Public Image Of Police

Print
The Public Image of the Police
Final Report to
The International Association of Chiefs of Police
By
The Administration of Justice Program
George Mason University
Authors (Alphabetical Order):
Catherine Gallagher
Edward R. Maguire
Stephen D. Mastrofski
Michael D. Reisig
October 2, 2001
Contact Person:
Stephen D. Mastrofski
Administration of Justice Program
George Mason University
10900 University Boulevard, MS 4F4
Manassas, VA 20110-2203
Telephone: [phone]
Fax: [phone]
Email: [e-mail]

CONTENTS
EXECUTIVE SUMMARY
CHAPTER 1. INTRODUCTION
Methodology
CHAPTER 2. THE GENERAL IMAGE OF POLICE
I. Introduction
II. General versus Specific Measures
III. How to Measure General Police Image
IV. General Police Image Over Time
V. Factors Influencing the General Image of the Police
Personal Characteristics of the Citizen
Nature of the Citizen’s Recent Contact with Police
Mass Media Portrayals of Police and Crime
VI. Police Image Compared to Other Major Social Institutions
VII. Police Image from Community Surveys
VIII. Conclusion
CHAPTER 3. PUBLIC PERCEPTIONS OF THE OUTCOMES OF POLICING
I. Introduction
II. Different Ways to Measure Police Outcomes
III. Police-crime Outcome-oriented Elements: Results from National Polling Data
IV. Community Outcome-oriented Elements and the Public’s
General Image of the Police
V. Responsibility for Crime Control: Neighborhood- and Citizen-level Differences
VI. Conclusions
CHAPTER 4. PUBLIC PERCEPTIONS OF POLICING PROCESSES
I. Introduction
II. Generic Dimensions of the Quality of Service
Attentiveness
Reliability
Responsiveness
Competence
Manners
Fairness
Integrity
III. Police-specific Dimensions of the Quality of Service
Stops and Searches
Use of Force
IV. Race and the Image of Police
V. The Relationship between Police Processes and the General
Image of the Police
VI. High Visibility Events and the Police Image Regarding Processes
VII. The Consequences of the Police Image
VIII. Conclusions
CHAPTER 5. IMPROVING PUBLIC PERCEPTIONS OF THE POLICE IMAGE:THE IMPACT OF COMMUNITY POLICING
I. The Impact of Community Policing on General, Outcome, and Process Measures
II. Conclusions
CHAPTER 6. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH
I. Review of Findings and Implications
General Police Image
Perceptions of the Outcomes of Policing
Public Perceptions of Policing Processes
Improving the Public Perception of the Police
A Perspective on the Findings
II. Recommendations for Future Research
Research Questions
Agenda for Future Data Collection
REFERENCES

EXECUTIVE SUMMARY
Study’s Purpose
The purpose of this study is to provide an overview of published research on the public image of the police. The report covers three types of police images: general perceptions of the police as an organization or institution, perceptions of police outcomes, and perceptions of police processes. The report considers research that reflects on improving the image of police. It summarizes the findings and discusses the implications for future research.
Methodology
Two types of reviews were conducted: a review of published research and a review of archived data sets pertaining to the image of the police held by the public. A comprehensive search of social science research literature was conducted to obtain a base for the literature review. We attempted to obtain all of the publications drawing on national surveys of police. We were selective in drawing upon surveys relevant to specific police agencies, using these where national surveys did not provide insights to important questions.
A thorough search of publicly available archives of national and major international surveys of the police image was also conducted. Surveys of samples drawn on a state, county, or municipality were not considered unless they offered some valuable insights to broader questions about the police image. Where available we obtained copies of the survey instruments (or those parts relevant to the police image) and basic characteristics of the sample. From this information we prepared a catalog that will allow IACP to view the entire scope of existing survey data on the police image that are already available. This catalog is provided separately in a form that is electronically accessible. Selected data from these surveys are presented in Exhibits in this report.
Major Findings and Recommendations
The public image of the police is complex. It has many aspects, grouped under three general categories: overall image, perceptions of police outcomes, and perceptions of police processes. There are different ways to measure each aspect. Findings can vary considerably according to which aspect is measured and how each is measured.
Polls of the adult population in the United States since the 1960s show that the majority of the public has an over-all positive view of the police. Depending on the year and the particular measure used, the percentage of respondents with a positive assessment of police has been between 51 and 81 percent. When asked to assess service to their own neighborhoods, respondents tend to produce even higher evaluations. Relatively few citizens offer a negative assessment of police.
The police consistently rank among the institutions and occupations in which the public expresses the highest confidence and trust.
Most citizens are satisfied with police service in their own neighborhood, and this level of satisfaction appears to vary little from one urban jurisdiction to another. Cross-jurisdiction research on this topic is limited to a small number of jurisdictions, however.
Citizens’ experiences with the police affect their over all assessment of the police. The more positive a citizen’s recent experience with the police, the more positive the citizen’s over-all assessment of the police. However, previously held views of police do not change easily and themselves tend to influence how citizens interpret their own experiences with the police.
The vast majority of the American public has not had a face-to-face contact with a police officer in the previous twelve months, so it will be difficult for police to make large improvements in their over all public image by the direct contact they have with the public.
Large portions of the American public report using the mass media as their primary source of information about crime, and these stories are the context for most mass media accounts of police work. News and entertainment media portray police and police work in a highly distorted fashion. The recent trend toward “tabloid-style” journalism – even in mainstream media – appears to reduce public confidence and trust in the police.
Between the 1980s and mid-1990s, increasing numbers of the American public gave police protection in their area a positive assessment.
Neighborhood residents hold both police and residents responsible for controlling crime in the neighborhood.
At the end of the 20th century, substantial majorities of the American public expressed positive views of how police treat the public. Police ranked highest in being helpful and friendly and lowest in treating people fairly.
The public image of honesty and ethical standards of police has improved substantially from 1997 to 2000.
The majority of the American public does not perceive police brutality in their area, but from the mid-1960s to the end of the 20th century the percentage who do perceive brutality has increased approximately threefold, accounting for a third of the public. This increase may be due at least in part to the public’s changing standards of what constitutes brutality. The public has become less accepting of police use of force during this time period.
Across nearly all indicators of the public image of the police, racial minorities consistently show lower assessments of police than do whites. These race effects appear to be particularly enduring for citizens’ assessments of police fairness and use of force.
The over-all legitimacy of the police depends much more on citizens’ perceptions of how the police treat them than on their perceptions of police success in reducing crime.
Public confidence in and support for the police depends more on citizens’ perceptions of police officers’ motives than whether the outcome was personally favorable to the citizen.
The public’s perceptions of how police treat them appear to affect their willingness to obey the law and obey the police.
Negative publicity about the police in one city that receives high visibility around the nation may have a nation-wide impact on the public’s view of the police, but the effect appears to be modest and not enduring.
When the public perceives major threats to the nation’s security, the overwhelming majority appear willing to give additional powers to the police that invade privacy and restrict liberty, but substantial portions of the public are also concerned about the possibility of police abuses of these powers.
Community policing may have some modest, long-term positive influence on citizens’ satisfaction with police, but it is unlikely to produce a “quick fix.”
The following represents a distillation of the major findings of this
study.
Between the 1980s and mid-1990s, increasing numbers of the American public gave police protection in their area a positive assessment.
Neighborhood residents hold both police and residents responsible for controlling crime in the neighborhood.
At the end of the 20th century, substantial majorities of the American public expressed positive views of how police treat the public. Police ranked highest in being helpful and friendly and lowest in treating people fairly.
The public image of honesty and ethical standards of police has fluctuated over the years but has improved substantially from 1977 to 2000.
At the end of the 20th century, a majority of the American public perceives racial profiling to be a widespread practice and a problem.
