Are there conditions under which minority bureaucrats are less likely to provide active representation? Th e authors address this question by testing the link between passive and active representation for race in a police department and in the particular instance of racial profi ling. Literature from three areas — racial profi ling, representative bureaucracy, and police socialization — is brought together. Th e fi ndings support the hypothesis that organizational socialization can hinder the link between passive and active representation. Furthermore, the authors fi nd that the presence of black police offi cers is related to an increase in racial profi ling in the division. Th is fi nding was unexpected and raises several important questions regarding active representation and race.
Previous research has established a clear link between passive and active representation for race. Numerous studies have concluded that minority bureaucrats implement policies or use their discretion to reduce the disparate treatment minority clients have historically received from various public bureaucracies ( Hindera 1993; Meier and Stewart 1992; Meier, Stewart, and England 1989; Selden 1997 ). But are there conditions under which minority bureaucrats are less likely to provide active representation?
Recent research examining the link between passive and active representation for gender has highlighted the role that institu- tional variables play in the trans- lation of passive into active representation ( Keiser et al. 2002 ). Institutional variables can promote or hinder the provision of active representation by bu- reaucrats. One such institutional variable is organiza- tional socialization. Some organizations rely heavily on socialization to structure and shape the actions of individual bureaucrats ( Kaufman 1960; Romzek 1990; Simon 1947 ). It is our contention that organi- zational socialization may hinder the link between
passive and active representation. 1 To date, little research has examined the role that socialization plays in the provision of active representation.
To address this gap in the literature, this research will test the link between passive and active representation for race in organizations that rely heavily on socialization — in this case, police departments. Police departments are notorious for their use of socialization to modify the behaviors and attitudes of their employees. Th erefore, this setting off ers a diffi cult test for the theory of representative bureaucracy. Police depart- ments meet the necessary criteria to facilitate active representation. Th e policy area of racial profi ling is highly salient for minority communities, and police offi cers exercise discretion in their work. At fi rst glance, we would expect to fi nd that the proportion of African American offi cers working in a police depart- ment is correlated with a decrease in the incidence of racial profi ling (measured by racial disparity in vehicle stops) by that police department, all else being equal. But does organizational socialization strip the racial identity of police offi cers, somehow changing them from “black” to “blue” and limiting their provision of active representation?
Racial Profi ling Tension between the police and racial/ethnic minority communi- ties has been and continues to be one of the most pressing issues facing American police organiza- tions ( Barak, Flavin, and Leighton 2001; Culver 2004; Websdale 2001; Williams 1998 ). Historically, American policing has a heritage of legally sanc-
tioned, disparate service delivery and the enforcement of racially motivated laws and statutes, inclusive of slave codes, black codes, and other oppressive policies and practices directed toward Africans, African Americans, and other marginalized populations ( Russell-Brown 1998 ). Consequently, the present-day
Vicky M. Wilkins Brian N. Williams 1 University of Georgia
Black or Blue: Racial Profi ling and Representative Bureaucracy
Vicky M. Wilkins is an assistant
professor in the Department of Public
Administration and Policy at the University
of Georgia. Her teaching and research
interests include public administration,
public personnel management, and
representative bureaucracy. Her research
has appeared in the American Political Science Review, Public Administration Review, Journal of Public Administration Research and Theory, Review of Public Personnel Administration , and Legislative Studies Quarterly. E-mail : firstname.lastname@example.org
Brian N. Williams is an assistant
professor in the Department of Public
Administration and Policy at the University
of Georgia. His primary areas of research
explore the relationships between
bureaucratic units and communities and
community policing efforts within
communities of color. He is the author of
Citizen Perspectives on Community Policing: A Case Study in Athens, Georgia (State University of New York Press, 1998), among
other research articles, book chapters, and
E-mail : email@example.com
Diverse Ethical Challenges Facing Today’s Public Administrators
Recent research examining the link between passive and active representation for gender has
highlighted the role that institutional variables play in the translation of passive into
Racial Profi ling and Representative Bureaucracy 655
relationship between minorities and the police in America is one that has been infl uenced by the histori- cal legacies of slavery, segregation, and discrimination ( Russell-Brown 1998; Walker 1977, 1980 ; Websdale 2001; Williams and Murphy 1990 ).
Th e public perception of racial profi ling is a contem- porary by-product of the legacy of American policing and one of the most hotly debated issues in law en- forcement today. 2 Consequently, racial profi ling has been defi ned in many diff erent ways; however, for the purposes of this article, we will utilize the defi nition of Ramirez, McDevitt, and Farrell, who defi ne racial profi ling as “any police-initiated action that relies on the race, ethnicity, or national origin rather than the behavior of an individual or information that leads the police to a particular individual who has been identifi ed as being, or having been, engaged in crimi- nal activity” (2000, 3). With its association of race/ ethnicity as a proxy for increased likelihood of crimi- nal behavior, racial profi ling refl ects Skolnick’s (1975) conception of America’s “symbolic assailant” and is symptomatic of a much larger issue of bias-based policing.
Contemporary racial profi ling has been cited as a police tactic whose practical origins are associated with the Ronald Reagan administration’s “war on drugs” ( ACLU 1999; Harris 1999 ). In 1984, the Drug Enforcement Administration (DEA) imple- mented Operation Pipeline to stem the fl ow of drugs into American cities. Th is nationwide highway interdiction program targeted private motor vehicles in a cooperative eff ort of the DEA and state and local law enforce- ment agencies. To assist in the interdiction eff orts, a profi le of a potential drug traffi cker was developed and utilized as a train- ing tool for local, state, and federal law enforcement offi cers ( ACLU 1999 ). Moreover, these eff orts resulted in the issuance of guidelines in 1985 by the Florida Department of Highway Safety and Motor Vehicles regarding the common characteristics of drug couriers. Th ese guidelines cautioned state troopers to be suspicious of specifi c drug courier characteristics, including certain racial and ethnic groups “associated” with the drug trade. Conse- quently, traffi c stops were initiated by offi cers using this race/ethnicity-based description.
