood, 1976–1988.” Research in Political Sociology 9:117–63. Black, Donald D. 1983. “Crime as Social Control.” American Sociological Review 48: 34–45. Blackwell, James W. 1990. “Ethnic Inequality and the Rate of Homicide.” Social Forces 69:53–70. Bonacich, Edna. 1972. “A Theory of Ethnic Antagonism: The Split Labor Market.” American Sociological Review 37 (5): 547–59. Boyd, Elizabeth A., Richard A. Berk, and Karl M. Hamner. 1996. “‘Motivated by Hatred or Prejudice’: Categorization of Hate-Motivated Crimes in Two Police Divisions.” Law and Society Review 30:819–50. Browning, Christopher R. 2002. “The Span of Collective Efficacy: Extending Social Disorganization Theory to Partner Violence.” Journal of Marriage and Family 64: 833–50. Bruch, Elizabeth E., and Robert D. Mare. 2006. “Neighborhood Choice and Neighborhood Change.” American Journal of Sociology 112 (3): 667–709. Bursik, Robert J., Jr. 1988. “Social Disorganization Theories of Crime and Delinquency: Problems and Prospects.” Criminology 26:519–52. Clogg, Clifford, Eva Petkova, and Adamantios Haritou. 1995. “Statistical Methods for Comparing Regression Coefficients between Models.” American Journal of Sociology 100:1261–93. Coleman, James S. 1988. “Social Capital in the Creation of Human Capital.” American Journal of Sociology 94:S95–S120. Comstock, Gary D. 1991. Violence against Lesbians and Gay Men. New York: Columbia University Press. CPD (Chicago Police Department). 2005. “Hate Crimes in Chicago.” Chicago: Chicago Police Department. Craig, Kellina M. 1999. “Retaliation, Fear, or Rage: An Investigation of African American and White Reactions to Racist Hate Crime.” Journal of Interpersonal Violence 14:138–51. De Sena, Judith. 1990. Protecting One’s Turf: Social Strategies for Maintaining Urban Neighborhoods. Lanham, Md.: University Press of America. FBI (Federal Bureau of Investigation). 1999a. Hate Crime Statistics, 1998. Washington, D.C: Government Printing Office. ———. 1999b. Hate Crime Data Collection Guidelines. Washington D.C.: U.S. Department of Justice, Federal Bureau of Investigation. Grattet, Ryken, Valerie Jenness, and Timothy Curry. 1998. “The Homogenization and Differentiation of Hate Crime Law in the United States, 1978 to 1995: Innovation This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 861 and Diffusion in the Criminalization of Bigotry.” American Sociological Review 63 (2): 286–307. Green, Donald P., J. Glaser, and Andrew Rich. 1998. “From Lynching to Gay Bashing: The Elusive Connection between Economic Conditions and Hate Crime.” Journal of Personality and Social Psychology 75 (1): 82–92. Green, Donald P., Lawrence H. McFalls, and Jennifer K. Smith. 2001. “Hate Crime: An Emergent Research Agenda.” Annual Review of Sociology 27:479–504. Green, Donald P., Dara Z. Strolovitch, and Janelle S. Wong. 1998. “Defended Neighborhoods, Integration, and Racially Motivated Crime.” American Journal of Sociology 104:372–403. Hagan, John, Hans Merkens, and Klaus Boehnke. 1995. “Delinquency and Disdain: Social Capital and the Control of Right-Wing Extremism among East and West Berlin Youth.” American Journal of Sociology 100 (4): 1028–52. Heitgard, Janet, and Robert J. Bursik, Jr. 1987. “Extracommunity Dynamics and the Ecology of Delinquency.” American Journal of Sociology 92 (4): 775–87. Herek, Gregory M., and Kevin Berrill. 1992. Hate Crimes: Confronting Violence against Lesbians and Gay Men. Newbury Park, Calif.: Sage. Hovland, Carl I., and Robert R. Sears. 1940. “Minor Studies of Aggression, VI: Correlation of Lynching with Economic Indices.” Journal of Psychology 9:301–10. Iganski, Paul. 2002. The Hate Debate: Should Hate Be Punished as a Crime? London: Profile Books. Jacobs, David, and Katherine Wood. 1999. “Interracial Conflict and Interracial Homicide: Do Political and Economic Rivalries Explain White Killings of Blacks or Black Killings of Whites?” American Journal of Sociology 105:157–90. Jacobs, James, and Kimberly Potter. 