The majority of the American public does not perceive police brutality in their area, but from the mid-1960s to the end of the 20th century the percentage who do perceive brutality has increased approximately threefold, accounting now for a third of the public. This increase may be due at least in part to the public’s changing standards of what constitutes brutality. The public has become less accepting of police use of force during this time period.
Across nearly all indicators of the public image of the police, racial minorities consistently show lower assessments of police than do whites. These race effects appear to be particularly enduring for citizens’ assessments of police fairness and use of force.
The over-all legitimacy of the police depends much more on citizens’ perceptions of how the police treat them than on their perceptions of police success in reducing crime. Public confidence in and support for the police depends more on citizens’ perceptions of police officers’ motives than whether the outcome was personally favorable to the citizen.
The public’s perceptions of how police treat them appear to affect their willingness to obey the law and obey the police.
Negative publicity about the police in one city that receives high visibility around the nation may have a nation-wide impact on the public’s view of the police, but the effect appears to be modest and not enduring.
When the public perceives major threats to the nation’s security, the overwhelming majority appear willing to give additional powers to the police that invade privacy and restrict liberty, but substantial portions of the public are also concerned about the possibility of police abuses of these powers.
Community policing may have some modest, long-term positive influence on citizens’ satisfaction with police, but it is unlikely to produce a “quick fix.”
The following summarizes the major limitations of the available research and lists recommendations for future research.
Different measures of the public’s image of the police can produce radically different results. Research is needed to identify the best survey items to accomplish specific research and evaluation purposes. Doing this will provide more valid and reliable measures for learning what the public image of the police is and what influences that image.
Very little is known about the relative importance of various sources of information on the police’s public image. Research is needed to learn how much influence is exerted by the public’s personal experiences with the police, what they learn second-hand from friends and acquaintances, and what they learn from the mass media. Knowing how much and in what ways each of these sources influence public opinion about the police will help police develop more effective strategies for improving the public’s evaluations of and support for the police.
Very little is known about the influence of nationally publicized events on the police image. Knowing how both negative and positive publicity in one community affects the public’s image of police in other communities will help police leaders learn how to deal more effectively with the consequences of those events in their local communities.
Very little is known about how much variation there is in levels of citizen satisfaction with the police from community to community, and even less is known about what types of communities and police agencies show the highest and lowest levels of satisfaction. Research on this topic will help to validate what most effectively enhances the police image. Given the tremendous diversity of communities and police agencies, the research must distinguish what works in different kinds of communities. Virtually all of the survey research on the police image has concentrated on relatively large urban jurisdictions.
Very little is known about contextual influences on patterns of public opinion about the police. Patterns may be different when crime is high compared to when crime is low, when there are strongly perceived threats to national security and when there are not.
Very little is known about the relationship between objective and subjective indicators of police performance. When the crime rate is going up or down does the public credit the police with this effect? Because police tend to rely heavily on objective measures of performance in dealing with crime and solving problems, it is important to know whether success or failure objectively measured translates into public credit and accountability when measured subjectively through public opinion surveys.
Little is known about the implications of public opinion for public behavior that is of concern to police. Are there thresholds of public satisfaction or dissatisfaction in a community that indicate a considerably increased likelihood of citizen support or resistance to the police? What are the consequences of shifts in the police image for the tenure of police leadership? Answers to these questions will help police leaders use poll results to predict short and long-term trends in citizens’ behaviors that are important to police.
The report concludes with a proposal for IACP to take a lead role in developing a data collection system that would enable its membership to track its progress in improving the police image and make it possible for researchers to answer the research questions listed above. The working name for this program is the Uniform Public Opinion Poll on Policing (UPOPP). The UPOPP system would be a voluntary program that would provide survey research planning to participating agencies. Those agencies would agree to conduct an annual public opinion survey in their jurisdictions. In addition to a common set of survey questions for all agencies, these surveys could also include questions crafted to suit the special needs of that department and the community it serves. Data would be archived by a research organization selected by IACP. In addition to providing advice on the design and implementation of the annual survey, the research organization would analyze the archived data, issuing an annual report on the state of the public image of police.
The following sections of the executive summary provide a more detailed description of findings and recommendations. Findings are divided into major sections on the general image of the police, perceptions of the outcomes of policing, perceptions of policing processes, and improving the public perception of the police. This is followed by a discussion that places the findings in perspective. The executive summary concludes with a discussion of priority issues for future research and an agenda for data collection.
The General Image of Police
The general image of the police offers an overview of the public’s perception of the police. Particular characteristics of the people, organization, or institution remain undifferentiated. Measures of the general image are useful because they provide a summary measure of the level of overall favorableness or support that the public holds for the police.
How one measures image makes a difference.
The more general the question, the more positive the response tends to be.
Slight changes in the wording of the question or the response options can make a big difference in how positive the image appears to be.
Questions that measure the “favorableness” of the police image tend to generate more positive responses than questions that ask about “confidence” in the police.
The majority of the public has a substantial degree of confidence in the police as a general institution.
Only a small percentage of polled citizens reports having very little or no confidence in the police in their community.
The proportion of people with confidence in the police can change several percentage points from year to year, but it often changes only 2-3 points per year.
Confidence in police has been declining slowly since 1996 (from 60 to 54 percent).
The trend in respect for the police is that the level of positive views of the police have been declining since the mid-to-late 1960s.
A more positive general image of the police is associated with the following characteristics of the public:
Being older. National samples of high school seniors consistently rate the job police (generally) are doing as substantially lower than do national samples of older persons.
Being of higher wealth or socio-economic status
Living in suburban (as opposed to urban) areas
Being white (as opposed to black)
Having positive attitudes about one’s own neighborhood
One major study of Chicago suggests that there is no difference between blacks and whites when socio-economic disadvantage of the neighborhood is taken into account. That is, blacks’ negativity toward police appears to be due to their concentration in disadvantaged areas.
Negative attitudes about the police by disadvantaged persons appears to be part of a more diffuse alienation from government, law, and the political process generally.
Citizens’ experiences with the police influence their general image of the police.
One study indicates that police courtesy/friendliness toward the citizen in a recent contact with police exerts the most powerful influence on the citizen’s general evaluation of the police. This holds for situations where the citizen’s contact was involuntary (traffic stops) and voluntary (breaking-and-entering complaints).
However, two studies have indicated that people’s prior general views of police have stronger influence on their evaluation of a subsequent specific contact than their evaluation of a specific contact has on subsequent general views of police.
The vast majority of the American public has not had face-to-face contact with a police officer in the previous twelve months.
Most citizens regard the mass media as their prime source of information about crime, and crime news is the context for most mass media accounts of police work.
The implicit message of much crime news is the inability to catch offenders.
There is an increasing trend in the news media to concentrate their coverage on a few sensational cases in a tabloid style of journalism. The net impact of tabloid-style coverage appears to be a decline in confidence in the police.
Entertainment media present images of police that distort the realities of every-day police work. Although more positively presented than attorneys and judges, police are more often than not presented as incompetent rule-breakers.
Since 1993 the police consistently rate among the top three institutions out of thirteen in public confidence (with the military clearly at the top and churches/organized religion and the police vying for second place). Police rate much higher than the rest of the criminal justice system.
Large majorities of adult citizens are satisfied or very satisfied with police service in their neighborhoods.
Although there is variation in satisfaction levels across urban jurisdictions, most fall within the 80-90 percent range.
The majority of school-age children in Cincinnati trust their local police, but a large portion do not, and this distrust is particularly strong among nonwhite students.
Public Perceptions of the Outcomes of Policing
Outcomes are identified by knowing the goals that people hold for the public – the consequences of doing police work. Police nowadays are expected to accomplish a variety of outcomes, including reducing crime and disorder, reducing fear of crime, solving neighborhood problems and improving quality of life, and developing greater community cohesion. However, most of the research on the public perception of the police image has focused on the impact of the police on crime and safety.
From the 1980s to the mid-1990s, increasing numbers of the public have offered positive assessments of the quality of police protection.