During the late 1980s and early 1990s, Lamberth (1996) studied the racial distribution of traffi c stops in New Jersey. Lamberth designed a research method- ology to ascertain the rate at which blacks were being stopped compared to the percentage of blacks driving
the same stretch of road. Lamberth’s analysis found that black and white drivers violated traffi c laws at the same rate, yet 73.2 percent of the cars stopped and arrested had black drivers or passengers, though these cars represented only 13.5 percent of the vehicles on the road. Since Lamberth published this work, a growing body of statistical evidence has suggested that blacks and Latinos are substantially more likely than whites to be stopped, questioned, or searched ( Close and Mason, 2003; Cordner, Williams, and Zuniga 2000; Rojek, Rosenfeld, and Decker 2004; Smith and Petrocelli 2001 ). Of particular note is Meehan and Ponder’s (2002) analysis, which found that African American motorists driving in neighborhoods with higher percentages of white residents were subjected to disproportionate surveillance and stopping by police compared to white motorists. Moreover, these African American motorists were less likely to have contraband than white motorists stopped in the same neighborhoods.
Methodological shortcomings (i.e., the diffi culty of empirically identifying the driving public, inadequate data collection because of offi cer disengagement, and questions related to the accuracy and reliability of the data) have made it diffi cult to state conclusively that departments are engaging in this discriminatory prac- tice ( Engle and Calnon 2004 ; Engle, Calnon, and Bernard 2002 ; Meehan and Ponder 2002; Ramirez, McDevitt, and Farrell 2000; Smith and Alpert 2002 ).
Nonetheless, the allegation of racial profi ling, and the percep- tion of the police that it creates among minorities, lessens the likelihood of meaningful collabo- ration with minority citizens and communities in the coproduc- tion of public safety and order.
Theory of Representative Bureaucracy Th e theory of representative bureaucracy concerns how the demographic characteristics of
bureaucrats aff ect the distribution of outputs to clients who share these demographic characteristics. Th e literature distinguishes between two forms of repre- sentation: passive and active. Passive representation is concerned with whether the bureaucracy has the same demographic origins — sex, race, income, class, religion — as the population it serves ( Mosher 1982 ). Studies of passive representation examine whether the composition of the bureaucracy mirrors the demo- graphic composition of the general population or whether minorities are underrepresented in the bu- reaucracy ( Cayer and Sigelman 1980; Cornwell and Kellough 1994; Hall and Saltzstein 1975; Kellough 1990; Kellough and Elliot 1992; Meier and Stewart 1992; Naff 2001 ). Th ese studies seldom examine the
… the allegation of racial profi ling, and the perception of the police that it creates among
minorities, lessens the likelihood of meaningful
collaboration with minority citizens and communities in the
coproduction of public safety and order.
656 Public Administration Review • July | August 2008
eff ects of representation or lack of representation on an agency’s policy outputs. Active representation, in contrast, is concerned with how representation infl u- ences policy making and implementation. Active representation assumes that bureaucrats will act pur- posely on behalf of their counterparts in the general population ( Pitkin 1967 ).
Early scholars assumed that passive representation would naturally translate into active representation, but recent work has identifi ed a couple of necessary conditions for the link to occur ( Keiser et al. 2002 ; Meier 1993 ). First, bureaucrats must have discretion in order to act on a given policy. In bureaucracies in which most decisions are dictated by rules, bureau- crats have few opportunities to shape outputs to re- ward a particular group within their clientele ( Meier 1993 ). Police offi cers are the quintessential street-level bureaucrats and clearly exercise a necessary amount of discretion ( Lipsky 1980 ). Th ere are numerous ways that minority police offi cers could infl uence outcomes for minority drivers, either through their own actions or by infl uencing the organization. Th e second neces- sary condition is that the policy area must be salient to the demographic group in question ( Keiser et al. 2002 ; Meier 1993 ; Selden 1997 ). As discussed earlier, racial profi ling is a highly salient issue in black communities.
Race and ethnicity are the most common demo- graphic characteristics examined by the existing research on both passive and active representation (Cayer and Sigelman 1980; Hindera 1993; Meier 1975 , 1993; Rehfuss 1986; Riccucci and Saidel 1997; Saltzstein 1989; Selden 1997; Th ompson 1976, 1978 ). In the case of race and ethnicity, numerous studies have concluded that minority bureau- crats implement policies or use their discretion to reduce the disparate treatment minority clients have historically received from public bureaucracies ( Hindera 1993; Meier and Stewart 1992; Meier, Stewart, and England 1989; Selden 1997 ). Th erefore, at fi rst glance, we would expect that passive representation leads to active representation in police departments. However, the high levels of organizational socialization present in police departments require that we modify this expectation.
Organizational Socialization Recent research on representative bureaucracy focuses on the role that institutional and contextual factors play in the link between passive and active representa- tion ( Keiser et al. 2002 ). One of the institutional variables considered is organizational socialization.
Scholars argue that administrators are socialized by the organizations they work in and adopt behaviors and preferences that are consistent with organizational goals, thereby minimizing the infl uence of their own personal values on bureaucratic behavior ( Downs 1967; Gawthrop 1969; Meier and Nigro 1976 ; Simon 1957 ; Th ompson 1976; Weber 1946 ). To ensure that administrative decisions are consistent with the goals and values of the organization, the organization attempts to instill in employees a common set of assumptions and way of looking at the world. Th is worldview values organizational loyalty above personal beliefs ( Downs 1967; Romzek 1990 ; Simon 1957 ). Simon claims that organizational identifi cation, “the process whereby the individual substitutes organiza- tional objectives … for his own aims” (1957, 218), depersonalizes administrative decision making, ensur- ing that all employees make “correct” decisions consis- tent with organizational objectives. Employees may be willing to adopt the organization’s values to increase the chance of promotion and career success, either because they feel peer pressure to do so or simply because they come to agree with and internalize the dominant organizational view ( Romzek 1990 ; Simon 1957 ; Th ompson 1976 ). Th erefore, organizational socialization may actually strip away the racial identity of black police offi cers and replace it with an organiza- tional identity. In essence, this process may transform those offi cers who are “black in blue” to simply blue.
Police Socialization Th e degree of cohesion and solidarity among police offi cers has long been noted as one of the most noticeable yet unusual aspects of the police profession ( Hahn 1971 ) and has been described metaphorically as a “blue-walled mosaic.” As such, this profession refl ects and proj- ects a sense of fraternal support and fi delity that, in turn, encour- ages and reinforces an overarch-
ing police culture. However, this police culture is not necessarily monolithic in nature. It is fractured and helps cultivate and reinforce certain subcultural norms that refl ect the functional areas of police work. Conse- quently, a management culture versus a street cop culture is visible at the subcultural level, as are other subcultures and related norms that refl ect departmen- tal, regional, international, status, and operational diff erences ( Reuss-Ianni 1983 ). Nonetheless, offi cers beyond their jurisdictional, operational, or hierarchi- cal boundaries are cognizant of the brotherhood they are members of, thereby highlighting the police process of occupational socialization.
