Racial Threat,Suspicion,and Police Behavior

Abstract

Racial bias in traffic enforcement has become a popular line of inquiry, but examinations into explanations for the disparity have been scant. The current research integrates theoretical insights from the racial threat hypothesis with inferences drawn from the empirical analyses of the factors that stimulate officer suspicion. The most intriguing finding from this beat-level examination of the structural predictors of several traffic stop outcome measures concerns the conditional effect of the racial composition of the beat on search rates. The analyses reveal that the search rate increases in areas where the proportion of Black residents is higher; however, this finding is observed only for White motorists. This finding is interpreted as indicating that structural characteristics of an area can provide cues to officers regarding who belongs in that environment. As a result, social control increases among groups whose racial characteristics are inconsistent with the neighborhood racial composition.

Keywords

policing racial threat suspicion racial profiling

Public and political awareness of racial disparity and bias in traffic enforcement has increased dramatically over the past decade. The terms racial profiling and driving while Black have entered the public lexicon and have rekindled interest in the causal processes that affect police behavior during encounters with people of color. Though the exact definition of racial profiling varies (Batton & Kadleck, 2004), at its core it involves police officers’ differentially targeting people of color for traffic stops, searches, and encounter dispositions. That is to say, race rather than behavior is seen as the primary factor that directs officers’ decision making during self-initiated encounters (Ramirez, McDevitt, & Farrell, 2000).

Regardless of how one defines racial profiling, there is a strong consensus among the public that racial profiling exists and that it is widespread (Gallup, 1999; Reitzel & Piquero, 2006; Reitzel, Rice, & Piquero, 2004). In addition to triggering legislative action, this perception has motivated governmental officials to sponsor numerous data collection projects to determine whether disparity exists within their jurisdictions (Withrow, 2006). The result is a proliferation of data being analyzed by researchers and forwarded to the public and to policy makers for their consumption. Research has focused on how to collect data, what information should be collected, how to analyze data, and what conclusions to draw from the results (Fridell, 2004). However, little attention has been given to why disparity is observed and whether disparity is observed because of discriminatory practices by the police or because of more legitimate reasons, such as deployment of organizational resources (Withrow, 2006), and few have forwarded coherent theoretical models to guide this inquiry.

In particular, little attention has focused on how the social environment and the characteristics of the citizens jointly effect the exercise of formal governmental crime control. More often than not, officers develop understandings of what types of people and behavior are “normal” within the areas they work (Klinger, 1997). Factors such as crime, racial composition, neighborhood distress, and previous experiences with citizens in communities shape officers’ understandings of what is expected and what (or who) belongs. A mismatch between what is expected and what is observed heightens an officer’s sense of awareness and increases his or her suspicion that criminal activity may be afoot. Officer-initiated encounters between the police and the public may be less a result of the citizens’ race alone and more a product of characteristics and behavior that are unexpected given the makeup of the local environment. It is important to consider whether citizens create suspicion because their race is inconsistent with that of the typical population of the area, because this would result in greater levels of social control being applied to Blacks in predominately non-Black areas, as well as more social control of Whites in increasingly non-White areas.

Drawing on racial threat conflict theory, as well as an understanding of the social cues that modulate police suspicion, this investigation extends our understanding of the influence of race on police behavior during traffic encounters. The decision to stop a motorist represents a limited, though important, mobilization of police resources, whereas the post-stop behavior decision to search or issue a citation or ticket represents a significant application of formal social control. Utilizing data collected from traffic stops in a large urban police department over a 9-month period, this article examines how the social structure of neighborhoods influences total and race-specific rates of stops, searches, and citations.

An Overview of Racial Profiling Research

Research on the racial profiling phenomena has been widespread. A comprehensive overview of the research on racial profiling is beyond the scope of the current inquiry because such information can be found elsewhere (Engel, Calnon, & Bernard, 2002; Fridell, 2004; Withrow, 2006). Rather, this section provides an overview of the racial disparity–traffic enforcement literature.

There is little doubt that racial minorities are disproportionately stopped by the police. In an early review of the state of traffic enforcement research, Engel et al. (2002) reported that all the published research to date, a total of 13 reports, demonstrates some level of disparity. A subsequent overview by Withrow (2006) indicates that 18 of 23 published studies demonstrate disparity in the rate in which Blacks are stopped by the police. This conclusion appears to be consistent with a Bureau of Justice Statistics analysis that revealed that whereas Blacks compose 9.8% of licensed drivers, they represent 11.6% of all drivers stopped by the police at least once in 1999 and 13.7% of drivers stopped more than once (Langan, Greenfeld, Smith, Durose, & Levin, 2001).

Racial disparities in stopping and/or ticketing have been found to be present in urban jurisdictions. Urban areas tend to have larger numbers and concentrations of racial minorities than do nonurban areas, and perhaps this concentration of threat populations increases the application of social control that focuses on members of racial minority groups. Harris (1999) reported that based on the examination of interviews with citizens and official records collected by several large departments in Ohio, Blacks are between 2.32 and 2.76 times more likely to be ticketed than Whites. M. R. Smith and Petrocelli (2001) found that minorities are 1.16 times more likely than Whites to be stopped by the police. Last, Rojek, Rosenfeld, and Decker (2004) examined 92 cities in Missouri, including large cities, suburban organizations, and other medium-sized cities, and all but one reported an overrepresentation of Blacks being stopped.

Suburban departments do not appear immune from the racial-profiling controversy. Meehan and Ponder (2002) noted that Blacks are more likely than Whites to be stopped in a predominately White Midwestern suburb; in fact, they found disparity to be greatest in the sections of the city characterized as the highest proportion of White residential population. However, Novak (2004) found a small but statistically insignificant overrepresentation of Blacks stopped in a suburban environment.

The decision to stop is merely one method by which the police can differentially enforce legal codes. Therefore, to more fully capture the role that race plays in the application of the law, we review studies that examine the correlates of several important post-stop behaviors.