Confidence in the ability of the police to achieve traditional crime-focused goals appears to be high based on a 1995 survey. 74 percent expressed confidence in the police to protect citizens from crime, 74 percent to solve crime, and 65 percent to prevent crime.
African Americans reported lower ratings than whites and Hispanics.
Residents appear to hold police at least partially responsible for outcomes at the neighborhood level.
There is a relationship between satisfaction levels of police service in the neighborhood and residents’ ratings of crime, disorder, and physical decay in the neighborhood.
Research at the community level suggests that residents believe that citizens share responsibility for controlling crime.
This relationship holds across neighborhoods with different levels of family income, home ownership, and length of occupancy – differing only according to ethnicity.
Hispanics appear more likely to hold police solely responsible for controlling crime.
Public Perceptions of Policing Processes
The processes of policing are how police do their work. The aspects of police processes that one might study are virtually infinite, but the public cares most about those that are captured by the notion of “service.” Service has many dimensions, some of which are generalizable to a wide variety of human services, not just policing: attentiveness, reliability, responsiveness, competence, manners, fairness, and integrity. Some features of service are peculiar to police – those aspects of their authority that empower them to intrude on citizens’ privacy and coerce them: stops and searches and use of force, for example.
The research on most aspects of the public’s view of policing processes is limited in scope. The following represent the major findings from the limited research available.
Police attentiveness to crime victims (providing them counseling on how to undo the negative effects of victimization) has a positive effect on their attitude toward the police.
Citizens’ expectations about how the police will perform affects their evaluation of how they perform in a specific situation.
Positive evaluations are associated with perceived performance that meets or exceeds prior expectations.
A substantial majority of the public rates their police as doing an excellent or pretty good job of being “helpful and friendly.”
A substantial majority of the public express positive attitudes about the fairness of the police, but a significant portion rate them as “only fair or poor.”
African Americans, younger people, singles, and low-income respondents tend to offer less positive evaluations of police fairness. The difference is particularly striking between African Americans and whites.
Racial disparities in assessments of police fairness may be caused in part by indirect exposure to unfair treatment by receiving second-hand accounts from others in their neighborhood.
Almost one in five respondents fear that the police will stop and arrest them when they are completely innocent.
The public image of the honesty and ethical standards of police has fluctuated considerably, but has improved substantially from 1977 to 2000.
In 2000 the police ranked 10th out of 32 occupations rated on honesty and ethical standards.
Approximately one in ten respondents in a national survey reported that they had been stopped by police because of their racial or ethnic background.
Blacks were seven times more likely to report this than whites.
Most vehicle operators stopped by police felt that they had been stopped legitimately, but there were significant differences by race and gender.
Men and blacks were less likely to feel that their stops were legitimate
Most vehicle operators who are searched by police feel that the search was not legitimate.
Blacks were substantially more likely to view the search as illegitimate than whites.
A study in England found that citizens were more likely to feel fairly treated when officers gave a good reason for the stop.
In 1999 59 percent of the American public perceived racial profiling by the police as “widespread,” and in 2000 75 percent viewed it as a problem in the United States.
Blacks are much more likely than whites to perceive that racial profiling by police is widespread.
Nearly all citizens who experience police force view the police behavior as improper.
Nearly all citizens view police force as appropriate when a citizen attacks the officer, and the majority approve when a suspect escapes from custody, but few citizens approve when a citizen says vulgar or obscene things to a police officer.
The public’s acceptance of police force is declining over time, especially for suspect escapes and the use of vulgar or obscene language.
Lowered thresholds of what constitutes brutality in the public’s mind may account for some of the significant increase in the public’s perception of brutality.
The citizen’s race is a significant influence on the citizen’s assessment of the quality of the police process.
Hispanics, and especially African Americans, evaluate police less favorably on the use of force, fairness, friendliness, and promptness.
For citizens who have had contact with the police within the previous two years, when the level of satisfaction with that contact is taken into account, race is no longer an significant influence on citizens’ assessments of police friendliness and promptness. However, race remains a significant influence on assessments of use of force and fairness.
A small, but growing number of studies indicates that citizens’ assessments of police processes have a powerful influence on their view of police legitimacy.
Trust in the motives of legal authorities, such as the police, has more impact on police legitimacy than the citizen’s view of the fairness or favorability of the outcome for that person. This is uniform across race/ethnic groups.
Citizens’ assessments of police competence and fairness are both significant predictors of the public’s general confidence in and support for the police, but the fairness assessment is by far the more powerful of the two predictors. This holds for both whites and racial minorities.
A study in Oakland during a time when police were using aggressive tactics to suppress gangs and gun-related crimes showed that public support for the police was influenced far more by how police interact with the public than whether crime was reduced.
The Rodney King incident may have had a nationwide effect on the public’s view of police honesty and integrity in the few years following the event, but the effect was modest and not enduring. Other high visibility events (Louima, Diallo, and Ramparts) showed no readily discernable effect.
In the immediate aftermath of the September 11, 2001 terrorist attacks on New York City and the Pentagon, the overwhelming majority of the public was willing to give additional powers to police to conduct surveillance, identify, and apprehend terrorists.
However, substantial proportions of the public were concerned about the possibility of police abuses of these powers.
A growing body of research suggests that how the public feels about the way police treat them affects the public’s behavior (obeying the law and obeying the police).
Most of this research is based on studies of citizen contacts with the police.
Improving Public Perceptions of the Police Image: The Impact of Community Policing
This report reviewed studies that evaluate the impact of community policing on the public’s perception of the police. Community policing’s philosophy emphasizes the importance of meeting the community’s needs, and there has recently been some research that allows an assessment of its impact. The aspect of community policing that has been evaluated is the attempt to increase citizen participation at the neighborhood level.
Community policing may have some modest positive influence on citizens’ satisfaction with the police.
Some evidence suggests that adoption of community policing programs is associated with perceptions of improved quality of neighborhood life and improvement in the image of the police.
Community policing reforms are unlikely to provide a quick fix, but entail a long-term commitment by police to work with citizens to address neighborhood ills.
A Perspective on the Findings
The public image of the police is complex, making generalizations difficult. There is no single best measure of the police image. Researchers should carefully select the best measures to suit the specific purposes of their research, taking care to avoid cynical selection of only those measures that tend to show the police in the most positive light.
At the beginning of the 21st century, the image of American police can be characterized as positive or mixed, depending upon one’s interpretation.
There is reason for both alarm and celebration. Many indicators show that American police are among the most trusted and admired institutions of contemporary society, while there are also many indicators that the American public – especially the young and disadvantaged – are wary of the police and see plenty of room for improvement. Although police appear to enjoy legitimacy with the majority of people in even the groups who are most disaffected, police leaders should not be complacent. Substantial portions of the disadvantaged are not so positive, and it is precisely these people whose cooperation and good will the police need in general and in the every-day work of the street officer. Even relatively low levels of public dissatisfaction with police are problematic if they are concentrated among groups who have a self-identity as “victims” of
policing.
There appear to be at least three ways in which the public forms negative impressions of police: the direct experiences of the public with the police, how the police are presented to the public through the press and entertainment media, and the standards and expectations the public holds for the police. This is the most complex, because – even when performance measured objectively is rising – the public’s standards and expectations may be rising even more rapidly. When this happens, the public – or certain segments of the public – remain continually dissatisfied as they “raise the bar” of police performance higher than the police are going. The police themselves may play a role in raising public expectations and standards, which ultimately affects the public’s assessment of their performance.
Issues for Future Research
A number of important issues are highlighted for future research. Addressing these issues and answering these questions will help police be more effective in improving public perceptions of them, their practices, and their accomplishments.
More research should be done to identify the best survey measures for specific purposes. Measurement issues include not only question wording, but also question ordering and how the survey itself is framed for the respondents.