1998. Hate Crimes: Criminal Law and Identity Politics. New York: Cambridge University Press. Jenness, Valerie, and Kendal Broad. 1997. Hate Crimes: New Social Movements and the Politics of Violence. New York: Gruyter. Jenness, Valerie, and Ryken Grattet. 1996. “The Criminalization of Hate: A Comparison of Structural and Polity Influences on the Passage of ‘Bias-Crime’ Legislation in the United States.” Sociological Perspectives 39 (1): 129–54. ———. 2001. Making Hate a Crime: From Social Movement to Law Enforcement. New York: Russell Sage. Kasarda, John, and Morris Janowitz. 1974. “Community Attachments in Mass Society.” American Sociological Review 39:328–39. Kornhauser, Ruth. 1974. Social Sources of Delinquency. Chicago: University of Chicago Press. Levin, Brian. 2002. “From Slavery to Hate Crime Laws: The Emergence of Race and Status-Based Protection in American Criminal Law.” Journal of Social Issues 58 (2): 227–45. Levin, Jack, and Jack McDevitt. 1993. Hate Crimes: The Rising Tide of Bigotry and Bloodshed. New York: Plenum. Martin, Susan. 1995. “A Cross-Burning Is Not Just an Arson: Police Social Construction of Hate Crimes in Baltimore County.” Criminology 33:303–26. Massey, Douglas S. “The Prodigal Paradigm Returns: Ecology Comes Back to Sociology.” Pp. 41–48 in Does It Take a Village? Community Effect on Children, Adolescents, and Families, edited by Alan Booth and Ann Crouter. Mahwah, N.J.: Lawrence Erlbaum Associates. Matsueda, Ross L. 1988. “The Current State of Differential Association Theory.” Crime and Delinquency 34 (3): 277–306. McDevitt, Jack, J. Balboni, L. Garcia, and J. Gu. 2001. “Consequences for Victims: A Comparison of Bias- and Non-Bias Motivated Assaults.” American Behavioral Scientist 45:697–713. McDevitt, Jack, J. Balboni, J. Nolan, and J. Weiss. 2000. Improving the Quality and This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp American Journal of Sociology 862 Accuracy of Bias Crime Statistics Nationally. Washington, D.C.: U.S. Department of Justice. McVeigh, Rory, Michael Welch, and Thoroddur Bjarnason. 2003. “Hate Crime Reporting as a Successful Social Movement Outcome.” American Sociological Review 68:843–67. Messner, Steven F., Suzanne McHugh, and Richard B. Felson. 2004. “Distinctive Characteristics of Assaults Motivated by Bias.” Criminology 42:582–618. Morenoff, Jeffrey D., Robert J. Sampson, and Stephen W. Raudenbush. 2001. “Neighborhood Inequality, Collective Efficacy and the Spatial Dynamics of Urban Violence.” Criminology 39 (3): 517–60. Myers, Samuel L., Jr. 1980. “Why Are Crimes Underreported? What Is the Crime Rate? Does It Really Matter?” Social Science Quarterly 61 (1): 23–43. Newton, Michael, and Judy Ann Newton. 1991. Racial and Religious Violence in America: A Chronology. New York: Garland. Nolan, James J. III, Yoshio Akiyama, and Samuel Berhanu. 2002. “The Hate Crime Statistics Act of 1990: Developing a Method for Measuring the Accuracy of Hate Violence.” American Behavioral Scientist 46 (1): 136–53. Olzak, Susan. 1990. “The Political Context of Competition: Lynching and Urban Racial Violence, 1882–1914.” Social Forces 69 (2): 395–421. ———. 1992. The Dynamics of Ethnic Competition and Conflict. Palo Alto, Calif.: Stanford University Press. Perry, Barbara. 2001. In the Name of Hate: Understanding Hate Crime. New York: Routledge. ———. 2002. “Defending the Color Line: Racially and Ethnically Motivated Hate Crime.” American Behavioral Scientist 46 (1): 72–92. Phillips, Scott, and Ryken Grattet. 2000. “Judicial Rhetoric, Meaning-Making, and the Institutionalization of Hate Crime Law.” Law and Society Review 34 (3): 567–606. Pinderhughes, Howard. 1993. “The Anatomy of Racially Motivated Violence in New York City: A Case Study of Youth in Southern Brooklyn.” Social Problems 40: 478–92. Portes, Alejandro. 1998. “Social Capital: Its Origins and Applications in Modern Sociology.” Annual Review of Sociology 24:1–24 Putnam, Robert D. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster. Quillian, Lincoln, and Devah Pager. 