Consider the decision to conduct a search. A large national survey of citizens indicates that searches occur during approximately 6.6% of all traffic stops; however, the likelihood of a search varied on the basis of the race of the driver: 11.0% of Blacks and 11.3% of Hispanics indicate being searched during traffic stops, whereas Whites report being searched only 5.4% of the time (Langan et al., 2001). Explorations into disparity in behavior during traffic enforcement indicate that Blacks are more likely than Whites to be searched during traffic encounters (Cox, Pease, Miller, & Tyson, 2001; Knowles, Persico, & Todd, 2001; Rojek et al., 2004; Schafer, Carter, Katz-Bannister, & Wells, 2006; Zingraff et al., 2000). Note, however, that others have found race to be unrelated to likelihood of conducting a search (Paoline & Terrill, 2005; M. R. Smith & Petrocelli, 2001).

There appears to be even less consensus in the empirical literature regarding the propensity for the police to issue formal citations or to arrest people of color. Survey research compiled by Langan and colleagues (2001) indicates that Blacks and Hispanics are more likely than Whites to be ticketed or arrested. Other studies find that people of color are significantly more likely than Whites to be arrested or issued a ticket or citation (Cox et al., 2001; Rojek et al., 2004). R. A. Brown and Frank (2005) reported that suspect race was unrelated to whether the driver is given a warning or a citation; however, Blacks are 14 times more likely than Whites to be physically arrested than issued a summons in lieu of arrest. Still other research indicates that Blacks are significantly less likely than Whites to be formally sanctioned, even after controlling for theoretically relevant factors (Novak, 2004; M. R. Smith & Petrocelli, 2001; Zingraff et al., 2000). The latter pattern of findings may perhaps be explained as a way for officers to correct for overrepresentation of minorities in earlier phases of the traffic stop. The decision to stop and search is not only highly discretionary but largely invisible from review because no paper trail is created pursuant to these low-visibility behaviors. This is not the case for citations, tickets, and arrests. M. R. Smith and Petrocelli (2001) suggested that this scenario might occur because officers may be more likely to stop minorities for reasons that would not support the issuance of a traffic summons or arrest (see also Fagan & Davies, 2000). The researchers concluded that high stop rates indicate that officers had broadened their suspicion for minorities.

In sum, the extant research seems to suggest that race may play an important role in the decision of whether to stop, search, and/or sanction an individual during the series of events that encompass a traffic stop encounter. What is not clear are the causal mechanisms that may be generating these effects. As we explicate below, we believe that racial conflict theory can provide some insight into this matter.

Conflict Theories of Social Control

The conflict perspective broadly conceptualizes social control as an instrument used by social elites to control those behaviors and populations that threaten their interests (Black, 1976; Blalock, 1967; Liska, 1992; Quinney, 1970, 1977; Turk, 1969). Although there is considerable disagreement concerning the identity of so-called elites, let alone the manner in which they are able to co-opt and influence institutions of social control (Bierne, 1979; Liska, 1987, 1997), there is a common theme that unifies this literature. Although the terminology may differ, virtually all proponents of conflict theory aver that the social control of dangerous groups and strata, rather than the mere suppression of illegal activities, represents the primary objective of crime control institutions (Liska, 1997; Quinney, 1970; Turk, 1969).

As a research proposition, the racial conflict perspective of social control reduces to what is commonly referred to as the threat hypothesis (Jackson, 1986; Liska, 1997). Drawing on research that indicates that the more powerful segments of society view racial and ethnic minorities as a criminal threat (Liska, Lawrence, & Sanchirico, 1982; Skogan, 1995; Swigert & Farrell, 1977), proponents of racial conflict theory believe that there is a strong connection between the racial composition of communities and crime control. Specifically, they posit that a high percentage of non-Whites (especially, Blacks) produce an emergent property, a perceived threat of crime, which motivates a White majority to pressure local authorities to increase the size and/or aggressiveness of legal institutions.

The racial threat hypothesis has spawned a substantial body of empirical research on the relationship between the racial composition of social aggregates and police-related outcome variables. Indeed, we have been able to locate an array of studies that examine the macro-social determinants of policing resources (expenditures and personnel) and police behavior (use of force and arrests) at various spatial levels of aggregation (cities, metropolitan areas, and states) and temporal levels (cross-sectional, panel, annual time series). Given these differences in research design, as well as in the model specifications, there should be little surprise that support for the racial conflict perspective has been mixed (Eitle, D’Alessio, & Stolzenberg, 2002; Greenberg, Kessler, & Loftin, 1985; Jackson & Carroll, 1981; Jacobs, 1979; Jacobs & Helms, 1999; Liska & Chamlin, 1984; Parker, Stults, & Rice, 2005).

Unfortunately, we were able to find only two studies that examine the relationship between the relative population size of racial minorities and police behavior at a smaller level of aggregation. Specifically, Petrocelli, Piquero, and Smith (2003) examined the causal impact of the percentage of Blacks on traffic stops, searches, and arrest measures across census tracts within Richmond, Virginia. The pattern of the results is less than clear. Consistent with the racial conflict theory, their findings indicate that the percentage of Blacks is positively related to the percentage of stops that led to a search. However, contrary to the racial threat hypothesis is their finding that percentage Black has no effect on the stop rate and is negatively associated with the percentage of stops that lead to at least one arrest or summons.

How is one to interpret these seemingly incongruous findings? As we attempt to explain below, we suspect that the community-level effect of race on the exercise of crime control may be more complex than what can be captured in simple additive model. That is to say, the relationship between the racial composition of neighborhoods and police behavior may depend on the level of suspicion that race engenders.