Research should be done to determine the relative importance of a variety of influences on the public’s image of the police: the personal experiences of the public, what the public learns second-hand from friends and acquaintances, and what they learn from the mass media. How influential the citizen’s prior view and expectation when they experience a personal contact or exposure to a new mass media presentation?
What is the importance of highly publicized events on the police image – both negative events and positive ones? How long do these effects last, and how much influence do events in one community exert on the public’s image of police in another jurisdiction?
What kinds of police agencies achieve the most positive public perceptions?
What influence does context have on the public’s evaluation? Are patterns the same when crime is high as when it is low? When crime is rising as when it is stable or falling? Do citizens’ standards and priorities change as the context changes?
What influence do police public relations campaigns have on the public’s image of the police? Which kinds of efforts are most successful?
How closely correlated are objective measures of police performance (e.g., crime rates) to subjective measures (i.e., those based on public opinion surveys)? Where subjective and objective measures point to different results, what accounts for those differences?
What impact does the public’s image of police have for the public’s actions (e.g., crime reporting, cooperating with police, participation in partnerships, obeying the law, voting and supporting funding of police programs)? What are the consequences of shifts in the police image for the tenure of police leadership?
Agenda for Future Data Collection
A program for routinely collecting data on the police image is proposed. The program would be sponsored by the IACP and conducted with the assistance of an expert research organization. The working name of this program is the Uniform Public Opinion Poll on Policing (UPOPP). It would provide for a nation-wide system of survey research coordinated and facilitated by UPOPP, but conducted by individual IACP participating agencies. Participation in the program would be voluntary. The UPOPP program would provide for high quality survey research that produced a standardized set of survey questions to be used by all participants, plus an optional set of questions designed to suit any special interests of each participating agency. The survey would be conducted annually by participating departments to allow the tracking of public image trends over time.
Under the proposed UPOPP model, the IACP would:
Commission the development of a standard survey instrument and methodology.
Underwrite the field testing and revision of this instrument and methodology.
Solicit participation from IACP members
Select and monitor a research organization that would conduct a number of activities.
The responsibilities of the research organization would include the following:
Keep abreast of studies and survey research relevant to the UPOPP mission.
Assist participating members in developing survey items that would deal with local issues of concern not covered in the standard survey instrument.
Provide limited technical assistance to participating police agencies regarding the conduct of the survey (e.g., sampling, data collection, contracting with a local survey research firm to collect the data).
Receive and archive copies of participating agencies’ survey instruments and data, establishing and enforcing technical standards for computerized data.
Issue an annual report to the IACP that summarizes analysis from the data archive for the current year, as well as showing comparisons and trends for previous years.
Participating agencies would be responsible for collecting the data and delivering a copy to the UPOPP research organization. Participating agencies would either conduct the survey themselves or arrange for a local survey research firm to do this.
A number of issues would need to be resolved in detail, presumably choices to be made by the IACP – with the advice of the UPOPP research organization.
The content of the standardized portion of the survey instrument,
The content of the local-option, customized portion of the survey instrument,
Arrangements to make participation by small agencies economically feasible,
Scope of services provided by UPOPP research organization to participating agencies,
Content of the UPOPP annual report and whether results for individual participating departments would be reported,
Archiving requirements, access to data, and follow-up cross-jurisdiction data analysis, and
Funding of the UPOPP program.
CHAPTER 1
INTRODUCTION
The International Association of Chiefs of Police engaged the Administration of Justice Program at George Mason University to conduct a review of existing knowledge of the public image of the police. This report presents the findings from that study.
We conceptualize the police image as falling into three general categories: over all image, outcomes, and process (Mastrofski, 1998). The over all (or general) image of police is diffuse and reflects perceptions, feelings, and evaluations that ask about the police in general, without regard to any particular characteristic or criterion. The following are examples of over all image:
Confidence in the police
Satisfaction with the police
Trust in the police
Respect for the police
Support for the police
Police performance in general
The benefit of such general descriptions is that they reflect an over all orientation of the public to the police. They give us a general sense of how positive or negative the public is toward the police. They are limited, however, in that they provide no indication of what it is about the police that pleases or displeases them. They are merely summary measures that cannot by themselves reveal the individual assessments and weights of those assessments that contribute to each respondent’s view. The following two types of items enable researchers to examine the internal “architecture” of the public’s image of police.
Police are expected to achieve a variety of outcomes, some of which have long been characterized as part of the police mission, and others of which have been more recently embraced under the rubric of community policing: reducing crime and disorder, reducing the fear of crime, solving neighborhood problems and improving the quality of life, and developing greater community cohesion.
Police are also expected to adhere to a wide variety of process-oriented standards. These include the following:
Integrity – avoidance of corruption & abuse of power for personal gain, dishonesty, tolerance of corruption, abuse of power, and dishonesty among fellow officers,
Fairness – treating people equally or equitably, regardless of race, sex, origin, etc.,
Civility – treating people with respect,
Responsiveness – giving people what they want, showing care and concern for their problems,
Police presence – being available and accessible to provide police services in a timely fashion,
Appropriate use of force – using only that force necessary to accomplish legitimate goals, and
Competence – having the knowledge and skills necessary to do their work.
While it is safe to assume that people have opinions about the extent to which the police do these things well, there has been no systematic, comprehensive analysis of the views of the American public on all of these qualities. And we need to know the relative importance of these things to the public in shaping their over all assessment of police. When people express confidence or satisfaction with the police, how much do each of these elements exert on the over all evaluation?
It is important to know how stable the public’s image of police is over time and to know what influences fluctuations in their image. To what extent are positive/negative fluctuations a function of highly publicized events (e.g., the Rodney King incident, the O. J. Simpson trial, the Diallo case), and to what extent are they the product of social, cultural, economic, and crime trends?
This is not merely an academic question, but rather one with profound implications for the profession. If the public’s perception of the police nationwide is dramatically affected by highly publicized (even if highly localized) events, then a police executive needs to be prepared to provide a balanced picture to put the publicized event in perspective. If the event is atypical of policing, then the chief must be prepared to provide convincing evidence to the contrary (that does not appear to be merely self-serving). If it is enough of a problem to warrant widespread concern, then the chief must be in a position to show that he or she appreciates the scope of the problem and is taking action to rectify matters. Responding to such image problems requires nearly always a “reactive” approach.
If, on the other hand, the public’s view of police is affected by trends in crime, the economy, and other social indicators, then a more proactive, long-term approach may be effective. This might include alerting police leaders to the need to attend to specific image problems or opportunities that are anticipated with shifts in social, economic, or crime trends. For example, does the image of police as effective crime fighters decline when crime escalates? Are the police held responsible when drug crime and disorder increases? Does their image benefit when these problems are on the decline? When crime is on the increase, does the public place greater emphasis on the crime control function than it does when it is on the decline? When crime is declining, does the public place greater stress on police adherence to law and equal treatment across social groups?
It is also important to account for variations in the police image at any given time. It is well documented that racial and ethnic minorities tend to have a less positive image of the police than do whites. Elderly have a more positive view of police than youths. More highly educated and higher income citizens also tend to be more positive. However, we do not know much about the relative importance each group places on these qualities. For example, do older people place greater weight on the ability of police to control crime and disorder than younger people? Do younger people place greater emphasis on how the police treat citizens when dealing with them (rather than the outcomes of those police efforts)?
It would also be useful to understand how the public weights the relative importance of accomplishing desirable outcomes (e.g., lower crime rates) and police adherence to certain standards about how they go about their business (e.g., legality, courtesy, responsiveness, fairness). There is a fundamental tension in the police role that highlights the classic ends-means philosophical problem: when, if ever, do the ends justify the means? Egon Bittner’s (1970) insight about modern expectations about policing is that Western democratic culture places a high value on accomplishing lofty goals, such as peace (absence of crime and disorder), and that it prefers that whenever possible this be accomplished by peaceful means (such as negotiation, persuasion, and education). Nonetheless, society finds need to invest someone with the authority to use “dirty” – that is, coercive – methods for accomplishing peace when peaceful means may not work. The police are the institution that receives that authority, but giving them legal authority does not eradicate societal concern about the philosophical dilemma of deciding when the ends justifies the means – and when they do not. Americans place a high value on achieving peace and minimizing crime, but they also appear to be very concerned that police accomplish these things by not simply pursuing the most effective and efficient methods. They care also about how the police pursue these goals.
