THE RELATIVE IMPORTANCE OF RACE AND
ETHNICITY ON CITIZEN ATTITUDES TOWARD
THE POLICE
Vincent J. Webb
Chris E. Marshall
University of Nebraska at OmahaThe empirical study of citizen attitudes toward the police has a rich tradition and is regarded as important for several reasons. Some authors such as Murty, Roebuck, and Smith (1990) have suggested that positive images of the police are necessary in order for the police to function effectively and efficiently. Decker (1981) has argued that the police organization as a public sector organization needs community support to meet its goals. He identifies the “attitude-effectiveness” link as especially important in an urban society where the police are primarily reactive and dependent on the public for initiating police activity (p. 80). Dunham and Alpert (1988) have pointed out that citizens in neighborhoods that reflect distinct cultures have different values concerning the appropriateness of different police practices. These values are reflected in attitudes toward the police, and police practices that are incompatible with culturally-based attitudes may result in ineffective policing (p. 506). Murty, Roebuck, and Smith (1990) echo a similar proposition. In their view and that of others (Radelet, 1986 and Skolnick and Bayley, 1986), negative attitudes toward the police result in “mutual ill feelings, lack of respect, disorder, and inefficient police functioning” (Radelet, 1986:280). Greene and Decker (1989) point out that the nature of citizen attitudes toward the police and police attitudes toward citizens is an important determinant of whether the two groups can work together to set up crime control programs. They note that this is an especially important consideration given the contemporary emphasis on policing strategies that strive to bring “the police and the community into greater interaction.”
urban centers such as the Rodney King case highlight the continued crisis in American police community relations and underscore the need to continue to examine the nature and source of citizen attitudes toward the police (ATP). Understanding the source of the attitudes can result in the development of more effective policies and programs for effective policing. In this paper, we are especially interested in reexamining factors which have been previously found to be predictors of ATP in light of more recent studies examining the complexity of ATP in various populations.
Historically, research on attitudes toward the police has taken at least two different directions: most research has focused on identifying the determinants of ATP; but more recently, researchers have begun to explore the fundamental and complex structure of attitudes toward the police. The research reported here attempts to bridge these two directions. From a policy perspective, if ATP is primarily a function of police-citizen interactions, then improving ATP to improve the effectiveness of policing seems to call for strategies that modify police and citizen behavior. If the source of ATP is the transmission of culturally-derived beliefs and values, then policies that focus on behavior may be insufficient. Rather, interventions directed toward impacting upon socialization and cultural values may be in order. One goal of this research is to address these policy considerations.
THE DETERMINANTS OF ATTITUDES TOWARD
THE POLICE: SELECTED LITERATURE
1959 study by Fultz suggested that contact with the police was the most important determinant of ATP.
Besides race, several early studies examined the connections between gender, education, income, age, and occupation with ATP. Decker’s (1981) review of research on attitudes toward the police is organized around individual-level and contextual variables. His review identifies four individual-level variables – race, socioeconomic status, age, and gender – as important predictors of ATP. Beyond those studies already mentioned, the finding that ATP varies by race is widely supported in studies such as those of Hahn (1971); Zeit (1965); Jacob (1971); Campbell and Schuman (1972); Furstenburg and Wellford (1973); Hadar and Snortum (1975); Skogan (1978); Davis (1990); and Murty, Roebuck, and Smith (1990). It is important to note that the race variable for most of the early studies consisted of blacks and whites, with few other racial and ethnic groups being studied. Some exceptions are the inclusion of Hispanics by Carter (1983) and Cubans by Sullivan, Dunham, and Alpert (1987) and Dunham and Alpert (1988).
As Decker notes, the general finding that ATP varies by age, with younger citizens having less favorable attitudes toward the police than older citizens, has less consistent support. Studies by Campbell and Schuman (1972); Walker (1972); Smith and Hawkins (1973); Hadar and Snortum (1975); Feagin (1974); Dunham and Alpert (1988); Sullivan, Dunham and Alpert (1987); and Murty, Roebuck, and Smith (1990) have tended to support this finding, while other studies such as those by Jacob (1971) and Davis (1990) find little support for the age-ATP relationship.