Suspicion and Social Control

Suspicion is a critical element into the exploration of race-based decision making by the police. Suspicion is part of the working personality and defining characteristic of the police (Herbert, 1997; Rubenstein, 1973; Skolnick, 1994), and it is a critical element of police culture (Crank, 1998). Officers are trained and socialized to identify suspicious behaviors and characteristics that are consistent with lawbreaking. This often includes behavioral cues, and it is a function of police officers’ learning what type of behavior is common among those involved in criminality. In seminal work on police and in a discussion of the symbolic assailant, Skolnick (1994) outlined several of these behaviors, including driving automobiles that do not look right, attempting to evade the police, being visibly rattled when near a police officer, loitering where children play, and wearing coats on hot days. These behaviors, though innocent and innocuous, can symbolize threat to police officers. As Crank (1998) aptly concluded, suspicion is “the talent of ironically converting safe areas into perilous ones” (p. 102), which in turn mobilizes police behavior.¹

The race of an individual is thought to have a significant impact on the formulation of suspicion, though few researchers have directly addressed this issue. Targeting minorities is explained as a function of the differential criminal involvement of the people in these groups. This is particularly the case for drug possession, weapon possession, and outstanding warrants. Police officers, perceiving Blacks to be differentially involved in these activities, target them accordingly (Harris, 2002). Some have pointed to the drug courier profile as the genesis of using race as a cue for criminality (Withrow, 2006). All things being equal, people of color may be perceived as being more suspicious than Whites, thus explaining their frequent contact with the police. Alpert, MacDonald, and Dunham (2005) expanded on this point to include the impact of race and location. They posited that when race and behavior are incongruent with a given environment, an officer’s suspicions become aroused. Alpert and colleagues’ empirical analysis reveals that although Black citizens are more likely to elicit nonbehavioral suspicions by officers, the racial composition of the neighborhood has no significant influence on the formulation of suspicion.

The formulation of suspicion may be triggered by more than the racial characteristics of individuals. Nonbehavioral characteristics, including race, may give rise to suspicion when coupled with the characteristics of the neighborhood. To continue with Skolnick’s discussion (1994), a key ingredient in the formulation of suspicion is to “look for the unusual,” which includes “persons who do not ‘belong’ where they are observed” (p. 44).

Minorities who are observed in predominately White neighborhoods may raise the suspicion of officers, thereby leading to officer mobilization. M. K. Brown (1981) commented that police are more likely to become suspicious and stop an individual if his or her race is incongruent with the racial composition of the environment. Brown offered examples of incongruity that leads to suspicion of deviance: “A young boy driving a new car is considered to be an adequate indicator of a potential car thief. The proverbial Black man in the all-White neighborhood late at night needs no comment” (p. 170). Additionally, Withrow (2004b, 2006) found that Blacks are consistently overrepresented among traffic stops in beats where the residential population is predominately White and that Whites tend to be overrepresented among those who are stopped in beats that are predominately populated by Blacks. Officers expect drivers within an area to be representative of those who live or utilize the roadways in the beat, and drivers who do not fit the racial characteristics of what is expected in that area are perceived as being suspicious. Knowledge of what is “normal” within the community provides officers a context to understand what is inconsistent within an area. Thus, it appears that racial groups who are perceived to be “out of place” draw additional attention by officers. Withrow (2006) went on to explain,

This may explain why a Black citizen in a White neighborhood attracts the attention and suspicion of a police officer and why a White citizen in a Black neighborhood attracts the attention and suspicion of a police officer. Both situations are equally inconsistent within the contexts of their locations. (p. 128)

Withrow’s theory of contextual attentiveness (2006) suggested that officers understand what is “usual, customary or expected” within an area and that they are “differentially attentive toward individuals or behaviors that appear inconsistent with predetermined conceptualizations” (p. 127) of what is expected. If incongruity is observed, then officers may seek out reasons to encounter this individual.

Related, Meehan and Ponder (2002) observed that within a predominately White suburban area adjacent to a larger, predominately Black urban area, Black drivers are more likely to experience social control depending on environmental context. Blacks are more likely to be stopped or be the subject of mobile data terminal queries (which is a proxy for criminal suspicion) in predominately White areas of the city that happen to be located further away from the larger Black urban city. Officers characterize Black drivers in predominately White areas to be out of place; therefore, the mere presence appears to be suspicious.

In short, the integration of the racial conflict perspective with an understanding of suspicion within the working personalities of officers can provide a new perspective into police–public traffic encounters. Racial conflict theory hypothesizes that as the relative population size of racial and/or ethnic minorities within macro-social units increases, so too does the application of social control to members of these subordinate groups. However, as explicated above, the culture-of-suspicion literature suggests that the macro-level relationship between the minority group population size and crime control outcomes might be tempered by the dynamics of neighborhood racial composition and the race of individuals encountered in these areas. Blacks encountered in predominately White neighborhoods may draw the suspicion of officers in a way similar to Whites who are observed in predominately Black neighborhoods. The statistical uniqueness of this race-out-of-place dynamic causes officers to treat people differently when their characteristics are inconsistent with what is expected in the environment. This research is positioned to reconcile these somewhat-competing perspectives by examining police application of social control across race.

Method

Kansas City, Missouri, served as the location for the research presented in this study. Kansas City is a large Midwestern city of approximately 441,000 people across 320 square miles, and it has approximately 1,300 sworn officers. In 2004, Kansas City had a crime rate slightly higher than that of comparably sized cities, with 7,338 Index I crimes, including homicides, rapes, robberies, burglaries, aggravated assaults, larcenies, motor vehicle thefts, and arsons (Kansas City Police Department, 2005). At the time that data for this study were collected, there were 780 sworn employees at the rank of patrol officer, and most were assigned to general patrol duties in one of the five patrol divisions. Additional officers were assigned to the traffic division, and their duties primarily consisted of patrolling interstates and highways that run within the city. Officers assigned to the patrol divisions as well as the traffic division were responsible for conducting routine traffic stops.

Pursuant, in part, to Missouri law requiring all law enforcement entities in the state to detect patterns of racial profiling and bias in traffic enforcement, the Kansas City Police Department has collected data on police–public contacts during traffic stops since 2003. The department requires officers to record a variety of information not mandated by statute, including the beat in which the encounter occurred.

The police beat is the unit of analysis for the present examination. During the period of this investigation, Kansas City had 70 police beats. We believe that this is an appropriate unit of analysis for the current inquiry because (a) policing is an inherently local function, (b) researchers have commented on how environment shapes police behavior (Klinger, 1997; D. A. Smith, 1986), and (c) the exercise of social control is expected to vary by place. These patrol beats tended to vary in size owing to population density, calls for service, and overall crime. As such, these units exhibited a great deal of variation in size. The typical beat was 4.50 square miles (7.24 km), but beat size ranged from 0.33 square miles ([0.53 km] in the central business district) to 41.40 square miles ([66.63 km] in an outlying area).