Two traditions in America make it difficult for police to pursue a single goal (e.g., crime reduction) without taking other, possibly conflicting goals into account (Goldstein, 1977; Miller, 1975). One is the tradition of distrust of concentrating too much power in any given public office, and finding ways to constrain or counterbalance powers that are given public officials, such as the police. The other is the tradition that police be public servants whose job is to provide their clientele things they want in the ways they want them. That is, police in America are not to be remote and isolated legal functionaries, but rather responsive and accessible servants of the public’s will: “of, by, and for the people,” to apply Lincoln’s well-known phrase.
What does this mean for the police image? Police are often targeted for reform because they are viewed as deficient in achieving the goals that are set before them, but they are also criticized and targeted for reform when their methods appear less than desirable. And when the police themselves seek support, their leaders often tend to emphasize their successes in accomplishing a particular outcome (e.g., reducing crime) or adhering to widely accepted standards (e.g., adhering strictly to legality). For example, the police leadership of New York City has in recent years sought to burnish the department’s image by claiming credit for reducing crime (Bratton, 1998). And the Federal Bureau of Investigation long ago established its credibility for incorruptibility by visibly promoting strict adherence to legal standards by its agents. An important question thus arises as we turn our attention to what shapes the image of the police that the public holds. What most burnishes the general police image and ultimately support for the police – positive feelings about police effectiveness in producing outcomes or positive feelings about how police go about their business?
At this point, no one has attempted to amass existing survey research on the public’s view of the police into a single, cohesive statement that will enable us to answer these questions. However, by conducting a comprehensive review of what is available from the existing research literature, we will be in a better position to state what is known and what remains to be determined. That is our goal in the remainder of this report.
Methodology
We began by conducting a thorough search of publicly available archives of national and major international surveys of the police image. Surveys of samples drawn on a state, county, or municipality were not considered unless they offered some valuable insights to broader questions about the police image. Several archives were used, including those housed at the University of Michigan, the Gallup, Roper, and Harris public opinion polls, and those conducted by news organizations. Where available, we obtained copies of the survey instruments (or those parts relevant to the police image) and basic characteristics of the sample (number of respondents, date conducted, how conducted). From this information, we prepared a catalog that will allow IACP to view the entire scope of existing survey data on the police image that are already available. Since this catalog would be lengthy in print, we have provided separately a database containing the information. It can be printed or it can be searched electronically. The catalog provides a useful tool for understanding previous research on the public image of the police. This resource should prove valuable in deciding on future surveys and survey items.
In addition, we also conducted a review of published research on the police image that is drawn from national (and in some cases local) public opinion surveys. This review draws primarily from academic and professional journals and research institute reports that are publicly available. Most of the existing literature focuses on the over all image of the police; surveys on the public’s image of specific aspects of police outcomes and police processes are far less frequent and systematic. The review summarizes the state of research and indicates the extent to which answers to questions raised in the initial research proposal are addressed by this research.
After reviewing and assessing all available data sets and published studies, we identify those aspects about the police image that remain unaddressed or about which there are inconclusive or conflicting findings. In the concluding chapter, we make a series of recommendations as to the kind of future research that might answer important questions about the image of policing in the United States.
Throughout the report we cite the specific sources from which we make our observations – both research reports and public opinion polls. We recognize that this may be distracting to readers desiring only to learn the results of the research, but we include the citations for two reasons: (1) some readers may wish to learn more about the specific research described, and (2) researchers whose work is cited deserve to have their work acknowledged.
CHAPTER 2
THE GENERAL IMAGE OF POLICE
I. Introduction
The public image of the police is measured a number of different ways. Sometimes surveys ask about “local” police, police in “your neighborhood” or police in “your area,” while other surveys ask about the police as a general institution. The terminology used to gauge public support also varies widely, with questions asking about whether respondents “approve of” or “trust” the police, have “confidence in” or “respect for” the police, or whether they “support” or have “favorable” views of the police. What makes these terms “general” is that the criteria or standards of performance remain unspecified. They do not ask the public to focus on either police processes or outcomes. The person answering this question could in good conscience choose both, neither, or perhaps something else entirely. And without additional information, we are unable to determine how much weight the survey respondent gives to specific aspects of police performance. Such questions are like those that ask the public to indicate whether they approve of the job that the president of the United States is doing – without specifying any particular aspect of that job. Such questions are useful, however, in that they give the survey respondent an opportunity to offer a summary that takes all of those aspects that are relevant to his or her view into account, weighting each, at least implicitly, as he or she prefers.
Not surprisingly, the terminology used in public opinion polls seems to make a difference in measuring the general image of the police. Another important element to consider in public opinion polls is whether citizens are voicing an opinion about their own previous experiences with the police, those of their neighbors, friends or family members, or simply general impressions based on a number of sources, from television and the media to opinions shared within the subcultures in which they are immersed. With all these questions in mind, it is difficult to come to terms with what constitutes the “general image” of police.
Why is the “general image” of police worth measuring? There are a number of important reasons. First, an understanding of the general image of the police among citizens provides an important indicator of support for the institution among its constituents. Understanding how the public views the police is a crucial first step in improving relationships between the police and communities. This is why community surveys are a prominent component of the community policing movement. Similarly, measurements of the public image of the police can be compared. By producing such measures, agencies can learn whether their image is improving or declining over time, or whether they are held in higher or lower esteem by their citizens than police in other communities.
Second, the general image of the police may affect the sorts of behaviors by the public that greatly interest the police. These include supporting tax initiatives or referenda designed to enhance the resources of local police agencies, to participate in co-production activities like neighborhood watch, providing the police with information useful to solving crime or improving the quality of life in neighborhoods. Communities with a poor image of the police will be less likely to support and help the police do their jobs, and more likely to file complaints, launch civil suits, rebel against the police, and produce media problems. Whether there is indeed a strong relationship between these public behaviors and the overall image of the police is an untested, but certainly plausible, thesis.
Finally, there is a small but growing body of evidence that those who view the authority exercised against them as illegitimate are more likely to rebel against authority, or in the case of the police, violate the law. For instance, research has shown that while arrest deters spouse assault among some offenders, it leads others to become even more angry and defiant, which actually increases their recidivism rates. Other research has found that domestic violence arrestees who thought they were treated fairly by police were least likely to reoffend (Paternoster, et al., 1997). While much research remains to be done on the link between the perceived legitimacy of the police and crime rates, there is some evidence to suggest that as institutions like the police lose legitimacy, an increase in crime and rebellion against the police and other legal and political institutions might result (LaFree, 1998; Tyler, 1990).
II. General versus Specific Measures
For quite some time police researchers have noted that different survey items regarding police image capture different levels of public support. For example, Reiss (1967:36) noted that citizens seem to be in a “double bind.” Citizens frequently express skepticism about police power, yet they view police power as a solution. Similarly, citizens respect the police function, but they lack trust in some of the ways police perform their duties. Finally, although they remain sympathetic the challenge of police work, citizens are hesitant to allow police discretion (Reiss, 1967).
White and Menke (1982) argue that the inconsistencies in the police image revealed in public opinion surveys result from question formatting. The proportions of citizens who reported a positive police image when presented general questions ranged from 75% to 80%. However, positive responses only ranged from about one-sixth to one-half when citizens were presented survey items that were more specific in nature.[1] White and Menke (1982:223) concluded that “general and specific items assess different universes of meaning and are not simply artifacts of meaningless comparisons of these measures.”