Th e occupational or organizational socialization of police offi cers attempts to separate those environmental and
Th e degree of cohesion and solidarity among police offi cers has long been noted as one of
the most noticeable yet unusual aspects of the police profession
and has been described metaphorically as a “blue-walled
Racial Profi ling and Representative Bureaucracy 657
structural infl uences that may aff ect the individual offi cer. As an important organizational function, it seeks to engender the requisite skills needed to be eff ective arbiters of community confl ict, providers of public safety, and maintainers of public order. Socialization theory argues that police behavior is determined more by work experience and peers than by offi cer predisposition. Th is premise is one of two major theories that seek to explain the origin of police offi cers’ attitudes and values, and it presupposes that police behavior, either good or bad, is a learned behavior that occurs subsequent to the hiring of the individual offi cer.
Caplow (1964), Manning (1970) , and others describe organizational socialization as the process by which members of an organization learn the required behav- iors and supportive attitudes needed to be recognized as a member of that organization. Its primary objec- tive is long-term internalization and action congru- ence with organizational values. Th is process occurs throughout all stages of an offi cer’s career and incor- porates both formal and informal dimensions.
Th e formal socialization process is hierarchical in nature and begins with a police cadet’s initial contact with his or her instructor in a sterile, preoccupational environment — the classroom of the local police acad- emy. Th is initial contact is followed by the subsequent postacademy “street” experiences of offi cers who are in the fi eld under the supervision of an assigned fi eld training offi cer. Th e fi nal phase of the formal socializa- tion process continues as offi cers come into contact and interact with more senior supervisors and manag- ers. Conversely, the informal socialization process transcends the police offi cer seniority continuum and aff ects both novice police cadets and more seasoned police offi cers. Th e informal method counters the hierarchical nature of the formal socialization process and is characterized by peer-to-peer or offi cer-to- offi cer interactions. Th ese interactions have been found to contradict the values articulated in the more formal socialization process.
Insights from Interviews with Individual Offi cers Past qualitative research off ers some insights into the potential impact of organizational socialization in general and police socialization in particular on Afri- can American police offi cers. It is possible that the behavior of black police offi cers is infl uenced by the strong need to fi t into the culture of the organization. A statement made by one offi cer highlights this pos- sibility: “Th is one [white] sergeant here that loves to refer to black people as ‘those people.’ If you’re black and you refer to black people as ‘those people’ you’re treated okay by him. Otherwise, you may not be. You get a few black cops who do that sort of thing, guys who have no backbone. Th ey’re treated by this one
sergeant as one of the guys” ( Leinen 1984, 38 ). Further evidence is off ered by the following statements, which were generated from an analysis of in-depth individual interviews and focus group discussions among African American police offi cers in regard to their perceptions of the infl uence of organizational socialization on their performance and attitudes about racial profi ling. 3
It [profi ling] is not only white offi cers. It’s black offi cers, too. Because you hear them say, “Well, he must be doing something to drive that vehicle.” Th e assumption is we can’t have African American males with affl uent parents, whose car they are driving, or they can’t be entrepreneurs or hard working people. It’s like, if you are driving this, then you must be involved in something illegal. Th e unfortunate thing is those thoughts don’t only come from people other than African Americans. Black people think that about black people too. Th at’s what really concerns me. I think we bought into the thought process of other people about our own people. (Personal communication, African American female police offi cer, 2003)
Sometimes you have to do some degree of it [profi ling] because that’s one of the instincts that we have and that we pick up over the years if you stay in it [police profession] long enough. Profi ling will come to you to a certain degree like anybody else (Personal communication, African American male police offi cer, 2003)
Before I left the streets working, there was a housing project and there was a big deal about the “stats.” I was watching all the other boys making them stats, but they were making illegal stats. No probable cause, but there was mighty trumped up charges just to say, “Hey Sarge, I’ve got this paperwork over here.” And I remember one day at roll call. I’d just had enough about the stats and I confronted the Sarge in front of everybody. And I said, “Sarge, you know the reason these guys got these stats is because they’re going around the corner getting them in the wrong kind of way. And I’m not going out of the way to get no stats.” He took me to his offi ce and said, “I’m trying to keep my guys motivated.” I said, “You’re doggone right you’re trying to keep them motivated — motivated to do the wrong thing. Gestapo.” (Personal com- munication, African American male police offi cer, 2003)
Th ese statements suggest that police socialization may have a desensitizing eff ect on black offi cers and may ultimately infl uence their attitudes and behavior during traffi c stops. Consequently, it is valuable to
658 Public Administration Review • July | August 2008
understand police socialization in the context of the police department we are examining.
Study Site Socialization: Some Contextual Considerations In this study, we analyze data from a traffi c stop study in San Diego, California, a large urban municipality. Th is local law enforcement agency employs in excess of 2,000 sworn police offi cers, and its procedures and practices refl ect those of most big-city police depart- ments. Th e department embraces the community- and problem-oriented philosophies that dominate the landscape of American police agencies and utilizes a decentralized approach to policing, with community collaboration in problem identifi cation and problem solving playing a vital role in the coproduction of public safety and public order.
Th e police offi cer selection process in this city is similar to that of other large, urban police depart- ments in the United States. Th is process begins with an initial screening of potential applicants. Each applicant is required to meet or exceed the following requirements: be at least 20.5 years of age on the day of the written exam and 21 years of age at the time of academy graduation; be a U.S. citizen or permanent resident alien who is eligible and has applied for U.S. citizenship prior to application for employment; possess a valid Class C driver’s license; and possess certifi cation of graduation from a high school in the United States or its equivalent (i.e., GED, successful completion of state profi ciency examination). Th is initial screening phase is followed by eight other steps in the offi cer selection process: written examination, background investigation, physical ability test, poly- graph exam, departmental interview, psychological evaluation, medical examination, and academy training.