The police conducted 106,267 traffic stops between January 1 and September 30, 2004. Officers recorded location, date, time, demographic characteristics of the party involved, nature of the encounter, officer behaviors during the encounter, and ultimate disposition.²

These stops represent discretionary proactive officer behavior and should be unaffected by race of the driver. Excluded from examination here is what the department called investigatory stops. In these situations, officers are stopping motorists because they (or their vehicles) fit the description of those who are wanted for questioning by the police. Race may be used as a factor that influences the stop of a vehicle during investigatory stops, without the action being considered discriminatory.³

The stops utilized in the proceeding analyses involve persons stopped by the police merely for violations of traffic laws, including speeding, failing to signal a lane change, driving left of center, and having equipment violations. These encounters are proactive, and officers possess a great deal of discretion regarding whether to stop the citizen and whether to engage in any post-stop behavior.

Only those traffic stops that occurred on surface streets are included in the analysis. The interstate system is designed to move traffic through and around the metropolitan area in a quick fashion, and access onto and off of interstates is necessarily limited. These interstates pass through a number of communities and beats, but there are not always opportunities to exit to these communities (or to access the highway from these communities). Although we posit that neighborhood characteristics affect officers’ decision making, it seems unrealistic to assume that these factors influence decision making on interstates. It is theoretically inconsistent to believe that community characteristics shape the application of social control of individuals who travel on arterial interstates. Furthermore, surface streets are more likely to be populated by persons who live and work in the community, rather than those who quickly pass through on the elevated interstate that runs through it. In short, the traffic stop that occurs on interstates is qualitatively different from that which occurs on surface streets. Consequently, they are excluded from the analysis. A total of 52,165 traffic stops occurring on surface streets were aggregated to the beat level.

Dependent Variables

As we discuss above, the analyses focus on three decision nodes associated with traffic stops: traffic stop rates, searches rates, and arrest/citation rates. Table 1 presents the descriptive statistics of these endogenous variables, as well those for the predictors.

Table 1.

Descriptive Statistics

Variable	Description	Range	M	SD
Dependent variables
Traffic stop rate	Log of traffic stop rate per 1,000	3.38–8.31	4.76	0.83
Traffic stop rate of Blacks	Log of traffic stop rate of Blacks per 1,000	3.34–8.56	5.08	0.88
Traffic stop rate of Whites	Log of traffic stop rate of Whites per 1,000	3.32–9.32	5.16	1.32
Search rate	Log of search rate per 1,000	0.28–6.02	2.05	1.00
Search rate of Blacks	Log of search rate of Blacks per 1,000	0.73–6.32	2.68	0.78
Search rate of Whites	Log of search rate of Whites per 1,000	0.87–6.84	1.97	1.54
Citation rate	Log of citation rate per 1,000	3.11–7.97	4.52	0.87
Citation rate of Blacks	Log of citation rate of Blacks per 1,000	3.04–8.20	4.79	0.96
Citation rate of Whites	Log of citation rate of Whites per 1,000	3.10–9.18	4.97	1.33
Independent variables
Percentage Black	Percentage Black population	0.02–0.99	0.41	0.33
Control variables
White-to-Black economic inequalities	Factor of White-to-Black unemployment; White-to-Black income	−0.99–4.35	0.00	1.00
Workload	Factor of calls for services; Index I property; Index I violent crime	–2.34–2.94	0.00	1.00
Social disorganization	Factor of percentage single-parent households; percentage poverty	−1.58–2.33	0.00	1.00
Population mobility	Factor of percentage living in residence less than 5 years; percentage rental units	−1.58–2.71	0.00	1.00
Traffic stop rate	Traffic stop rate per 1,000 residents	29.49–4,084.53	203.22	491.25
Black traffic stop rate	Black traffic stop rate per 1,000 Black residents	28.21–5,196.25	293.38	653.51
White traffic stop rate	White traffic stop rate per 1,000 White residents	27.74–11,111.11	517.29	1,407.22

Total traffic stop rate is computed as the total number of traffic stops divided by beat population, for 1,000 residents. To discern whether the effect of the racial composition is conditioned by the racial characteristics of drivers, we also calculated race-specific stop rates. The Black stop rate is the total number of traffic stops involving Black motorists divided by the number of Blacks living within the beat, multiplied by 1,000; the White stop rate is the number of traffic stops involving White drivers divided by the total number Whites living in the beat, multiplied by 1,000.

Search rates were calculated from the total number of traffic stops observed in the beat. The total search rate is the number of traffic stops in which an officer conducted a search divided by the total number of traffic stops in that area, multiplied by 1,000. Black and White search rates were calculated in similar fashions in that the Black search rate represents the number of Black drivers searched divided by the number of Black drivers stopped, multiplied by 1,000. White search rates were the number of White drivers searched divided by the total number of White drivers stopped, multiplied by 1,000.

Citation rates were also calculated from the number of traffic stops in each beat. The total citation rate is the number of stops resulting in a ticket divided by the number of traffic stops, multiplied by 1,000. Black citation rate is estimated as the number of Blacks cited during a traffic stop divided by the total number of Blacks stopped, multiplied by 1,000. White citation rates are measured as the number of Whites cited during a traffic stop divided by the number of Whites stopped overall, then multiplied by 1,000. Each dependent measure was observed to be highly skewed. Therefore, we transformed each measure by its natural logarithm to achieve a normal distribution.⁴

Independent Variables

Social and demographic characteristics were estimated using U.S. Census data from 2000. Black population is calculated as the number of Blacks residing in a beat divided by the beat population. As can be seen from Table 1, there exists wide variation in the racial composition of beats in this city, ranging from 2% Black population to 99% Black population. This variation in the racial demographics of social units is typically found in large urban jurisdictions.

It is necessary to simultaneously control for variables that should reasonably influence police officer discretion. The economic conflict perspective contends that economic disadvantage in general and ascribed inequalities in particular are associated with a greater reliance on formal social control (Blau & Blau, 1982; Jacobs, 1979; Parker et al., 2005; B. W. Smith & Holmes, 2003). To control for potentially confounding social processes, we include two indicators of interracial economic disadvantage in our model specifications: the ratio White-to-Black unemployment and the ratio of White-to-Black household income. Preliminary data analyses indicate that these two variables are too highly correlated (r = .74) to be included in the same equation. Therefore, we use factor analysis to combine these variables into a single measure of economic disadvantage.