In sum, the available evidence suggests that survey items tapping into general evaluations of the police image will yield more favorable results. However, the evidence suggests also that a complete picture of how the public perceives the police can only be pieced together by administering both general and specific survey items to respondents. Specific items are useful in identifying the particular aspects of police performance that are least attractive to the public, which will enable police organizations to target those areas for improvement.
III. How to Measure General Police Image
In a recent review of several national polls, Shaw and his associates (1998) found that a majority of Americans hold a favorable image of the police in general, as well as the police in their communities. However, their study shows that different questions generate different levels of support for the police.
One common question used to measure general police image among citizens asks “how much confidence you, yourself, have” in the police as an institution in American society. Results from eight Gallup polls conducted annually from 1993 to 2000 show that, on average, slightly more than 56% of respondents reported either “a great deal” or “quite a lot” of confidence in the police in general (see Exhibit 1). The percentage of those responding in positive terms ranged only slightly over time, from 52% in 1993 to 60% in 1996. The results from polls asking about citizens’ levels of confidence in the police in their community were nearly identical. The percentage of citizens with a positive police image, averaged across five national polls, was 58%. Differences in levels of confidence across the five polls, which were conducted between 1981 and 1996, appear rather marginal, ranging from 55% in 1981 and 1991, to 60% in 1985, 1995 and 1996 (see Exhibit 1). The message from these results is that a majority of Americans have confidence in the police as an institution and the police in their community, but a substantial portion are also less confident. This requires careful consideration.
It would be wrong to conclude from the polls noted above that a sizable portion of citizens have little or no confidence in the police. For example, results from the eight Gallup surveys conducted between 1993 and 2000 show that only between 10% and 12% of respondents said that they had “very little” confidence or “none at all.” A significantly higher percentage, ranging from 29% in 1996 to 35% in 1993, reported that they had “some” confidence in the police. Again, the results from the five polls asking about levels of confidence of “police in your community” are very similar. The percentage of citizens who reported that they had “very little” confidence or “none” were a small minority, ranging from 12% in 1996 to 19% in 1981. The percentage of citizens who said that they had “some” confidence in the police in their community was significantly higher (range = 24% [1981] to 31% [1991]). Overall, then, only a small minority of citizens who were polled reported that they had very little or no confidence in the police as an institution, and in the police in their community.
The key issue is what to make of the response, “some confidence.” Does that reflect positively or negatively on police? The answer, of course, depends upon one’s expectations. Perhaps one way to think about it is to consider whether a chief would be pleased to announce to the press that about one fourth of the citizen’s of his jurisdiction reported that they had “some” confidence in their police. We suspect that most chiefs – especially those who embrace community policing – aspire to higher levels of public confidence, so it seems likely that most would view an assessment of “some” confidence as a signal that improvement is possible and even, perhaps, needed.
Different questions have generated varying levels of public support for the police. Another common survey question asks citizens about their “overall opinion” of their “local police department.” The results reported by Shaw and his associates show that an overwhelming majority of respondents rate their local police favorably (see Exhibit 1). For example, a Gallup poll conducted in March 1991 revealed that 82% of respondents rated their local police department either “very favorably” or “mostly favorably.” A more recent poll, which was conducted by Princeton Survey Research Associates in 1997, also showed that an overwhelming majority of citizens (81%) rated their local police in favorable terms (Shaw et al., 1998:414-415). When comparing these findings to those reported above, it appears as though survey items asking citizens about how “favorably” they rate the police in their community generate more positive results when compared to items that inquire about levels of “confidence.”
We can only speculate as to the reason for this difference. Perhaps questions asking respondents to make a “favorableness” assessment are tied more closely to the respondents’ personal, subjective expectations, whatever they may be. The question would seem to allow respondents to incorporate a broader range of police characteristics: intentions, effort, and outcome. In contrast, a question about “confidence” in police would seem to encourage respondents to focus more on objective, observable results, perhaps downplaying intention and effort. In the same way that one may have a favorable impression of a doctor with a good bedside manner, one could still be less positive about the confidence placed in him or her in curing a difficult disease.
IV. General Police Image Over Time
It is difficult to draw sweeping conclusions about changes in the general image of police over time because there are few data sets that collect comparable information over time. There are several reasons why these “longitudinal” data sets are rare. The primary reason is that designing survey questions is a demanding, technical task – one that requires repeated testing among respondents to ensure that they understand the question, that it is not confusing, that the response options make sense, and that it measures the phenomenon of interest. Survey questions, or the response options that go along with them, frequently change format based on feedback from survey respondents, the whims of the researcher, a lack of awareness of prior research, or any one of several other possible explanations. Readers need to use caution when the format of a question changes even slightly, since any deviation from the trend may be due to the change in the question format rather than a change in the quantity being measured (such as the public image of the police).
To illustrate this point, consider a set of polls conducted in the early 1980s using the same questions on equivalent samples of respondents, but with a slight difference in the response options. In 1981, CBS News and the New York Times polled a national random sample of adults about the degree of confidence they had in the police in their community. The response options were: A Great Deal, Quite a Lot, Some, Very Little, and No Opinion. Fifty-five percent of respondents reported that they had a great deal or quite a lot of confidence. When the survey was repeated about 8 months later in 1982 by the Gallup Corporation, the response options changed: A Great Deal, Quite a Lot, Not Very Much, None at All, and No Opinion. Now 76% of respondents reported having a great deal or quite a lot of confidence. A 1984 survey having the same format as the 1982 survey found similar results, with 75% of respondents reporting a great deal or quite a lot of confidence in police. Then in 1991 the original question format was adopted again, and once again in this poll, 55% of respondents had a great deal or quite a lot of confidence in the police. It is unlikely that there was a sharp increase in public confidence in the police from 1981 to 1982, which then leveled off until 1984, followed by a gradual reduction through 1991. The more plausible interpretation is that a small adjustment in the response options provided to the survey respondents made an enormous difference in the findings. In this case, “some” was replaced with “not very much” and “very little” was replaced with “none at all.” Terminology matters.
Despite the problems with compiling a longitudinal data series, we do have some limited information on trends in the general image of the police. The data series are imperfect; they are missing data for some years, while in other instances polls were conducted twice in the same year. As we have already shown in Exhibit 1, public confidence in the police from 1993 to 2000 has not experienced dramatic changes. An average of about 56% of respondents have a great deal or quite a lot of confidence in the police. In 1993, this figure was 52%, rising to 60% in 1996, and declining back to 54% in 2000.
There are a number of plausible explanations for these changes, such as highly publicized crises and crime trends. These influences have not been explored with rigor, but we can offer an illustration to show how one type of explanation might be explored in future research. Exhibit 2 shows the relationship between public confidence in the police and crime rates for the 1993-2000 period depicted earlier in Exhibit 1. In this chart, crime rates and public confidence rates for 1994-2000 are expressed in percentage increments above or below their 1993 levels. If declining crime rates promote higher levels of public confidence, then as the yellow and green lines (crime rates) decline, the red line (public confidence) should rise. The chart fails to show the predicted relationship across the entire time period. Even as crime declined at a fairly steady rate during these years, public confidence either remained the same or declined for five of the seven time periods. Public confidence rose only between 1994 and 1996, early in the process. What could account for this pattern? It is possible that the public adjusts its expectations over time, requiring increasingly greater levels of performance to express confidence in police. This is only speculative, however, based on a very limited illustration. We would prefer to have a much longer time period to compare these trends. Also, it is more plausible that if citizens hold police accountable for the crime rate, this relationship would be more readily discernable if we were comparing confidence in the public’s assessment of its own police department to the crime rate in that jurisdiction. Perhaps too the relationship would be clearer if we had data over this time period for the public’s assessment of the police ability to reduce crime (rather than a general image question).