Academy training consists of 32 weeks of college-level training. During this time period, recruits take courses that cover the principles of law enforcement, criminal law, rules and evidence, search and seizure, laws of arrest and control methods, traffi c laws, juvenile laws, fi rst aid, care and use of fi rearms, patrol theory and methods, the criminal justice system, and physical conditioning and self-defense. Th is preservice training concludes with six weeks of fi eld training. Upon the successful completion of this curriculum, recruits graduate with the rank of Police Offi cer I and begin a 12-week fi eld training program under the tutelage of an experienced fi eld training offi cer. Based on this high level of occupational socialization within the San Diego Police Department, we expect a conditioning eff ect whereby the confl uence of the formal and infor- mal dimensions of police socialization shapes the values and behaviors of police offi cers, regardless of race, and limit or negate the provision of active representation.
Study Design and Methodology Th e data for this project come from the vehicle stop forms collected by the San Diego Police Department in 2000. Starting on January 1, 2000, police offi cers in San Diego were required to complete a vehicle stop form each time they stopped a vehicle. Use of the ve- hicle stop forms was discontinued in 2001. Vehicle stop forms recorded the context of traffi c stop encounters (date and time), the reason for the stop, driver demo- graphics (gender, race/ethnicity, age), the actions taken during the stop (search, search authority, search out- come), and the disposition of the stop (citation, arrest). Offi cers turned these forms in at the end of their shifts. Th e data on the forms was entered into a database by personnel at police headquarters. For this project, we analyze the 168,901 stop forms compiled during 2000. 4
Th e only piece of individual-level information the police offi cer was required to provide was the division he or she was assigned to. Given this, our level of analysis for this project is aggregated to the division. Th e San Diego Police Department consists of eight divisions. Th ese divisions (North, Northeast, East, Southeast, Central, West, South, and Mid-City) are geographically arranged throughout the area and serve as the fi rst line of command for the police offi cers.
Dependent Variable To test the relationship between racial profi ling and representative bureaucracy, we use the most common measure of racial disparity, vehicle stops. Scholars and politicians often focus on the diff erence between the percentage of stops when the driver is black and some measure of the black population in the area ( Harris 1999; Lundman and Kaufman 2003; Novak 2004; Smith and Petrocelli 2001 ). 5 Often, the measures of the black population are problematic. Some studies use the city’s black population or even the state’s black population. We try to improve on this measure by using the black driving-eligible population, measured by the population that is 15 years or older in the division. 6 Our dependent variable is the diff erence between the percentage of stops in the division involv- ing a black driver and the percentage of the black driving-age population in the division for each month in 2000. Th is yields a sample size of 96 for our analy- sis. For example, in May 2000, 7 percent of the ve- hicle stops in the Northern division involved black drivers, while only 1 percent of the driving-eligible population in that division was black, producing a racial disparity measure of 6 percent for that month. Table 1 presents the vehicle stops by the race/ethnicity of the driver in each division compared to the driving- age population for 2000.
Independent Variable To test the link between passive and active representa- tion, our variable of interest is the percentage of black sworn police offi cers in each division in 2000. 7 Th e
Racial Profi ling and Representative Bureaucracy 659
mean of this variable is 9.6 percent. Th e division with the greatest representation of black police offi cers is the Southeast division, in which 22.7 percent of the offi cers are black. 8 Th e theory of representative bu- reaucracy and past research on active representation for race suggest that as the percentage of minority police offi cers in a division increases, the racial dispar- ity in that division will decrease. Our understanding of organizational socialization leads us to modify our expectations. Instead, we expect to fi nd that as a result of intense organizational socialization, black police offi cers no longer think of themselves as “black,” thereby negating the possibility of active representa- tion. In addition, the overwhelming pressure to fi t into the culture of the organization may produce a “blue” mentality on the part of black offi cers and may result in an increase in racial disparity in the division.
Control Variables Several community factors may infl uence vehicle stops, and we can control for some of them in our model. First, we include the median income of the division according to the 2000 U.S. Census. 9 Previous research has found that racial disparity in enforcement increases in high-poverty areas ( Cox et al. 2001; Fagan and Davies 2000; Smith and Petrocelli 2001 ; Spitzer 1999). It is possible that all police offi cers, regardless of race or ethnicity, target certain types of drivers. Find- ings from in-depth individual interviews and focus group discussions with police offi cers, as well as with police executives, suggest that the notion of the “low- hanging fruit” infl uences police practice based on the race and socioeconomic status of the community.
I think one of the things we have to be careful of is the temptation by a lot of offi cers to grab the low-hanging fruit … what I mean by that is focusing our attention and enforcement eff orts on certain populations and communities that may not be as vocal or have access to as much power, as lets say a more affl uent white commu- nity. It is not uncharacteristic for members of poorer minority communities to express some concern that they have been targeted and receive a diff erent type of police response. (Personal communication, retired African American male police executive, 2002)
Given this, we expect to fi nd higher racial disparity in divisions with lower median incomes. An additional way to get at the eff ects of poverty on police behavior is to control for the monthly city unemployment rate. We hypothesize that unemployment is positively related to racial disparity in traffi c stops.
Another community factor that is often associated with offi cer behavior is the racial composition of the population in the division. Studies fi nd that black motorists are more likely to be stopped or searched in neighborhoods with higher populations of white residents ( Meehan and Ponder 2002; Smith and Petrocelli 2001 ). To control for the com- position of the division, we include a measure for the percentage of the division population that self- identifi ed as white/Caucasian in the 2000 census. We expect to fi nd that racial disparity increases in divisions where the population is predominantly white/Caucasian.
Th e fi nal community variable that we control for is the monthly overall crime rate of the city. Th e mea- sure of crime rates includes seven crimes: homicide, forcible rape, robbery, aggravated assault, burglary, motor vehicle theft, and larceny. Research suggests that area crime rates are positively related to the inci- dences of racial profi ling ( Cox et al. 2001; Fagan and Davies 2000; Smith and Petrocelli 2001 ; Spitzer 1999). We hypothesize that increased crime rates increase enforcement, which, in turn, may increase disparity. Table 2 presents the summary statistics for all of our variables.