A crime measurement was included to control for officer workload and deployment. Withrow (2006) argued that there is a significant and positive relationship between the numbers of officers deployed in a beat and the number of crimes in that beat. If there were more officers located in beats with more crime, then the level of exposure that drivers have to police officers (along with the likelihood of being stopped for traffic offenses) would be expected to increase. Thus, included in the models is the number of calls for service, the number of violent index crimes, and the number of property index crimes reported to police during the period in which stop data were collected. These data were provided by the police department. These measures were also factor analyzed and loaded into a single factor using varimax rotation.

Social disorganization theory suggests that social conditions that negatively affect the social cohesion of a community may lead to an increase in the application of social control (Bursik & Grasmick, 1993; Sampson & Groves, 1989). Therefore, we include four structural antecedents of social disorganization in the models: the percentage of single-parent households, the percentage of residents living in poverty, the percentage of residents living in the same dwelling less than 5 years, and the proportion of rental units in the area. These measures were also factor analyzed and loaded into two orthogonal factors. The first comprises the percentage of single-parent households and percentage poor, whereas the second includes the population mobility and rental measures.

Last, for post-stop discretionary decisions (e.g., the rate in which motorists are searched and the rate in which motorists are cited), we introduce the unlogged estimates of the traffic stop rates.

Results

The analyses proceed as follows. First, because theory and research suggest that the racial composition effect may be contingent on the visibility of the police behavior under investigation, we estimate the effects of the structural predictors on the total stop rate, the total search rate, and the total citation rate. Second, to isolate the contextual race-out-of-place effect from the aggregate racial threat effect of the percentage of Blacks, we also estimate race-specific equations for the stop, search, and citation rates, respectively.

Table 2 presents the ordinary least squares estimates regarding the effects of the causal variables on the total and racially disaggregated traffic stop rates. The Model A column reports the unstandardized and standardized coefficients along with the associated standard errors for the overall stop rates, whereas the Model B and Model C columns report those for the Black and White stop rates, respectively.

Table 2.

Ordinary Least Squares Estimates for Traffic Stop Rates

	Model A: Total Stop Rates			Model B: Stop Rates for Blacks			Model C: Stop Rates for Whites
Variable	b	SE	β	b	SE	β	b	SE	β
Percentage Black	0.771	0.436	0.303	−0.269	0.484	−0.099	1.587*	0.621	0.394
Control variables
White-to-Black economic inequalities	−0.145	0.108	−0.175	−0.117	0.120	−0.132	0.230	0.153	0.175
Workload	0.079	0.087	0.095	0.119	0.096	0.135	0.042	0.123	0.032
Social disorganization	0.031	0.118	0.037	−0.107	0.131	−0.121	0.268	0.168	0.204
Population mobility	0.454***	0.089	0.546	0.341***	0.099	0.384	0.421***	0.127	0.320
Constant	4.452***	0.195		5.192***	0.217		4.510***	0.278
Adjusted R²	0.323			0.264			0.453

Note: n = 70.

p < .05. ***p < .001.

The results are rather clear. With the exception of the population mobility, which is positive and significant in all three equations, none of the control variables affect the total or disaggregated traffic stop rates. Of particular interest, however, is the partial association between percentage Black and traffic stops. Contrary to the predictions of the racial threat hypothesis, the percentage of Blacks has no appreciable impact on either total or Black traffic stop rates. Yet consistent with our speculation that the influence of the racial composition of communities might be tempered by expectations of the police (the race-out-of-place thesis), percentage Black positively affects the stop rate for Whites (β = .394, p < .05).

Once the police have initiated a traffic stop, they have the option to either allow the citizen to leave without taking any further action or decide that the situation warrants the exercise of more intrusive forms of crime control. Table 3 presents the results of the multivariate analyses of the macro-level determinants on these activities—namely, the total and race-specific searches.

Table 3.

Ordinary Least Squares Estimates for Search Rates

	Model D: Total Search Rates			Model E: Search Rates for Blacks			Model F: Search Rates for Whites
Variable	b	SE	β	b	SE	β	b	SE	β
Percentage Black	1.194***	0.353	0.390	0.293	0.343	0.123	1.471*	0.562	0.311
Control variables
White-to-Black economic inequalities	−0.033	0.085	−0.033	0.009	0.084	0.012	0.069	0.143	0.045
Workload	0.150*	0.070	0.150	0.203***	0.068	0.260	0.046	0.117
Social disorganization	0.339***	0.095	0.339	0.034	0.095	0.043	0.543**	0.152	0.352
Population mobility	0.061	0.080	0.061	0.006	0.008	0.007	0.102	0.120	0.066
Traffic stop rate	0.001***	0.000	0.562
Black traffic stop rate				0.001***	0.000	0.738
White traffic stop rate							0.000***	0.000	0.369
Constant	1.335***	0.153		2.302***	0.152		1.160***	0.252
Adjusted R²	0.714			0.540			0.676

Note: n = 70.

p < .05. **p < .01. ***p < .001.

Overall, the pattern of findings for the search rate equations suggests that the control variables do a better job of accounting for variations in search rates than in traffic stop rates. As expected, workload positively affects total and Black search rates, whereas the social disorganization factor positively affects total and White rates. Consistent with consensus models of society—especially, the production function variant of this approach (Borcherding & Deacon, 1972; Schwab & Zampelli, 1987)—the traffic stop rates are the best predictors of total search rates (β = .562, p < .001), Black search rates (β = .738, p < .001), and White search rates (β = .369, p < .001).

Nonetheless, independent of the opportunity for searches (traffic stop rates), the percentage of Blacks positively affects the total search rate (β = .390, p < .001). Moreover, in accord with what we found for the traffic stops, it is positively related to the search rate for Whites (β = .310, p < .05) but unrelated to the search rate for Blacks.

Table 4 contains the ordinary least squares regression estimates of the effects of the structural predictors on an alternative escalation of crime control: citation rates. Once again, we find that the control variables more effectively explain and predict variations in the actual performance of crime control than they do the initial decision to make a traffic stop. Comparable to what we report for search rates, workload is positively related to each of the three citation rates, whereas the traffic stop rates evidence the largest partial effects within each citation rate equation. Although population mobility has no noticeable impact on search rates, it is positively related to both the total and White citation rates.