It is important to note that although the general image of police is fairly good and fluctuations from year-to-year tend to be quite small, it has been declining steadily since 1996 when measured in terms of public confidence. This is a particularly noteworthy pattern when one considers the enormous investment that the police profession and taxpayers have made in community policing reforms. During this period, billions of federal dollars have been spent to promote community policing, and according to surveys of police leaders, nearly all support it. The press on community policing has been almost entirely positive (Mastrofski and Ritti, 1999). Under this onslaught of good feelings about community policing, it is remarkable that general attitudes about police have changed so little and in fact have declined over the last five years.
Another indicator of how the general image of the police is changing over time comes from a series of public opinion polls on the public’s respect toward the police. Exhibit 3 shows that in the 1960s, a period of turbulence for the American police, two public opinion polls found that an average of 74% of respondents had a great deal of respect for the police. When the polls were repeated four more times in the 1990s, the average number of respondents with a great deal of respect for the police had dropped to just under 59%. The surveys were not conducted in equal time intervals, and they skipped more than two decades, so Exhibit 3 may be masking a much more complicated story. Nonetheless, the decline in respect for police from the 1960s to the 1990s is still quite striking. It tracks fairly closely with another public opinion trend during that period, which is a decline in the percent of Americans who trust their government to do what is right (LaFree, 1998:102).[2]
The most complete longitudinal series on the general image of the police results from the yearly Monitoring the Future surveys conducted by the University of Michigan (Pastore and Maguire, 1999). Since 1987, a nationally representative sample of at least 2,300 high school seniors is asked to report “how good or bad a job is being done for the country as a whole by the police and other law enforcement agencies?” Exhibit 4 shows that the perceived performance of the police declined from 1987 to 1992, fluctuated erratically through 1996, and then began to increase again through 1999. On average, only about 31.4% of seniors surveyed during this period view the police as doing a good or very good job. The remainder, nearly 70% on average, view the police as doing a fair, poor, or very poor job. The difference between the general image of the police among random samples of high school seniors and adults is pronounced. Age is one of many variables thought to influence the public image of the police. It is to these variables that we now turn.
We conclude this section with another methodological caveat. All of the survey questions we have considered (and will consider) tend to “force” or “channel” respondents to offer an opinion, when they may have no opinion or one that was so weak as to manifest itself only because the issue was raised by the survey researcher. What this means is that potentially many respondents who had not heretofore given the question (e.g., confidence in the police) much thought are now placed in a psychological state by virtue of being questioned that they feel pressure to offer an opinion. One of the ways survey researchers have developed to relieve that artificial pressure to offer an opinion that is weak or nonexistent is to replace “no opinion” with “...or haven’t you thought much about this recently?” The latter provides those with very weakly held views to select a “face-saving” option, and it more accurately portrays the state of the public’s mind about the issue. Another option is to preface all questions about the police with a general question about how much the respondent has recently thought about the performance of the police. This allows researchers to distinguish views based on how important the topic has been to the respondent. If researchers are attempting to predict what citizens will do as a consequence of their opinions about police (e.g., how they will vote, whether they will participate in police programs, whether they will obey the law), knowing how important this topic is to each citizen would be a valuable piece of information.
V. Factors Influencing the General Image of the Police
This section considers some of the factors shaping the general image of the police. It is incomplete by design, since some of the factors thought to influence the general image are the specific components of the police image which we have not covered yet. One of the most compelling arguments about the general image of the police is that it is shaped by the outcomes the police produce (such as crime control) and the processes they use to produce those outcomes (including fairness and other aspects of the policing process). Since these will be covered in detail in the remainder of this report, for now we focus on three kinds of influences on the general image of the police: the personal characteristics of the citizen who is asked to make the evaluation,[3] the nature of the contact that the citizen has recently had with the police, and mass media portrayals of the police and crime.
Personal Characteristics of the Citizen
Research on the factors influencing the public image of the police typically draws on the “usual suspects”: age, race, sex, income and socio-economic status, victimization history (which will be explored later), and other individual level factors thought to influence attitudes more generally. Since in matters of policing, race is a crucial variable, we examine it apart from the others shortly.
Race. One of the most persistent findings in public opinion polls about the police is that whites are more satisfied with police than nonwhites. This finding has been consistent over the past four decades, emerging from dozens of studies and polls, both in the United States and abroad (Bayley and Mendelsohn, 1969; Bradley, 1998; Cao, Frank, and Cullen, 1996; Huang and Vaughn, 1996). For instance, in a study of citizen satisfaction with police in 12 cities, conducted by the Bureau of Justice Statistics in 1998, 90% of whites were satisfied with police, compared with 76% of blacks and 78% of those of other races (Bureau of Justice Statistics, 1999). These aggregate racial differences held in 10 of the 12 cities. In Madison, Wisconsin, an equal number of blacks and whites (97%) were satisfied with police, while in Tucson, Arizona more blacks (91%) were satisfied than whites (88%).
Age. Most of the research shows a positive relationship between age and attitudes toward police. Younger people routinely report less satisfaction with the police than older people (Brown and Coulter, 1983; Hindelang, 1974; Jesilow, Meyer, and Namazzi, 1995; Huang and Vaughn, 1996; Smith and Hawkins, 1973). The one study not reporting such an effect was based on a sample of juveniles, suggesting that age may matter when comparing juveniles to adults, but among juveniles, age may not matter as much (Hurst and Frank, 2000). Since this conclusion is based on one study in a single city, it should be viewed with caution. Another study reported that elderly respondents held less favorable attitudes toward the police than younger adults (Huang and Vaughn, 1996; Zevitz and Rettammel, 1990). These last two research findings suggest that the relationship between age and attitudes toward the police may be curvilinear. In practical terms, this means that juveniles are less satisfied with the police than adults, but that among juveniles, age does not matter. Then, as people age, their satisfaction with police continues to increase, until a certain age level, beyond which attitudes toward the police begin to decrease again. This is mere speculation on our part, since the research on the effects of age on satisfaction with police is not sufficiently developed to warrant firm conclusions.
Gender. The relationship between gender and satisfaction with police is unclear. At least two studies have found that males hold more positive views than females (Brown and Coulter, 1983; Thomas and Hyman, 1977). Other studies have found that females hold more positive views than males (see Huang and Vaughn, 1996, p. 35). Still another study has found that gender had no effect (Hurst and Frank, 2000). We are not sure why the effects of gender are so erratic across different studies.
Socio-economic Status. Poorer people, and those from lower socio-economic classes tend to report less satisfaction with police than those who are wealthier. For instance, Benson (1981) found that respondents from lower social classes were less satisfied with police. Similarly, Brown and Coulter (1983) found that income and education both had a positive effect on satisfaction with treatment by the police (a variable that can be viewed as both an indicator of the general image of the police and as an indicator of the image associated specifically with police process). However, both Hindelang (1974) and Jesilow, Meyer, and Namazzi (1995) report that education had no effect on attitudes toward police. Decker (1981) notes an important concern about the role of socio-economic status (SES). As we will discuss shortly with regard to race, it is not clear whether it is the individual’s socio-economic status that influences attitudes toward police, whether it is the status of the neighborhood, or whether these two variables interact. As we will demonstrate shortly, if SES works in the same fashion as race, neighborhood effects may be more important than individual attributes like SES.
Other influences. Race, age, gender, and SES are the individual variables most commonly considered in research on citizen satisfaction with police. Nonetheless, there are other scattered research findings that may be important to consider. For instance, several researchers have found that people living in the suburbs have better attitudes toward the police than people living in urban areas (Hindelang, 1974; Hurst and Frank, 2000; Thomas and Hyman, 1977). Another study confirms what might be viewed as common sense, that juveniles with a commitment to delinquent norms are less satisfied with police (Leiber, Nalla, and Farnworth, 1998).