The third individual-level variable considered by Decker (1981) in his review is socio-economic status. In Decker’s assessment, research in support of a SES-ATP connection is more equivocal than for the relationships between race and attitudes toward the police, and age and ATP. Two early studies (Walker, 1972; Jacob, 1971) provided some support for a SES-ATP connection. More recent studies including Davis (1991) and Dunham and Alpert (1988) seem to challenge the existence of an important relationship between socio-economic status and attitudes toward the police. As Decker (1981) points out, SES is intertwined with neighborhood culture that may be the more important predictor of ATP.
Contextual-level variables including neighborhood culture, experiences with police, victimization, and experience with police programs make-up the second category of variables considered by Decker in his 1981 review. Some early studies such as those conducted by Jacob (1971); and Schuman and Gruenberg (1972) show that the interaction between neighborhood and race is an important explanatory variable with regard to ATP. The more recent work of Dunham and Alpert (1988) also underscores the importance of neighborhood. In their study, they concluded that “It is clear that there is more variation on attitude toward police practice among neighborhoods than within the neighborhoods.” They concluded that the linkage was strong enough to justify the use of different police practices in different neighborhoods. A study conducted by Murty, Roebuck, and Davis (1990) produced similar findings. Their analysis showed that the most important determinant of the probability of negative or positive attitudes toward the police was a neighborhood/residence variable. Citizens residing in low-crime neighborhoods tended to have more positive attitudes toward the police than citizens who resided in high-crime neighborhoods. And, finally, Smith, Graham, and Adams have included in their models for attitudes toward police, neighborhood qualities such as the percent of neighborhood population that is nonwhite and the percent of persons below the poverty level (1991:25-26). These racial composition factor showed strength in explaining attitudes toward the police in the sense that nonwhites living in “areas with high concentration of nonwhites have less favorable view of police than nonwhites living in areas that are racially mixed” (Smith et al., 1991:35).
A second contextual-level variable that has received considerable attention in research on ATP is experience with the police or police contact. Decker has noted that citizen contacts with the police can be one of the two types, voluntary and involuntary (1981:83). In general, it appears that positive voluntary contacts have little impact on ATP, while negative contacts – voluntary or involuntary ones – have an important impact (Jacob, 1971; Walker, 1972; Furstenberg and Wellford, 1973; Parks, 1976; Winfree and Griffiths, 1977; and Murty, Roebuck, and Smith, 1990).
variables of neighborhood culture and contact with the police as predictors of ATP have research support, unlike victimization and experience with police programs.
THE NATURE AND MEASUREMENT OF
ATTITUDES TOWARD THE POLICE
Much of the research on attitudes toward the police mentioned above has emphasized independent variables – the determinants of attitudes toward the police – and subsequently paid less attention to the dependent variable, attitudes toward the police. Historically, ATP research has encompassed several different attitude dimensions and several different measures of the dependent variables. Generally, the dependent variable, attitudes toward the police, has been ignored in ATP research, or perhaps if not ignored, taken for granted. For example, the excellent Decker review of ATP research (1981) provides no discussion of the measurement of ATP. Few empirical studies of ATP have employed rigorous measurement scale development techniques, and only a few have evaluated ATP measures using standard techniques. It can be argued that until recently, ATP research has been unsophisticated with regard to the measurement of attitudes toward the police.
Traditionally, the measurement of ATP has employed many evaluative dimensions including assessments of police performance, police services, police resources, police officer characteristics, and police practices. Such measures have been applied uniformly across study populations without a great deal of consideration for the appropriateness of the measure for different groups, especially different racial and ethnic groups. As Sullivan, Dunham, and Alpert point out after considering ATP differences among ethnic groups:
In other words, research comparing blacks and whites may make inappropriate use of a common ATP measure if the underlying structure of ATP differs for blacks and whites. These authors point to research of Scaglion and Condon (1980) as providing evidence of dissimilar cognitive structures for blacks and whites.
In their 1987 article, Sullivan, Dunham, and Alpert report on research that examines the structure of attitudes toward the police for different age and ethnic groups. These researchers selected thirty questions that had been previously used in ATP studies, and administered the questions to samples of Anglo, black, and Cuban adults and students. Using factor analysis, they could identify seven relatively distinct ATP factors and show that the subpopulations in their study exhibited attitude structures that were similar in some respects and dissimilar in others. They concluded that age and ethnicity are the variables primarily responsible for differences in attitude structure.