Table 4.

Ordinary Least Squares Estimates for Citation Rates

	Model G: Citation Rates			Model H: Citation Rates for Blacks			Model I: Citation Rates for Whites
Variable	b	SE	β	b	SE	β	b	SE	β
Percentage Black	−0.078	0.355	−0.029	−1.022**	0.354	−0.350	1.259*	0.521	0.310
Control variables
White-to-Black economic inequalities	−0.054	0.086	−0.062	−0.043	0.087	−0.045	0.053	0.132	0.040
Workload	0.227**	0.071	0.261	0.249***	0.071	0.261	0.244*	0.108	0.184
Social disorganization	0.188	0.096	0.216	0.127	0.098	0.133	0.205	0.141	0.154
Population mobility	0.183*	0.081	0.211	0.028	0.081	0.029	0.237*	0.111	0.179
Traffic stop rate	0.001***	0.000	0.635
Black traffic stop rate				0.001***	0.000	0.690
White traffic stop rate							0.000***	0.000	0.502
Constant	4.327***	0.154		4.911***	0.157		4.217***	0.233
Adjusted R²	0.614			0.671			0.624

Note: n = 70.

p < .05. **p < .01. ***p < .001.

The effect of the percentage of Blacks on the citation rates illustrates how the failure to disaggregate the dependent measure can mask countervailing causal processes. Inspection of Table 4 indicates that percentage Black is unrelated to the total citation rate. Note, however, that the percentage of Blacks negatively affects the citation rate for Blacks (β = –.350, p < .01) but positively affects the citation rate for Whites (β = .310, p < .05). Thus, it would appear that the differential impact of the racial composition of beats on the race-specific citation rates counterbalance one another, thereby producing null effects on the total citation rate.

Some previous research has questioned whether traffic offending rates or offending severity differs across races while offering empirical evidence that racial minorities may have higher offending rates (Alpert Group, 2004; Lange, Johnson, & Voas, 2005). Though this cannot be confirmed in the current data, this disparity in behavior implies that the rate in which Blacks are stopped, searched, and cited would be significantly higher than that for Whites. If racial minorities indeed violate traffic laws more frequently or more severely than Whites do, then disparity in stops, searches, and citations would be more a function of behavior rather than race. But in the current analysis, Blacks have a lower stop rate (5.08 versus 5.16 per 1,000) and citation rate (4.79 versus 4.97 per 1,000) than do Whites—though the overall search rate for Blacks is higher than it is for Whites (2.68 versus 1.97 per 1,000). Thus, if minorities were indeed engaged in more frequent or serious rates of traffic offending here, we would expect this difference to manifest in the form of consistently higher rates of social control, not in the form of social control being applied at parity.

In sum, the multivariate analyses of total and race-specific traffic stop, search, and citation rates suggest that the racial composition of communities, independent of workload and opportunity, has a substantial impact on the crime control activities of the police. The theoretical implications of these, as well as our other findings, are discussed below.

Discussion

This study attempts to integrate the racial conflict theory with the police suspicion literature to explain and predict the influence that racial composition of the beat has on the exercise of social control. To reiterate, the racial conflict perspective posits that as the relative size of racial minorities in a beat increases, so too would the application of social control of minorities. However, as we explicate above, the police suspicion literature suggests that the macro-level relationship between minority group population size and law enforcement practices might be tempered by the dynamics of neighborhood composition and the race of individuals encountered in these areas. Thus, individuals with racial characteristics that are inconsistent with the racial makeup of the beat would differentially draw suspicion to themselves, which in turn increases the likelihood that social control would be applied. This race-out-of-place perspective provides insight regarding how officers exercise discretionary choices to stop, search, and issue citations to citizens.

Generally, the results from the ordinary least squares models provide little direct support for the racial threat perspective. After controlling for other theoretically relevant variables, we found no relationship between racial composition of the beat and the rate in which citizens are stopped or cited by the police. However, results did reveal that as percentage Black increased, so too did the search rate of those stopped. After we disaggregated these data by race, we discovered that this was true only for the search rate of White motorists and not for Black motorists.

In contrast, our analyses provide substantial support for the race-out-of-place perspective. A positive relationship is observed between percentage Black and the rate of social control for Whites for all three measures of social control. As the proportion of Black residents in a beat increases, Whites are more likely to be considered out of place; thus, Whites in this environment generate suspicion in officers, and social control is applied. Meanwhile, the rate in which Blacks are stopped or searched is unaffected. Social control is not differentially applied to Blacks in this context because the presence of Blacks in an increasingly “Black beat” is not considered out of place but is rather expected, and this subsequently does not draw the attention of the police.

An examination of the citation rate for Blacks provides even greater support for the race-out-of-place perspective. As percentage Black increases, the citation rate for Blacks decreases. Black motorists in beats with higher proportions of Black residential composition do not draw the suspicion of officers, and social control is less likely to be applied.

To further illustrate this effect, it may prove helpful to conceptualize the measure of the racial composition in an alternative manner. We operationalized the racial makeup of beats as the percentage of Blacks because the racial threat hypothesis is stated in terms of the threat that is believed to emerge from increases in the relative population size of racial minorities. However, it would also be accurate to measure the racial composition of a beat by estimating the percentage non-Black. According to the measurement of variables presented here, percentage non-Black would be estimated as 1 minus percentage Black. This conceptualization of beat racial complexity suggests the following: As percentage non-Black of a beat increases, the citation rate for Blacks increases. This too is consistent with the race-out-of-place thesis. Thus, citation rates for White drivers increases as percentage Black increases, and citation rates for Blacks increase as percentage non-Black increases.

In sum, this research extends the macro-level effect of race variables on the enforcement of traffic laws by drawing attention to the complexities of this relationship. Specifically, the current investigation leads us to conclude that when making decisions to apply social control, officers consciously or subconsciously rely on cues between the environment and individuals observed within the environment. Here, like in other research, it is clear that race matters; however, it is clear that social context matters as well. Officers appear more likely to be applying social control to groups whose racial characteristics are inconsistent with what is expected within the area. The structural characteristics provide officers cues regarding who belongs and who does not. If an officer perceives that someone’s racial characteristics are incongruent to what is expected in the environment, the result is increased suspicion and, ultimately, increased police mobilization. This is true for Whites in predominately non-White beats and, though more limited, for Blacks in predominately non-Black beats.