Methodology To test the relationship between racial profi ling and the presence of minority offi cers in the eight divisions for the months of 2000, we employ ordinary least squares regression. Because pooled cross-sectional time-series data set often exhibit heteroscedasticity (i.e., correlated error terms across units at each time point), we use panel-corrected standard errors ( Beck and Katz 1995 ). A second problem for our data set is correlation among error terms within the same unit (division) over time (i.e., autocorrelation). To control for bias in the standard errors, we specify an AR1 process. 10
Table 1 Vehicle Stops by Race/Ethnicity of Driver and by Division, 2000
Black/African American (percent) White (percent)
Division Stops Population Difference Stops Population Difference
Northern 4 1 3 76 80 – 4 Northeastern 5 3 2 61 67 – 6 Eastern 8 4 4 64 78 – 14 Southeastern 34 27 7 12 15 – 3 Central 15 13 2 40 31 9 Western 10 6 4 64 69 – 5 Southern 4 5 – 1 15 20 – 5 Mid-City 22 15 7 29 47 – 18
660 Public Administration Review • July | August 2008
Findings Th e results of the model predicting the relationship between the presence of black offi cers and racial pro- fi ling are presented in table 3 . 11 Th e independent variables in the model predict 54 percent of the vari- ance in the dependent variable. 12 However, it is the parameter estimates that are most interesting. 13
We expected to fi nd that because of high levels of socialization, the percentage of black police offi – cers in the division would not reduce the racial disparity in vehicle stops for that division. Th is hypothesis is supported. Th e coeffi cient is both positive and statistically signifi cant ( p < .01). In other words, as the presence of black police offi cers increases, so does the racial disparity in vehicle stops in that division. Substantively speaking the impact is large: A 1 percent increase in black police offi cers in a division leads to a 1.98 increase in the racial disparity measure for a division in a month.
Two of the community control variables are also sta- tistically signifi cant ( p < .01) and support our hypoth- eses. First, as expected, the median income of the division is negatively related to racial disparity. Racial disparity in vehicle stops increases in poorer divisions, although the eff ect size is small. A $1,000 increase in median income is associated with a .25 decrease in
racial disparity. Second, the composition of the popu- lation of the division is positively and signifi cantly related to racial disparity. Our hypothesis that racial profi ling would increase in divisions where the popu- lation is predominantly white/Caucasian is supported. A 1 percent increase in the percentage of the division population that is white/Caucasian increases the racial disparity measure by .26. Neither the unemployment rate nor the crime rate appears to be related to racial profi ling.
Discussion In this article, we tested the hypothesis that the link between passive and active representation can be hindered by organizational socialization. Previous studies of active representation and race in organiza- tions with lower levels of socialization (i.e., public schools, the Equal Employment Opportunity Com- mission, and the Farmers Home Administration) have consistently found that black bureaucrats use their discretion to reduce negative outcomes or increase positive outcomes for black clientele. We posed the
question, are there conditions under which minority bureau- crats do not provide active repre- sentation? We addressed this question in an agency with high levels of formal and informal socialization, and the answer appears to be yes. Th e structure and processes of an organization
can aff ect the representation provided by the bureau- crats working there. Th e presence of black police offi cers is related to an increase in racial disparity in the division. Th is fi nding raises several important questions, most of which require individual-level data to address.
Th e results of our analysis, coupled with the state- ments that emerged from the analysis of in-depth individual interviews and focus group discussions with African American police executives and offi cers, seem to suggest that the pressure to conform to the organization or to achieve the goals of the organiza- tion weighs heavily on black offi cers and aff ects their
Table 2 Description of Variables
Variable Mean Standard Deviation
Minimum and Maximum
Percentage black police offi cers in division
9.6 5.5 3.7 – 22.7
Median income for division (thousands of dollars)
42.9 12.3 24.4 – 63.5
Unemployment rate (percent)
3.9 0.37 3.2 – 4.6
Crime rate 1917 127 1715 – 2115 Percentage of division population that is white/ Caucasian
49 25.85 14 – 79
Table 3 Regression Estimates for Racial Disparity in Vehicle Stops in the San Diego Police Divisions, 2000
Variable Unstandardized Coeffi cient (panel-corrected standard errors)
Standardized Coeffi cient (panel-corrected standard errors)
Percentage of black police offi cers 1.98 (.363) *** 1.32 (.242) *** Median income – .25 (.04) *** – .37 (.06) *** Unemployment rate .41 (1.03) .02 (.05) White/Caucasian population .26 (.04) *** .81 (.12) *** Crime rate .0001 (.001) .002 (.024) Constant –15.55 (6.51) Rho .67 .67 N 96 96 R-squared .54 .54 F 77.54 *** 77.54 ***
*** Denotes signifi cance at better than .001 (two-tailed test).
We posed the question, are there conditions under which minority bureaucrats do not
provide active representation?
Racial Profi ling and Representative Bureaucracy 661
attitudes and ultimately their behavior. However, more systematic individual-level qualitative research is necessary to explore the causal relationship behind these fi ndings.
Implications for Future Research Th e fi ndings of this study suggest several avenues for future research. Qualitative individual-level data would allow us to further explore a number of inter- esting questions raised by our fi ndings. First, how does the notion of “low-hanging fruit” infl uence the behavior of offi cers of any race or ethnicity? Second, do police offi cers, especially black police offi cers, feel a strong need to fi t into the culture of the organization, and what is the impact of this socialization on their behavior? In other words, do black offi cers feel pres- sure to demonstrate that they are just as “professional” as their colleagues, leading them to refrain from pro- viding active representation? Finally, how strong of a social construction is profi ling, be it racial, socioeco- nomic, or neighborhood, and what is the eff ect on police offi cers?
Another research avenue would involve examining individual-level traffi c stop data, which would allow us to explore the causal mechanism behind the aggre- gate fi ndings. With aggregate-level data, it is impos- sible to know whether black police offi cers, because of high levels of organizational socialization, actually profi le black drivers or whether the presence of large numbers of black offi cers in a division changes the behavior of other offi cers in the organization, resulting in an increase in racial profi ling. We could also exam- ine how rank or tenure in the police department infl uences traffi c stop behavior. By examining the number of years that a police offi cer has served on the force, we would have a better test of the eff ect of organizational socialization on offi cer behavior. Also, what about the eff ect of partnering? In particular, how does the partnering of two black offi cers, two white offi cers, or a black and a white offi cer infl uence traffi c stop behavior? Similarly, what about the eff ect of stratifi cation? For example, does traffi c stop behavior vary when black offi cers occupy senior positions in the department?