In another article Dunham and Alpert report findings from research on ATP differences in five Miami neighborhoods that they conceptualize to be “culturally distinct” (1988:507). The neighborhoods studied varied based on ethnicity (Cuban, black, Anglo) and social class. Using attitude scales and statements gleaned from previous research that they regarded as having “withstood the test of time with regard to reliability and the validity of testing” (Dunham and Alpert, 1988:509), they administered thirty items to a sample of high school students, a sample of police officers, and a sample of residents. They then conducted a factor analysis and were able to identify five ATP domains: Demeanor, Responsibility, Discretion, Ethnic, and Patrol. Their analysis shows relatively distinct differences in ATP among neighborhoods, and considerable consensus in ATP within neighborhoods.
METHODS AND PROCEDURES
The research reported here is part of a more general study of urban conditions and services in the metropolitan area of Omaha, Nebraska (650,000 inhabitants). The sample includes 229 blacks, 210 Hispanics, and 351 whites. Nearly all of the blacks and Hispanics in the sample are from two relatively distinct neighborhoods, while the whites are from neighborhoods scattered across the city. The data analyzed here were collected through a telephone survey that was conducted in the spring of 1991.
ATP Measures
Five scales were used to measure attitudes toward the police. Four of these scales were replications of those developed and used by Dunham and Alpert (1988). The fifth scale was one that we have developed and used in three previous studies.
The four scales taken from Dunham and Alpert (1988) were Officer Demeanor (ODEM), Responsibility for Crime Control (RCC), Discretion (DISC), and Active Patrol Strategies (APS).2 It should be noted that the names of these scales have been slightly modified for purposes of this paper. The fifth scale taps attitudes toward the professional and personal characteristic of police officers, and is called the Officer Characteristics (OCHR) scale. Dunham and Alpert (1988:511-512) describe their scales as follows:
Demeanor (ODEM Scale): This scale consists of eight questions measuring the subject’s perceptions of the general demeanor of police officer or his/her orientation toward citizens.
Responsibility (RCC Scale): This scale consists of two questions concerning the role of the police and citizens in controlling crime.
Discretion (DISC Scale): Two questions are included in this scale that measure agreement with the need for variability in enforcing the law and especially in stretching procedural safeguards in some neighborhoods or areas.
Finally, the fifth scale Officer Characteristics (OCHR Scale) consists of nine items designed to measure agreement or disagreement with personal and professional characteristics or attributes of police officers such as honesty, prejudice, communication skills, response skills, and so on.
Dunham and Albert reported reliability coefficients (a’s) for their scales as .88 for Demeanor, .71 for Responsibility, .43 for Discretion, and .66 for Patrol. The reliability coefficients obtained in the present study are: Demeanor (.75); Responsibility (.85); Discretion (.17); Patrol (.54) and Officer Characteristics (.73). It should be noted that several items taken from Dunham and Albert were slightly modified to make them city/department specific. Appendix A provides the wording for each of the 23 items used in this study. A five point, Likert-type scale was used with the responses ranging from Strongly Agree to Strongly Disagree. Keying of responses was adjusted so that higher item or scale scores reflect positive attitudes toward the police.
Item and scale means and standard deviations for the complete sample as well as for blacks, whites, and Hispanics are presented in Appendix B; correlation matrices for the indices and items is provided in Appendix C. Nearly all of the items comprising the Demeanor and Character scales have means exceeding 3.0 which indicates slightly positive ATP. Never-the-less, a comparison of means indicates significant differences between ethnic groups on all but two of the items that make up these two scales. Averages for the items that comprise the Responsibility, Discretion, and Patrol scales are lower suggesting less agreement with the scale items. A comparison of the means indicates significant differences among ethnic groups on all but one of the items.
In general, the differences in item means on all of the scales is consistent with findings from previous research: whites tend to be most positive followed by Hispanics and then blacks. The analysis of item means also supports Sullivan, Dunham, and Albert’s (1987) finding that the ATP structures of different ethnic groups exhibit both similarities and differences.