To be sure, the macro-level nature of this study makes it difficult to examine the influence of behavioral cues on formulating suspicion, which is the primary limitation of this study. Some researchers have noted the importance of considering behavioral and nonbehavioral cues on officer behavior (e.g., Alpert et al., 2005; M. R. Smith, Makarios, & Alpert, 2006), and research by Johnson (2006, 2007) explained that behavioral actions that create suspicion vary across race. A more comprehensive examination on the influence of suspicion and social control should include behavioral and nonbehavioral conditions, and such a study would likely be accomplished by using data from systematic social observations. Related, behavior such as citizen hostility could not be introduced here (such behavior has been considered a central factor in understanding the application of officer social control). These behavioral aspects would be most appropriately examined at the individual level or using multilevel modeling.

A future examination of the influence of race in traffic enforcement needs to take behavioral and situational factors into consideration, but it should also consider environment. The findings presented here recognize that the ecological context contributes to the examination of racial profiling and that disaggregation by race may be required to fully understand the complexities of racial bias during traffic enforcement. Failure to examine contextual factors may mask the way that social control is exercised.

Footnotes

The authors declared no potential conflicts of interests with respect to the authorship and/or publication of this article.

The authors received no financial support for the research and/or authorship of this article.

Notes

References

Alpert

G. P.

MacDonald

J. M.

Dunham

R. G.

(2005). Police suspicion and discretionary decision making during citizen stops. Criminology, 43, 407-434.

Alpert Group. (2004). Miami–Dade racial profiling study. Retrieved January 20, 2008, from http://www.aclufl.org/issues/racial_justice/miamidaderacialprofilingreport.cfm

Batton

Kadleck

(2004). Theoretical and methodological issues in racial profiling research. Police Quarterly, 7, 30-64.

Bierne

(1979). Empiricism and the critique of Marxism on law and crime. Social Problems, 26, 373-385.

Black

(1976). The behavior of law. New York: Academic Press.

Blalock

H. M.

Jr. (1967). Toward a theory of minority-group relations. New York: John Wiley.

Blau

(1982). The cost of inequality: Metropolitan structure and violent crime. American Sociological Review, 47, 114-129.

Borcherding

T. E.

Deacon

R. T.

(1972). The demand for services of non-federal governments. American Economic Review, 62, 891-901.

Brown

M. K.

(1981). Working the street: Police discretion and the dilemmas of reform. New York: Russell Sage Foundation.

10.

Brown

R. A.

Frank

(2005). Police–citizen encounters and field citations: Do encounter characteristics influence ticketing? Policing: An International Journal of Police Strategies and Management, 28, 435-454.

11.

Bursik

R. J.

Grasmick

H. G.

(1993). Neighborhoods and crime: The dimensions of effective community control. Lanham, MD: Lexington Books.

12.

Cox

S. M.

Pease

S. E.

Miller

D. S.

Tyson

C. B.

(2001). Interim report of traffic stops statistics for the state of Connecticut. Rocky Hill, CT: Division of Criminal Justice.

13.

Crank

J. P.

(1998). Understanding police culture. Cincinnati, OH: Anderson.

14.

Eitle

D’Alessio

S. J.

Stolzenberg

(2002). Racial threat and social control: A test of the political, economic, and threat of black crime hypotheses. Social Forces, 81, 557-576.

15.

Engel

R. S.

Calnon

J. M.

(2004). Comparing benchmark methodologies for police–citizen contacts: Traffic stop data collection for the Pennsylvania state police. Police Quarterly, 7, 97-125.

16.

Engel

R. S.

Calnon

J. M.

Bernard

T. J.

(2002). Theory and racial profiling: Shortcomings and future directions in research. Justice Quarterly, 19, 249-274.

17.

Fagan

Davies

(2000). Street cops and broken windows: Terry, race and disorder in New York City. Fordham Urban Law Journal, 28, 457-504.

18.

Fridell

L. A.

(2004). By the numbers: A guide for analyzing race data from vehicle stops. Washington, DC: Police Executive Research Forum.

19.

Gallup. (1999). Racial profiling is seen as widespread, particularly among young black men. Retrieved November 21, 2001, from http://www.gallup.com/poll/releases/pr991209.asp

20.

Greenberg

D. F.

Kessler

R. C.

Loftin

(1985). Social inequality and crime control. Journal of Criminal Law and Criminology, 76, 684-704.

21.

Harris

D. A.

(1999). The stories, the statistics, and the law: Why “driving while Black” matters. Minnesota Law Review, 84, 265-326.

22.

Harris

D. A.

(2002). Profiles in injustice: Why racial profiling cannot work. New York: New Press.

23.

Herbert

(1997). Policing space: Territoriality and the Los Angeles police department. Minneapolis: University of Minnesota Press.

24.

Jackson

P. I.

(1986). Black visibility, city size, and social control. The Sociological Quarterly, 27, 185-203.

25.

Jackson

P. I.

Carroll

(1981). Race and the war on crime: The sociopolitical determinates of municipal police expenditures in 90 non-Southern U.S. cities. American Sociological Review, 46, 290-305.

26.

Jacobs

(1979). Inequality and police strength: Conflict theory and coercive control in metropolitan areas. American Sociological Review, 44, 913-925.

27.

Jacobs

Helms

(1999). Collective outbursts, politics, and punitive resources: Toward a political sociology of spending on social control. Social Forces, 77, 1497-1523.

28.

Johnson

R. R.

(2006). Confounding influences on police detection of suspiciousness. Journal of Criminal Justice, 34, 432-442.

29.

Johnson

R. R.

(2007). Race and police reliance on suspicious non-verbal cues. Policing: An International Journal of Police Strategies and Management, 30, 277-290.

30.

Kansas City Police Department. (2005). Kansas City, Missouri Police Department annual report: 2004. Kansas City, MO: Author.

31.

Klinger

D. A.

(1997). Negotiating order in patrol work: An ecological theory of police response to deviance. Criminology, 35, 277-306.

32.

Knowles

Persico

Todd

(2001). Racial bias in motor vehicle searches: Theory and evidence. Journal of Political Economy, 109, 203-299.