Individual-level analysis is one of several future re- search avenues suggested by our study. We would also like to look beyond the vehicle stop and examine the relationship between the presence of minority police offi cers and vehicle searches. In addition, we are inter- ested in examining how organizational socialization aff ects additional law enforcement outcomes. Finally, it would be useful to collect data from additional locations to allow us to test whether the fi ndings of this study are infl uenced by self-selection. In other words, the blacks who choose to become police offi – cers in San Diego may be diff erent from those in other areas, and this could aff ect the fi ndings. 14
In sum, our fi ndings make several contributions to the theory of representative bureaucracy. Most notably, our fi ndings support our contention that institutional context aff ects whether passive representation leads to active representation. Specifi cally, our results empiri- cally support the hypothesis that organizational socialization can hinder the translation of passive representation into active representation. We provide evidence that institutional factors, such as socializa- tion, can reverse the benefi ts of passive representation. Further research into the eff ect of organizational socialization in additional policy areas is needed. Researchers should seek to identify cases in which organizational socialization can be treated as an inde- pendent variable in the analysis so that its direct infl u- ence can be ascertained. In addition, we see no reason why organizational socialization would not also aff ect the translation of passive into active representation for gender and ethnicity as well, although future research is needed to explore this in more detail.
Acknowledgments Th e authors thank Ken Meier, Brandon Bartels, and the three anonymous PAR reviewers for their insight- ful comments and constructive suggestions. We are also grateful to Chris Haley for her assistance with the data.
Notes 1. It is important to note that our data do not allow
us to directly test the eff ect of organizational socialization. Instead, we are able to test the relationship between key factors and the presence of racial profi ling in an organization in which socialization is known to be intense. To test the infl uence of organization socialization directly, we would need additional data.
2. According to the Racial Profi ling Data Collection Resource Center at Northeastern University, all 50 states and the District of Columbia currently have community, faith-based, or civil rights organizations that are actively seeking to address this issue. Likewise, hundreds of law enforcement agencies are engaged in collecting data on traffi c stops and searches in response to legislation, court settlements, executive directives, policy decisions, consent decrees, or voluntary eff orts.
3. Th ese statements come from interviews and focus group discussion held in Nashville, Tennessee, as part of the Bias-Based Policing Prevention Strate- gies Project ( Williams, Speer, and Peters 2003 ). Not all of the offi cers interviewed were affi liated with the Metropolitan Nashville Police Department.
4. Although it is possible that police offi cers were not completing a form for every stop they made, qualitative research collected during focus groups suggests that the vehicle stop forms collected should be representative of all vehicle stops.
662 Public Administration Review • July | August 2008
Th e focus group participants did not believe that offi cers were falsifying information on the stop forms or biasing the data by systematically completing forms for some stops and not others. Rather, participants felt that a few offi cers were never fi lling out the forms, whereas most offi cers were completing forms when they had time but not when they were busy ( Cordner, Williams, and Velasco 2002 ).
5. We use the term “black” instead of “African American” to be consistent with the vehicle stop form and because the police offi cers made their assessments on the basis of skin color, without any information regarding the driver’s ethnicity.
6. We realize that although this measure is an improvement on previous measures used in the literature on racial profi ling, it is still problematic. Th e number of 15-year-olds in the division is a rough surrogate for the driving population in the division and does not account for commuter traffi c. Ideally, we would have the resources to conduct traffi c surveys similar to those completed by Lamberth (1996) in New Jersey. Given the lack of resources necessary to collect these addi- tional data, we off er the best available alternative.
7. It would be preferable to have a time-varying measure for this variable; unfortunately, only the annual numbers are available. We took two steps to assess the stability of these numbers over time. First, we compared the percentage of black police offi cers in each division in 2000 with the same numbers for 2005 (the next year this report was available). We found these measures were still highly correlated at .82. In addition, we inter- viewed a data analyst with the San Diego Police Department to discuss transfer policies in the department ( Haley 2006 ). Th e data analyst reported that most transfers occur between service areas of the divisions and not between divisions. Transfers between divisions are consid- ered only on an as-needed basis as the depart- ment’s and the individual division’s needs are reevaluated. Both of these pieces of information suggest that the percentage of black police offi cers in each division is relatively stable within a given year.
8. Unfortunately, the data are not available to separate the offi cers by rank, so it is not possible to know whether the offi cers are on patrol or in a supervisory position. Th is would be important to test because researchers have found that when African Americans gain access to upper levels of an organization, they create an internal environ- ment that is more conducive to active representa- tion (Meier, Stewart, and England 1989). In addition, Wilkins and Keiser (2006) found that it was female supervisors in child support enforce- ment agencies, not female caseworkers, who provided active representation.
9. We divided median income by 1,000 to make it easier to interpret the coeffi cient.
10. Th e model was estimated numerous ways with and without panel-corrected standard errors, with a random eff ects model, with and without the lagged dependent variable, and with and without the AR1 error process, and the substantive fi ndings remained unchanged. In addition, alternative specifi cations of the model were tested. First, we ran the model using only the percentage of stops in the division involving a black driver as the dependent variable and using the same set of independent variables. Th is model produced substantive fi ndings identical to those of the reported model. Next, we again used the percentage of stops in the division involving a black driver as the dependent variable but, instead of controlling for the white/Caucasian percentage of the population, we controlled for the percentage of black driving-age population in the division. Th is specifi cation also produced similar substantive fi ndings, although the signifi – cance of several variables dropped because of problems with multicollinearity. Th e percentage of black driving-age population in the division is highly correlated with the percentage of the police offi cers in the division that are black (correlation = .91, VIF = 20.00). Given this, we chose to use the reported dependent variable. Th is allowed us to avoid the multicollinearity concerns and remain consistent with the racial profi ling literature.
11. We report both standardized and unstandardized coeffi cients in the table; for clarity, we refer to the unstandardized coeffi cients when interpreting the coeffi cients.
12. To check for multicollinearity in the model, we examined the bivariate correlations and the variance infl ation factors and no problems were detected.
13. Because our dependent variable is the diff erence between two terms, we test to see whether the variable of interest — the percentage of black police offi cers — is moving only with one of the terms. To do this, we examine the bivariate correlations between the percentage of black police offi cers in the division and each term of the dependent variable. We fi nd that the percent- age of black offi cers is correlated with the per- centage of stops involving a black driver and the percentage of the driving eligible population that is black at .89 and .91, respectively.