Predictor Variables
Demographic variables. Two demographic variables, gender and age were used. Gender was coded 1 for male and 0 for female.3 Age was grouped into three categories with 1 representing the youngest and 3 the oldest.
Social Class. Two indicators were used, education and income. Education was coded using five values with 1 representing the lowest number of years (less than grade eight) and 5 the highest (graduate/professional school). Income was coded using eleven values with 1 representing the lowest and 11 the highest.
Neighborhood. Two indicators were used: percentage of persons in the respondent’s zip code area with income at less that 75% of the poverty level and the percentage of persons in the respondent’s zip code area who are white. The data for this pair of variables was derived from the 1990 U.S. Census of Population and Housing.
Contact with police. Contact with police used two indicators, whether or not a respondent had called for service, and whether or not a respondent had been stopped by police. For both indicators a value of one represents contact and 0 represents no contact.
Race-ethnicity. Based on self-identification, respondents were placed into one of three categories, black, Hispanic, and white. Since we were interested in the separate effects of each race-ethic category, we constructed two indicator variables: one for Hispanics and one for blacks. For both variables, a “1” indicates that the respondent was from the racial-ethnic group with the other racial-ethnic groups being the reference groups.
ANALYSIS AND FINDINGS
Initially, each of the five ATP scales was regressed on the five sets of variables using Ordinary Least Squares (OLS) regression. The results of this analysis are presented in Table 1. The total R2 was significant for each model.
The demographic and race variable sets make statistically significant contributions on the Officer Demeanor Scale (ODEM), with the race variable set being by far the largest contributor (.094) to the total amount of variance explained by the five variable sets (R2=.139). The police contact variables contribute very little in terms of R2 and are non-significant, while the demographic and social class variable sets contribute significantly, but much less strongly than does race.
An examination of the regression analysis for the Officer Characteristics (OCHR) scores shows that four of the five variable sets (demographic, social class, police contact, and race) contribute significantly to the total R2 explained by the model (.218). Nearly all of the variance explained is due to the race variable set (R2 change=.147).
The five variable sets account for very little of the variance in Responsibility for Crime Control (RCC) scores. The total R2 for the model is .124 with the social class and race variable sets making significant contributions. Again, a good part of the variance accounted for by the model is due to the race variable set (R2=.022); however, in this model, social class arises as a stronger contributor (R2=.096).
The total R2 for the Discretion scale (DISC) is only .047, the race and social class variable sets making the only significant contribution. Race, social class, and demographic factors are the variable sets making significant contributions to the regression model for Active Patrol Strategies (APS) (R2 change=.004 and R2 change=.020 and R2 change=.022, respectively).
Table 2 provides both standardized (beta) and raw regression (b) coefficients, and t tests for the five ATP regression models. Four variables – age, having been stopped by the police, Hispanic, and black – are significant predictors of Officer Demeanor (ODEM) scores. Older respondents have more favorable ODEM scores; those having been stopped by the police, Hispanics, and blacks are inclined to provide less positive ODEM scores in this model. The magnitudes of the betas showed that being black (–.25) is the most important predictor variable followed by being Hispanic (–.12), Age (.11), and Stopped (–.10).
American Journal of Police, V
ol. XIV
, No. 2
1995
55
ODEM OCHR RCC DISC APS
Independent R2 R2 R2 R2 R2
Variable Sets Change F Change F Change F Change F Change F
Demographic .017 5.20* .027 7.81** .002 .80 .003 .82 .022 7.00** Social Class .019 5.68* .019 5.79* .096 34.32** .014 4.14* .020 6.49* Neighborhood .003 .93 .008 2.74 .002 .55 .001 .45 .004 1.33 Police Contact .006 1.85 .017 5.25* .002 .76 .000 .15 .002 .68 Race .094 31.49** .147 52.40** .022 7.92** .029 8.94** .004 1.32 TOTAL .139 11.42** .218 15.66** .124 9.03** .047 2.92** .053 3.40**
* p < .05
Male and older respondents have more positive OCHR scores than females and younger respondents. Respondents having been stopped by the police, Hispanic respondents, and black respondents tend to have less favorable OCHR scores than do their counterparts. The relative size of these five betas indicate that black (–.31) is the most important predictor variable followed by Stopped by the Police (–.16), Hispanic (–.14), Age (.11), and Gender (.10).