33.

Langan

P. A.

Greenfeld

L. A.

Smith

S. K.

Durose

M. R.

Levin

D. L.

(2001). Contacts between police and the public: Findings from the 1999 national survey. Washington, DC: Bureau of Justice Statistics.

34.

Lange

J. E.

Johnson

M. B.

Voas

R. B.

(2005). Testing the racial profiling hypothesis for seemingly disparate traffic stops on the new jersey turnpike. Justice Quarterly, 22(2), 193-223.

35.

Liska

A. E.

(1987). A critical examination of macro perspectives on crime control. Annual Review of Sociology, 13, 67-88.

36.

Liska

A. E.

(1992). Social threat and social control. Albany: State University of New York Press.

37.

Liska

A. E.

(1997). Modeling the relationship between macro forms of social control. Annual Review of Sociology, 23, 67-61.

38.

Liska

A. E.

Chamlin

M. B.

(1984). Social structures and crime control among macrosocial units. American Journal of Sociology, 90, 388-395.

39.

Liska

A. E.

Lawrence

J. J.

Sanchirico

(1982). Fear of crime as a social fact. Social Forces, 60, 760-761.

40.

Meehan

Ponder

(2002). Race and place: The ecology of racial profiling African American motorists. Justice Quarterly, 19, 399-430.

41.

Novak

K. J.

(2004). Disparity and racial profiling in traffic enforcement. Police Quarterly, 7, 65-96.

42.

Paoline

E. A.

Myers

S. M.

Worden

R. E.

(2000). Police culture, individualism, and community policing: Evidence from two police departments. Justice Quarterly, 17, 575-606.

43.

Paoline

E. A.

Terrill

(2005). The impact of police culture on traffic stop searches: An analysis of attitudes and behavior. Policing: An International Journal of Police Strategies and Management, 28, 455-472.

44.

Parker

K. F.

Stults

B. J.

Rice

S. K.

(2005). Racial threat, concentrated disadvantage and social control: Considering the macro-level sources of variation in arrests. Criminology, 43, 1111-1134.

45.

Petrocelli

Piquero

A. R.

Smith

M. R.

(2003). Conflict theory and racial profiling: An empirical analysis of police traffic stop data. Journal of Criminal Justice, 31, 1-11.

46.

Quinney

(1970). The social reality of crime. Boston: Little, Brown.

47.

Quinney

(1977). Class, state and crime: On theory and practice of criminal justice. New York: McKay.

48.

Ramirez

McDevitt

Farrell

(2000). A resource guide on racial profiling data collection systems: Promising practices and lessons learned. Washington, DC: Bureau of Justice Assistance.

49.

Reitzel

Piquero

A. R.

(2006). Does it exist? Studying citizens’ attitudes of racial profiling. Police Quarterly, 9, 161-183.

50.

Reitzel

J. D.

Rice

S. K.

Piquero

A. R.

(2004). Lines and shadows: Perceptions of racial profiling and the Hispanic experience. Journal of Criminal Justice, 32, 607-616.

51.

Rojek

Rosenfeld

Decker

(2004). The influence of driver’s race on traffic stops in Missouri. Police Quarterly, 7, 126-147.

52.

Rubenstein

(1973). City police. New York: Farrar, Strauss & Girox.

53.

Sampson

R. J.

Groves

W. B.

(1989). Community structure and crime: Testing social-disorganization theory. American Journal of Sociology, 94, 774-802.

54.

Schafer

J. A.

Carter

D. L.

Katz-Bannister

Wells

W. M.

(2006). Decision-making in traffic stop encounters: A multivariate analysis of police behavior. Police Quarterly, 9, 184-209.

55.

Schwab

R. M.

Zampelli

E. M.

(1987). Disentangling the demand function from the production function for local public services. Journal of Public Economics, 33, 245-260.

56.

Skogan

W. G.

(1995). Crime and the racial fears of White Americans. Annuals of the American Academy of Political and Social Science, 539, 59-72.

57.

Skolnick

J. H.

(1994). Justice without trial: Law enforcement in democratic society (3rd ed.). New York: Macmillan.

58.

Smith

B. W.

Holmes

M. D.

(2003). Community accountability, minority threat, and police brutality: An examination of civil rights criminal complaints. Criminology, 41, 1035-1064.

59.

Smith

D. A.

(1986). The neighborhood context of police behavior. In Reiss

A. J.

Jr. Tonry

(Eds.), Communities and crime (pp. 313-342). Chicago: University of Chicago Press.

60.

Smith

M. R.

Makarios

Alpert

G. P.

(2006). Differential suspicion: Theory specification and gender effects in the traffic stop context. Justice Quarterly, 23, 271-295.

61.

Smith

M. R.

Petrocelli

(2001). Racial profiling? A mulitvariate analysis of police traffic stop data. Police Quarterly, 4, 4-27.

62.

Swigert

V. L.

Farrell

R. A.

(1977). Normal homicides and the law. American Sociological Review, 42, 16-32.

63.

Turk

(1969). Criminality and the legal order. Chicago: Rand McNally.

64.

United States v. Arvizu, 534 U.S. 266 (2002).

65.

Walker

(2001). Searching for the denominator: Problems with police traffic stop data and early warning system solution. Justice Research and Policy, 3, 63-95.

66.

Withrow

B. L.

(2004a). A comparative analysis of commonly used benchmarks in racial profiling: A research note. Justice Research and Policy, 6, 1-22.

67.

Withrow

B. L.

(2004b). Driving while different: A potential theoretical explanation for race-based policing. Criminal Justice Policy Review, 15, 344-364.

68.

Withrow

B. L.

(2006). Racial profiling: From rhetoric to reason. Upper Saddle River, NJ: Pearson Prentice Hall.

69.

Worden

R. E.

(1995). Police officers’ belief systems: A framework for analysis. American Journal of Police, 14, 49-81.

70.

Zingraff

M. T.

Mason

H. M.

Smith

W. R.

Tomaskovic-Devey

Warren

McMurray

H. L.

. (2000). Evaluating North Carolina state highway patrol data: Citations, warnings, and searches in 1998. Retrieved November 21, 2001, from http://www.nccrimecontrol.org/shp/ncshpreport.htm