2001. “Black Neighbors, Higher Crime? The Role of Racial Stereotypes in Evaluations of Neighborhood Crime.” American Journal of Sociology 107:717–67. Rieder, Jonathan. 1985. Canarsie: The Jews and Italians of Brooklyn against Liberalism. Cambridge Mass.: Harvard University Press. Sampson, Robert J. 1999. “What ‘Community’ Supplies.” Pp. 241–92 in Urban Problems and Community Development, edited by Ronald F. Ferguson and William T. Dickens. Washington, D.C.: Brookings Institution Press. ———. 2002a. “Organized for What? Recasting Theories of Social (Dis)Organization.” Pp. 95–110 in Crime and Social Organization, edited by Elin Waring and David Weisburd. New Brunswick, N.J.: Transaction. ———. 2002b. “Transcending Tradition: New Directions in Community Research, Chicago Style.” Criminology 49 (2): 213–30. Sampson, Robert, and Byron Groves. 1989. “Community Structure and Crime: Testing Social Disorganization Theory.” American Journal of Sociology 95:774–802. Sampson, Robert J., Jeffrey Morenoff, and Felton Earls. 1999. “Beyond Social Capital: Spatial Dynamics of Collective Efficacy for Children.” American Sociological Review 64:633–60. Sampson, Robert J., and Stephen Raudenbush. 2004. “Seeing Disorder: Neighborhood This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp Racially Motivated Crime 863 Stigma and the Social Construction of ‘Broken Windows.’” Social Psychology Quarterly 67:319–42. Sampson, Robert, Stephen Raudenbush, and Felton Earls. 1997. “Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy.” Science 277:918–24. Sampson, Robert J., and William J. Wilson. 1995. “Toward a Theory of Race, Crime, and Inequality.” In Crime and Inequality, edited by John Hagan and Ruth Peterson. Stanford, Calif.: Stanford University Press. Savelsberg, Joachim J., and Ryan D. King. 2005. “Institutionalizing Collective Memories of Hate: Law and Law Enforcement in Germany and the United States.” American Journal of Sociology 111:579–616. Shaw, Clifford R., and Henry D. McKay. 1942. Juvenile Delinquency in Urban Areas. Chicago: University of Chicago Press. Skogan, Wesley G. 1984. “Reporting Crimes to the Police: The Status of World Research.” Journal of Research in Crime and Delinquency 21:113–37. Soule, Sarah A. 1992. “Populism and Black Lynching in Georgia, 1890–1900.” Social Forces 71 (2): 431–49. Soule, Sarah A., and Nella van Dyke. 1999. “Black Church Arson in the United States, 1989–1996.” Ethnic and Racial Studies 22 (4): 724–42. Statistics Canada. 1999. Hate Crime in Canada: An Overview of Issues and Data Sources. Canadian Centre for Justice Statistics, catalogue 85-551-XIE. Strom, Kevin J. 2001. Hate Crimes Reported in NIBRS, 1997–1999. Bureau of Justice Statistics Special Report, U.S. Department of Justice, NCJ 186765. Sutherland, Edwin H. 1947. Principles in Criminology, 4th ed. Philadelphia: Lippincott. Suttles, Gerald D. 1972. The Social Construction of Communities. Chicago: University of Chicago Press. Thomson, Irene Taviss. 2005. “The Theory that Won’t Die: From Mass Society to the Decline of Social Capital.” Sociological Forum 20:421–48. Tolnay, Stewart E., and E. M. Beck. 1995. A Festival of Violence: An Analysis of Southern Lynchings, 1882–1930. Urbana: University of Illinois Press. van Dyke, Nella, Sarah A. Soule, and Rebecca Widom. 2001. “The Politics of Hate: Explaining Variation in the Incidence of Anti-gay Hate Crime.” Research in Political Sociology 9:35–58. Waldinger, Roger. 1995. “The ‘Other Side’ of Embeddedness: A Case-Study of the Interplay of Economy and Ethnicity.” Ethnic and Racial Studies 18 (3): 555–80. Weitzer, Ronald, and Steven A. Tuch. 2005. “Racially Biased Policing: Determinants of Citizen Perceptions.” Social Forces 83 (3): 1009–30. Wisconsin v. Mitchell. 1993. 508 U.S. 476. This content downloaded from 128.95.71.159 on Thu, 2 Apr 2015 21:41:18 PM All use subject to JSTOR Terms and Conditions http://www.jstor.org/page/info/about/policies/terms.jsp