The Hispanic variable and the two social class variables are the significant predictors of Responsibility for Crime Control (RCC) scores. The beta is .18 for the Hispanic factor; the betas for education and income are –.12 and –.19, respectively.
None of the variables is a significant predictor of the Discretion scores (DISC scale). With respect to the Active Patrol Strategy scores (APS scale), age is the one significant predictor with a beta of –.12, indicating that younger respondents have higher APS scores; the second significant predictor is education suggesting (beta=.17) that more highly educated respondents have higher APS scores.
DISCUSSION
The findings from this analysis are generally consistent with those found in the attitudes toward the police research literature. Race variables have the greatest effects, while age, gender, and police contact have significant, but smaller effects upon certain ATP dimensions. Contrary to some of recent research efforts, though, our results demonstrated an effect of social class influences on two of the attitude toward police dimensions.
Overall, this analysis indicates that the largest effects are for the ATP scales that tap officer demeanor and officer characteristics. None of the variables used in the analysis explains much of the variation in the ATP scales that tap responsibility for crime control, discretion, and active patrol strategies.
American Journal of Police, V
ol. XIV
, No. 2
1995
57
ODEM OCHR RCC DISC APS
Independent ß b t ß b t ß b t ß b t ß b t
Demographic
Gender .06 .55 1.58 .10 .89 2.52** .03 .12 .84 –.01 –.04 –.35 .06 .21 1.54 Age .11 .59 2.60** .11 .65 2.78** .02 .05 .59 .06 .10 1.45 –.12 –.24 –2.70**
Social Class
Education –.02 –.05 –.35 –.03 –.10 –.64 –.19 –.28 –4.65** .08 .08 1.70 .17 .23 3.92** Income .03 .04 .64 –.00 –.00 –.06 –.12 –.07 –2.76** –.02 –.01 –.47 –.02 –.01 –.33
Neighborhood
% LT Poverty –.04 –.02 –.38 –.17 –.10 –1.75 –.05 –.01 –.50 –.08 –.01 –.78 –.14 –.03 –1.37 % White .08 .01 .66 –.00 –.00 –.03 –.12 –.01 –1.01 –.02 –.00 –.18 –.21 –.01 –1.59
Contact Police
Called Police .02 .13 .38 .01 .13 .37 –.04 –.16 –1.17 .02 .06 .55 .03 .09 .67 Stopped –.10 –.89 –2.47** –.16 –1.45 –3.97*** –.01 –.04 –.31 .00 .01 .11 –.04 –.12 –.86
Race
Hispanic –.12 –1.17 –2.53** –.14 –1.43 –3.34** .18 .71 3.76*** –.02 –.06 _.38 .07 .25 1.37 Black –.25 –2.41 –2.83** –.31 –3.13 –3.44*** –.04 –.16 –.44 –.14 –.41 –1.47 .01 .04 .11
Hispanic respondents were most likely to agree with statements that asserted that only the police can control crime at the neighborhood level. The pattern is different when responses to statements about police discretion are examined: there were no significant contributors to the model specified. While the racial group – black or Hispanic – seemed to have the expected directional influence on this discretion scale, they did not significantly influence the response in conjunction with the remaining factors in the model. It would seem reasonable that future investigations would benefit from an incorporation of an interaction of the racial and neighborhood characteristics as well as racial and social class characteristics. For example, it seems that the interaction of the level of poverty with racial composition might have an almost additive effect upon attitudes toward the police. Moreover, physical environmental factors such as the condition of buildings or the quality and effectiveness of street lighting may also play a role.
Two significant effects were found for the statements dealing with active patrol strategies. Younger respondents were most likely to disagree with statements that reflected aggressive patrol strategies. This may be some artifact of the fact that the younger person will more frequently fall victim to the carrying out of these police strategies. The second significant effect suggested that lesser educated citizens are also more likely to disagree with statements reflecting aggressive patrol strategies. Again, the level of activity by police among less educated citizens is greater and increased, aggressive patrol strategies come to fruition in the neighborhoods of those citizens.