14. Th anks to an anonymous reviewer for raising this possibility.
References American Civil Liberties Union (ACLU) . 1999 .
Driving While Black: Racial Profi ling on Our Nation’s Highways . http://www.aclu.org/
Racial Profi ling and Representative Bureaucracy 663
racialjustice/racialprofi ling/15912pub19990607. html [ accessed March 20, 2008 ].
Barak , Gregg , Jeanne Flavin , and Paul Leighton . 2001 . Class, Race, Gender, and Crime: Social Realities of Justice in America . Los Angeles : Roxbury .
Beck , Nathaniel , and Jonathan N. Katz . 1995 . What to Do (and Not to Do) with Time-Series Cross- Section Data . American Political Science Review 89 ( 3 ): 634 – 47 .
Caplow , Th eodore . 1964 . Principles of Organization . New York : Harcourt, Brace & World .
Cayer , N. Joseph , and Lee Sigelman . 1980 . Minorities and Women in State and Local Government: 1973 – 1975 . Public Administration Review 40 ( 5 ): 443 – 50 .
Close , Billy R. , and Patrick L. Mason . 2003 . Beyond the Traffi c Stop: Intersections of Race, Ethnicity, Gender, and the Decision to Search . Tallahassee : Florida Department of Transportation .
Cordner , Gary , Brian Williams , and Maria Zuniga . 2000 . Vehicle Stops for the Year 2000: Executive Summary . San Diego, CA . San Diego Police Department .
Cordner , Gary , Brian Williams , and Alfredo Velasco . 2002 . Vehicle Stops in San Diego: 2001 . San Diego, CA . San Diego Police Department . http://www. sandiego.gov/police/pdf/stoprpt.pdf [ accessed March 20, 2008 ].
Cornwell , Christopher , and J . Edward Kellough . 1994 . Women and Minorities in Federal Government Agencies: Examining New Evidence from Panel Data . Public Administration Review 54 ( 3 ): 265 – 76 .
Cox , Stephen M. , Susan E. Pease , Daniel S. Miller , and C . Benjamin Tyson . 2001 . Interim Report of Traffi c Stop Statistics for the State of Connecticut . Rocky Hill, CT : Division of Criminal Justice . http://web.wtnh.com/Report_Narrative.pdf [ accessed March 24, 2008 ].
Culver , Leigh . 2004 . Adapting Police Services to New Immigration . New York : LFB Scholarly Publishing .
Downs , Anthony . 1967 . Inside Bureaucracy . Prospect Heights, IL : Waveland Press .
Engle , Robin Shepard , and Jennifer M. Calnon . 2004 . Comparing Benchmark Methodologies for Police- Citizen Contacts: Traffi c Stop Data Collection for the Pennsylvania State Police . Police Quarterly 7 ( 1 ): 97 – 125 .
Engle , Robin Shepard , Jennifer M. Calnon , and Th omas J. Bernard . 2002 . Th eory and Racial Profi ling: Shortcomings and Future Directions in Research . Justice Quarterly 19 ( 2 ): 250 – 73 .
Fagan , Jeff rey , and Gavin Davies . 2000 . Street Stops and Broken Windows: Terry, Race, and Disorder in New York City . Fordham Urban Law Journal 28 ( 2 ): 457 – 504 .
Gawthrop , Louis C . 1969 . Bureaucratic Behavior in the Executive Branch: An Analysis of Organizational Change . New York : Free Press .
Hahn , Harlan . 1971 . A Profi le of Urban Police . Law and Contemporary Problems 36 ( 4 ): 449 – 66 .
Haley , Chris . 2006 . Telephone interview with authors , May 17 , Athens, Georgia .
Hall , Grace , and Alan Saltzstein . 1975 . Equal Employment in Urban Governments: Th e Potential Problem of Interminority Competition . Public Personnel Management 4 ( 6 ): 386 – 93 .
Harris , David A . 1999 . Th e Stories, the Statistics, and the Law: Why “Driving While Black” Matters. Minnesota Law Review 84 ( 2 ): 265 – 326 .
Hindera , John J . 1993 . Representative Bureaucracy: Imprimis Evidence of Active Representation in the EEOC District Offi ce . Social Science Quarterly 74 ( 2 ): 95 – 108 .
Kaufman , Herbert . 1960 . Th e Forest Ranger: A Study in Administrative Behavior . Washington, DC : Resources for the Future .
Keiser , Lael , Vicky Wilkins , Kenneth J. Meier , and Catherine Holland . 2002 . Lipstick and Logarithms: Gender, Identity, and Representative Bureaucracy . American Political Science Review 96 ( 3 ): 553 – 65 .
Kellough , J. Edward . 1990 . Integration in the Public Workplace: Determinants of Minority and Female Employment in Federal Agencies . Public Administration Review 50 ( 5 ): 557 – 66 .
Kellough , J. Edward , and Euel Elliot . 1992 . Demographic and Organizational Infl uences on Racial/Ethnic and Gender Integration in Federal Agencies . Social Science Quarterly 73 ( 1 ): 1 – 11 .
Lamberth , John . 1996 . A Report to the ACLU . New York : American Civil Liberties Union .
Leinen , Stephen . 1984 . Black Police, White Society . New York : New York University Press .
Lipsky , Michael . 1980 . Street-Level Bureaucracy: Dilemmas of Individual in Public Services . New York : Russell Sage Foundation .
Lundman , Richard J. , and Robert L. Kaufman . 2003 . Driving While Black: Eff ects of Race, Ethnicity, and Gender on Citizen Self-Reports of Traffi c Stops and Police Actions . Criminology 41 ( 1 ): 195 – 220 .
Manning , P. K . 1970 . Talking and Becoming: A View of Organizational Socialization . In Understanding Everyday Life: Toward the Reconstruction of Sociological Knowledge , edited by Jack D. Douglas , 239 – 58 . Chicago : Aldine .
Meehan , Albert J. , and Michael C. Ponder . 2002 . Race and Place: Th e Ecology of Racial Profi ling African American Motorists . Justice Quarterly 19 ( 3 ): 399 – 430 .
Meier , Kenneth J . 1975 . Representative Bureaucracy: An Empirical Analysis . American Political Science Review 69 ( 2 ): 526 – 42 .