In general these findings reaffirm the importance of race and ethnicity on attitudes toward the police. We were unable to separate potential neighborhood effects from racial and ethnic effects in the present study since the vast majority reside in geographical areas that are not well defined as neighborhoods. Our input on this issue results from the neighborhood composition variables dealing with poverty and racial composition, but the results are constrained by measuring these through using zip code information that is simply not the same as a “neighborhood.” The neighborhood composition variables were not important predictors in any of the five models.
neighborhood is an important factor affecting attitudes toward the police. Neighborhoods as meaningful units of social organization vary in importance and character from one city to the next, and this fact needs to be taken into consideration when making generalizations about the importance of neighborhoods in the development and transmission of attitudes toward police.
Another factor that, in future work, may prove useful is the mobility of youth. It seems reasonable to expect that for youth, socialized in both formal and informal settings, will learn attitudes toward the police in a variety of venues. Of course, for many youth, school is the major, formal setting for such transference and shaping of attitudes. In addition, attitudes are conveyed in such favorite spots for young people as shopping malls, cruising strips, and street-corners. Moreover, these settings may or may not be within the immediate neighborhood of residence; in other words, some youth will be more mobile than their peers and, thereby, vary their socialization opportunities more than the less mobile youth. Omaha is a case in point: each high school serves several different neighborhoods, as do popular youth gathering spots such as malls and cruising strips. Future research in the area of attitudes toward the police should take this fluid mobility of youth through a myriad of sites of socialization. This mobility surely has some influence upon the type of attitudes and views which youth are exposed to and adopt.
From a policy perspective, the findings reported here would seem to indicate that improving attitudes toward the police requires the modification of shared values and perceptions that seem to be related only indirectly to police behavior. Two significant effects on ATP for contact with police were detected, and both of those concerned stops by the police. Policies aimed at modifying the nature of police-citizen contact, while socially desirable, may have little impact on ATP among a community’s minority residents, since these residents have less favorable attitudes regardless of police contact.
Future studies of ATP would do well to focus more specifically on the development of attitudes toward the police. Studies that identify the stages of development of ATP in youth may prove to be extremely useful in understanding the distribution of different dimensions of ATP among various populations. Additional studies of the various dimensions of ATP that result in reliable and valid instruments for measuring ATP is another area of research that is needed.
reaffirm the proposition that race and ethnicity play important roles; strategies for providing police services need to take this into consideration. It seems fairly clear that a more diversified approach to providing police services is reasonable, since not all racial/ethnic groups have the same attitudes toward police. It is also reasonable that community-relations programs aimed at treating the minority problem will, based upon this study and others like it, almost certainly enjoy limited effectiveness. Our study suggests that these community-relations programs will have to be less top-down in orientation, and less single-mindedly aimed at dealing with the minority problem. Instead, a more bottom-up orientation is suggested in improving the attitudes toward the police of various minority groups; policies dealing with this issue would be better conceived as incremental works-in-progress. These “works-in-progress” will draw and utilize information delineating the specific problems of the individual minorities. This deconstruction of the minority problem to the problems of many groups seems clearly indicated by the results of research reported herein. If police are going to meet the needs of the wide diversity of ethnic groups within the typical American urban center this reconceptualization of the problem will be important. Further, we believe that this deconstruction of the problem will have the consequence of improving the attitudes toward police among the members of the various ethnic and racial communities.
NOTES
The data used in this paper are from a survey conducted using funds provided by the Nebraska Urban Conditions Program, administered by the University of Nebraska at Omaha’s Center for Public Affairs Research (CPAR). The authors gratefully acknowledge the Center’s Director, Dr. Russell L. Smith, and the Center’s Research Staff including Jerome A. Deichert, E. David Fifer, and Tim K. Himberger for assistance in conducting this survey. Also, the authors would like to acknowledge the excellent contribution of Vic Bumphus in the early stages of the data analysis.
2. Originally, we planned to use the Ethnic scale reported by Dunham and Alpert (1988). Pre-testing indicated that a substantial number of respondents were likely to refuse to respond to the items for that scale, (e.g. “The police are justified in regarding a Hispanic as one who needs to be watched more than others”). This sort of item may not work very well when using telephone survey methodology.
3. In the case of nominal-level variables, we have dummy coded the data as described in Pedhazur (1982:274-279).
REFERENCES
Campbell, A. and H. Schuman (1972). “A Comparison of Black and White Attitudes and Experiences in the City.” In C. Harr (ed.), The End of Innocence: A
Suburban Reader. Glenview, IL: Scott Forseman.
Carter, D. (1983). “Hispanic Interaction With the Criminal Justice System in Texas: Experience, Attitudes, and Perceptions,” Journal of Criminal Justice, 11:213-227.
Decker, S. (1981). “Citizen Attitudes Toward The Police,” Journal of Police
Science and Administration, 9(1):81-87.
Dunham, R. and G. Alpert (1988). “Neighborhood Differences In Attitudes Toward Policing: Evidence For A Mixed-Strategy Model of Policing In A Multi-Ethnic Setting,” The Journal of Criminal Law and Criminology, 79(2):504-521.
Feagin, J. (1970). “Home Defense and the Police: Black and White Perspectives,”
American Behavioral Scientist, 13:797-814.
Fultz, L. (1959). Public Relations and the Police: A Survey of Public Opinion. Houston, TX: University of Houston.
Furstenberg, F. and C. Wellford (1973). “Calling the Police: The Evaluation of Police service,” Law and Society Review, 7:393-406.
Gourley, D. (1953). Public Relations and The Police. Springfield, IL: Charles C. Thomas Publishing, Co.
Greene, J. and S. Decker (1989). “Police and Community Perceptions of the Community Role in Policing: The Philadelphia Experience,” The Howard
Hadar, I. and J. Snortum (1974). “The Eye of the Beholder: Differential Perceptions of Police by the Police and Public,” Criminal Justice and Behavior, 2:37-54.
Hahn, H. (1971). “Ghetto Assessments of Police Protection and Authority,” Law
and Society Review, 6:183-184.
Jacob, H. (1971). “Black and White Perceptions of Justice in the City,” Law and
Society Review, 5:69-89.
Murty, S.; R. Komanduri; B. Julian; and J. Smith (1990). “The Image of the Police in Black Atlanta Communities,” Journal of Police Science and
Administration, 17(4):280-287.
Parks, R. (1976). “Police Response to Victimization Effects on Citizen Attitudes and Perceptions.” In W. Skogan (ed.), Sample Surveys of the Victims of Crime. Cambridge, MA: Ballinger.
Pedhazur, E. (1982). Multiple Regression in Behavioral Research: Explanation
and Prediction (2nd Edition). New York, NY: Holt, Rinehart and Winston.
President’s Commission on Law Enforcement and Administration of Justice (1967). Task Force Report: The Police. Washington DC: Government Printing Office.
Radelet, L. (1986). The Police and The Community. New York, NY: Macmillan Publishing Co.
Skogan, W. (1978). “Citizen Satisfaction with Police Services: Individual and Contextual Effects,” Police Studies Journal, 8:469-479.
Skolnick, J. and D. Bayley (1986). The New Blue Line: Police Innovation in Six
American Cities. New York, NY: The Free Press.
Smith, D.; N. Graham; and B. Adams (1991). “Minorities and the Police: Attitudinal and Behavioral Questions.” In M. Lynch and E. Britt Patterson (eds.),
Race and Criminal Justice. New York, NY: Harrow and Heston.
Smith, P. and R. Hawkins (1973). “Victimization, Type of Police Contacts and Attitudes Toward the Police,” Law and Society Review, 8:135-152.
Sullivan, P.; R. Dunham; and G. Alpert (1987). “Age Structure of Different Ethnic and Age Groups Concerning Police,” The Journal of Criminal Law and
Criminology, 78(1):177-196.
Winfree, T. and C. Griffiths (1971). “Adolescent Attitudes Toward the Police.” In T. Ferdinand (ed.), Juvenile Delinquency: Little Brother Grows Up. Beverly Hills, CA: Sage Publishing Co.
Zeitz, L. (1965). “Survey of Negro Attitudes Toward the Law,” Rutgers Law
Review, 19:288-316.