— — — . 1993 . Latinos and Representative Bureaucracy: Testing the Th ompson and Henderson Hypothesis . Journal of Public Administration Research and Th eory 3 ( 4 ): 393 – 414 .
664 Public Administration Review • July | August 2008
Meier , Kenneth J. , and Lloyd G. Nigro . 1976 . Representative Bureaucracy and Policy Preferences: A Study in the Attitudes of Federal Executives . Public Administration Review 36 ( 4 ): 458 – 69 .
Meier , Kenneth J. , and Joseph Stewart , Jr . 1992 . Th e Impact of Representative Bureaucracies: Educational Systems and Public Policies . American Review of Public Administration 22 ( 3 ): 157 – 71 .
Meier , Kenneth J ., Joseph Stewart , Jr ., and Robert England . 1989 . Race, Class and Education: Th e Politics of Second-Generation Discrimination . Madison : University of Wisconsin Press .
Mosher , Frederick C . 1982 . Democracy and the Public Service . 2nd ed . New York : Oxford University Press .
Naff , Katherine C . 2001 . To Look Like America: Dismantling Barriers for Women and Minorities in Government . Boulder, CO : Westview Press .
Novak , Kenneth . 2004 . Disparity and Racial Profi ling in Traffi c Enforcement . Police Quarterly 7 ( 1 ): 65 – 96 .
Pitkin , Hannah Fenichel . 1967 . Th e Concept of Representation . Berkeley : University of California Press .
Ramirez , Deborah , Jack McDevitt , and Amy Farrell . 2000 . A Resource Guide on Racial Profi ling Data Collection Systems: Promising Practices and Lessons Learned . Washington, DC : U.S. Department of Justice . http://www.ncjrs.gov/pdffi les1/bja/184768. pdf [ accessed March 24, 2008 ].
Rehfuss , John . 1986 . A Representative Bureaucracy: Women and Minority Executives in California Career Service . Public Administration Review 45 ( 5 ): 454 – 60 .
Reuss-Ianni , Elizabeth . 1983 . Two Cultures of Policing: Street Cops and Management Cops . New Brunswick, NJ : Transaction Books .
Riccucci , Norma M. , and Judith R. Saidel . 1997 . Th e Representativeness of State-Level Bureaucratic Leaders: A Missing Piece of Representative Bureaucracy Puzzle . Public Administration Review 57 ( 5 ): 423 – 30 .
Rojek , Jeff , Richard Rosenfeld , and Scot Decker . 2004 . Th e Infl uence of the Driver’s Race on Traffi c Stops in Missouri . Police Quarterly 7 ( 1 ): 126 – 47 .
Romzek , Barbara S . 1990 . Employee Investment and Commitment: Th e Ties Th at Bind . Public Administration Review 50 ( 3 ): 374 – 82 .
Russell-Brown , Katheryn . 1998 . Th e Color of Crime: Racial Hoaxes, White Fear, Black Protectionism, Police Harassment, and other Macroaggressions . New York : New York University Press .
Saltzstein , Grace Hall . 1989 . Black Mayors and Public Policies . Journal of Politics 51 ( 3 ): 525 – 47 .
Selden , Sally Coleman . 1997 . Th e Promise of Representative Bureaucracy: Diversity and Responsiveness in a Government Agency . Armonk, NY : M.E. Sharpe .
Simon , Herbert A . 1947 . Administrative Behavior: A Study of Decision-Making Processes in Administrative Organizations . New York : Free Press .
— — — . 1957 . Administrative Behavior: A Study of Decision-Making Processes in Administrative Organizations . 2nd ed . New York : Macmillan .
Skolnick , Jerome H . 1975 . Justice without Trial: Law Enforcement in Democratic Society . 2nd ed . New York : Wiley .
Smith , Michael R. , and Geoff rey P. Alpert . 2002 . Searching for Direction: Courts, Social Science, and the Adjudication of Racial Profi ling Claims . Justice Quarterly 19 ( 4 ): 673 – 703 .
Smith , Michael R. , and Matthew Petrocelli . 2001 . Racial Profi ling? A Multivariate Analysis of Police Traffi c Stop Data . Police Quarterly 4 ( 1 ): 4 – 27 .
New York Attorney General’s Offi ce . 1999 . Th e New York City Police Department’s “Stop and Frisk” Practices: A Report to the People of the State of New York from the Offi ce of the Attorney General . http://www.oag.state.ny.us/ press/reports/stop_frisk/stp_frsk.pdf [ accessed March 24, 2008 ].
Th ompson , Frank J . 1976 . Minority Groups in Public Bureaucracies: Are Passive and Active Representation Linked? Administration & Society 8 ( 2 ): 201 – 26 .
— — — . 1978 . Civil Servants and the Deprived: Socio-Political and Occupational Explanations of Attitudes toward Minority Hiring . American Journal of Political Science 22 ( 2 ): 325 – 47 .
Walker , Samuel . 1977 . A Critical History of Police Reform: Th e Emergence of Professionalism . Lexington, MA : Lexington Books .
— — — . 1980 . Popular Justice: A History of American Criminal Justice . New York : Oxford University Press .
Weber , Max . 1946 . From Max Weber: Essays in Sociology . Edited and translated by H. H. Gerth and C. Wright Mills . New York : Oxford University Press .
Websdale , Neil . 2001 . Policing the Poor: From Slave Plantation to Public Housing . Boston : Northeastern University Press .
Wilkins , Vicky M. , and Lael R. Keiser . 2006 . Linking Passive and Active Representation for Gender: Th e Case of Child Support Agencies . Journal of Public Administration Research and Th eory 16 ( 1 ): 87 – 102 .
Williams , Brian N . 1998 . Citizen Perspectives on Community Policing: A Case Study in Athens, Georgia . Albany : State University of New York Press .
Williams , Brian N ., Sheila Peters , and Paul W. Speer . 2003 . Bias-Based Policing in Nashville: Police and Community Perceptions . http://www.police. nashville.gov/news/reports/docs/ MNPDFinalFocusGroupReport.pdf [ accessed March 24, 2008 ].
Williams , Hubert , and Patrick V. Murphy . 1990 . Th e Evolving Strategy of Policing: A Minority View . Perspectives on Policing 13 . Washington, DC : National Institute of Justice .