Appendix A:
ATP SCALE ITEMS
Officer Demeanor:
Omaha police officers are usually courteous
Omaha police officers are respectful toward people like me
Omaha police officers use more than they need to in carrying out their duties
Most police officers are usually friendly Most police officers are usually rude
Most police officers give people a chance to explain Police officers enjoy kicking people around
Police offices show concern when you ask them questions
Officer Characteristics:
Omaha police officers are honest Omaha police officers are physically fit
Omaha police officers are prejudiced against minority persons Omaha police officers are hardworking
Omaha police officers are well trained
Omaha police officers are not able to answer citizens’ questions correctly Omaha police officers don’t communicate very well
Omaha police officers respond quickly to calls for service Most police officers are liars
Responsibility for Crime Control:
Only the police can control crime in this area Only the police can control crime in my neighborhood
Discretion:
In order to prevent crimes and catch criminals, the police are sometimes required to stretch the search and seizure laws and other procedural safeguards The police are more strict in some neighborhoods than in others
Active Patrol Strategies:
In some neighborhoods, crime preventing requires that police officers stop people walking down the street and ask them where they are going or what they are doing
ol. XIV
, No. 2
1995
Overall Omaha Sample White Black Hispanic
ATP ITEMS Mean SD Mean SD Mean SD Mean SD F
Officer Demeanor Scale: 27.58 4.26 28.99 3.29 25.65 5.13 27.48 3.75 40.83**
Omaha police officers are usually 3.39 1.12 3.39 1.15 3.33 .98 3.43 .95 .55 Omaha police officers are respectful 3.54 .97 3.56 .98 3.37 .90 3.50 .81 2.97* Omaha police officers use more force 3.41 .91 3.53 .84 2.70 .99 3.09 .88 57.07** Most police officers are usually friendly 3.68 .75 3.72 .73 3.46 .89 3.64 .72 7.81* Most police officers are usually rude1 3.82 .68 3.89 .60 3.36 .90 3.57 .71 37.39**
Most police officers give people a 3.42 .88 3.45 .87 3.21 .92 3.34 .86 4.88* Police officers enjoy kicking people 3.82 .74 3.93 .64 3.13 .91 3.50 .79 71.13** Police officers show concern when you 3.60 .78 3.64 .76 3.36 .86 3.44 .78 9.93**
Officer Characteristics Scale: 30.47 4.49 32.24 3.63 28.08 5.02 30.27 3.86 57.89**
American Journal of Police, V
ol. XIV
, No. 2
1995
65
Only the police can control crime in 2.35 .90 2.31 .88 2.56 .99 2.76 .98 15.67**
Officer Discretion Scale: 5.09 1.28 5.35 1.32 4.68 1.12 5.12 1.28 17.92**
Police are sometimes required to stretch 2.82 .98 2.87 .99 2.57 .89 2.71 .91 6.57* The police are more strict in some 2.42 .80 2.47 .81 4.68 .60 2.39 .78 16.99**
Officer Patrol Strategies: 5.82 1.62 5.72 1.54 5.94 1.73 5.85 1.60 1.22
Police stop people walking down the 2.70 .96 2.64 .93 3.07 1.03 2.96 .94 14.54** Good police officer one who patrols 3.04 .98 3.07 .98 2.88 1.00 2.91 .97 3.05*
Notes:
1Scoring reversed to correspond to other scale items.
ol. XIV
, No. 2
1995
Officer Officer Officer Officer Officer
Demean Charac- Respon- Discre- Patrol < –––– > Independent Variables
teristics sibility tion Strategy
.042 .042 .044 –.041 .079a Gender 1.000
.122b .142a .062 .041 –1.29b Age –.055 1.000
.054 .054 –.262b .110b .140b Education .033 –.167b 1.000
.112b .114b .220b .068 .047 Income .091a –.144b .348b 1.000
–.290b –.361b .094b –.200b .037 % LT Poverty .029 .012 –.269b –.400b 1.000
.286b .336b –.050 .197b –.050 % White –.014 –.059 .228b .296b –.909b1.000
.000 –.026 –.030 .024 .015 Call Police? –.039 –.089a .012 .045 .029 –.005 1.000
–.070 –.120b –.111b .016 .020 Police Stop? .228b –.222b .129b .160b –.055 .071a .200b1.000
–.287b –.333b .030 –.207b .039 Black .015 .062 –.189b –.253b .827b–.910b–.014 –.074a1.000
–.012 –.023 .192b .014 .006 Hispanic .084a –.190b –.200b –.142b –.131b .317b .050 –.018 –.380b