Social scientists have not confirmed what these kinds of differences mean, but there are two theoretical approaches worth considering. One is that people with different characteristics have different experiences and that their opinions about the police are grounded in the objective reality of those experiences. If youths are more likely to be stopped, searched, cited, arrested, and warned than elderly people, then their negative views of police are perfectly understandable as an outgrowth of the different experiences of these two groups. The other theoretical perspective is that people with different backgrounds have different expectations or standards for police – and different ways to interpret events. If a person brings a negative preconception of the police to an experience, then they may be more inclined to focus on police actions that are consistent with that viewpoint and ignore those which are not, or they may simply interpret a given police action in a way that is consistent with that viewpoint.
Neither of the above theories has been thoroughly tested. However, a growing body of evidence suggests that racial differences in attitudes toward the police may not be a simple function of individual race, but that they are also influenced by broader social structural issues like (1) subcultural attitudes toward the police that are independent of individual experiences, and (2) the characteristics of the neighborhoods where respondents live. Decker (1981) found that community level predictors of individual attitudes toward the police included neighborhood culture and community beliefs about the police. Apple and O’Brien (1983:83) found that "an increase in the number of blacks in the neighborhood increases the opportunity for blacks to associate with others who have negative attitudes toward the police, and this results in an overall increase in their negative sentiment toward the police." Jesilow, Meyer, and Namazzi (1995) found that not liking things about one 's neighborhood was associated with negative attitudes about the police. Leiber, Nalla, and Farnworth (1998:169) concluded that “the imposition of legal authority and social control in certain neighborhoods engenders a pervasive resentment and resistance, and that youthful residents of those neighborhoods harbor a general disrespect for the law itself."
Neighborhood Effects. The most striking (and convincing) evidence for neighborhood effects comes from a massive study called the Project on Human Development in Chicago Neighborhoods. Sampson and Bartusch (1999) found that:
Once neighborhood economic disadvantage is taken into account, blacks ' views are found to be similar to whites '. Blacks appear to be more cynical toward or dissatisfied with the police only because they are more likely to live where disadvantage is concentrated. Even in neighborhoods where the rate of violent crime is high, there is no difference between the races in attitudes toward the police. Racial differences disappear when neighborhood context is considered. Thus, residents ' estrangement from the police is better explained by neighborhood context than race.
Although this is one of the largest and most carefully constructed studies of attitudes toward the police, it is still important to keep in mind that it is based on only one city. One of the enduring lessons of social science is that research evidence – even good research evidence – needs to be replicated over place and time before it can be generalized.
Sampson and Bartusch’s (1999) findings on the relationship between attitudes toward police and the legitimacy of the law also highlight a theme pointed out by researchers in the past. The public’s image of the police is often part of a larger attitudinal complex toward social, legal, and political institutions (LaFree, 1998). For instance, Benson (1981) found that political alienation influences ratings of the police, but the effect varies across social class and perceived integrity of the police. Thus, those who are poor and/or nonwhite may not only express unfavorable opinions toward the police, but may feel alienated from the political process more generally. Brown and Coulter (1983) found that citizens who rate the quality of local government higher tend to be more satisfied with the police. Albrecht and Green (1977) found that attitudes toward the police are strongly related to attitudes toward attorneys, judges, courts, and the legal system. Attitudes toward the police are also related to attitudes toward the larger political system, though the relationship is not as strong. Finally, attitudes toward the police are also related to degree of involvement in the political system, though this relationship is the weakest of those considered. Albrecht and Green (1977: 81) conclude that
The same respondents who express negativism toward their local police also feel generally alienated from the legal and political process and more cynical about the effectiveness of the operation of that process, especially in terms of its relationship to them and others like them....The implications are both clear and important. Programs designed to change public attitudes toward the police are not going to be generally successful unless they consider the broader, fundamental value system of which these attitudes are a part.
With this notion in mind, we will later (in section VI) examine the image of the police compared with other major social institutions.
Nature of the Citizen’s Recent Contact with Police
One way to think about the influences on the public’s general image of the police is to consider the different ways in which members of the public might acquire their impressions. People may acquire their images from direct personal experience – their contacts with the police. They may also acquire them indirectly through people with whom they associate – family, friends, coworkers, and acquaintances. And people may acquire their images of the police through the mass media – news, entertainment, and educational. A fair amount of survey research has focused on the impact of people’s direct experiences with the police on their general attitudes toward the police (Dean, 1980; Gibson, 1989; Koenig, 1980; Roberts and Stalans, 1997:149-152; Scaglion and Condon, Tyler, 1990). The general thrust of this research is that how citizens experience the police personally shows a significant impact on their general assessment of the police. Positive experiences are associated with a positive image and negative experiences with a negative image. Negative experiences appear to have a more powerful effect than positive experiences. The extent of the difference in impact between positive and negative contacts varies with the type of measure of general support used by the researcher and the composition of the survey sample.[4]
One recent study provides a detailed analysis of different aspects of the citizen’s contact experience with police and its impact on the citizen’s general level of satisfaction with the police department. Reisig and Chandek (2001) surveyed a sample of citizens who had recently had contact with police in a Midwestern city. They considered a number of possible influences on the citizen’s evaluation, including the level of courtesy/friendliness of the officer, the citizen’s age, sex, and race. The researchers found that the strongest predictor of the citizen’s satisfaction with the police department in general was how courteous/friendly the officer was with the citizen. This held for both traffic stops, where the citizen’s contact was involuntary and breaking-and-entering encounters, where the contact was at the citizen’s request. Interestingly, minority citizens tended to rate police significantly lower than white citizens in traffic stops, but the effects of the officer’s demeanor toward the citizen was over three times more powerful as a predictor of the citizen’s over-all evaluation of the police department generally. The citizen’s race was not a significant predictor of the over-all evaluation of police respondents gave in breaking-and-entering contacts. The researchers expected that the citizen’s expectations about what services the police would provide would exert an influence on their general evaluations of the police.[5] They hypothesized that if the police performed more service than expected, citizens would form a more positive general impression of the police, and if they gave less than expected, they would form a more negative overall impression. They found that although this was true for the evaluations that citizens gave for the specific police encounter, that effect did not transfer to their general impressions of the department.
The above studies examined the impact of the public’s specific contacts with police on their general views of police, but it is also possible, and indeed likely, that people’s general views of police influence the way they evaluate a specific experience. Two studies have examined this question, and both have found that prior general (sometimes called “global”) views of police have a stronger influence on the public’s evaluations of a specific contact with police than their evaluation of a specific contact has on the subsequent general evaluation of police (Brandl et al., 1994; Tyler, 1990). One research team concluded,
These findings are consistent with the proposition that citizens’ evaluations of their personal experiences with the police are affected by stereotyping and selective perception; those who hold generally favorable views of the police are more likely to evaluate their contacts with the police favorably, and those who hold generally unfavorable views are more likely to evaluate their contacts unfavorably (Brandl et al., 1994:131).
As valuable as police contact studies are, they overlook one important fact. The vast majority of Americans rarely have direct contact with the police, which means that these people will be drawing heavily on other sources to form their impressions of the police. A large 1999 Bureau of Justice Statistics survey shows that 79 percent of the respondents reported no face-to-face contact with the police in the previous year, and only 4 percent reported more than one contact (see Exhibit 5) (Langan et al., 2001). And the vast majority of the contacts people did have were routine traffic stops and requests for assistance. Relatively few involved the citizen as a suspect or victim in a serious crime or other emergency. So, the vast majority of the American public has not had recent contact with the police and of those few who have had recent contact, their experiences were not the sort of dramatic situations in which their safety was immediately threatened or their freedom and reputations were at stake. These people’s general impressions of the police when interviewed were thus probably heavily influenced by (a) memories of their own experiences with police in the more distant past, and (b) impressions given directly by individuals with whom they have frequent contact, and (c) impressions given by the mass media.
We did not find survey research pertaining to category b above, but a few important implications can be drawn from common sense. First, a police-citizen encounter can have an impact that goes well beyond the immediate public participants if those participants share their accounts and views of those experiences with their friends, neighbors, and acquaintances. To take an extreme case, if the minister of a church is treated rudely by a police officer, there is tremendous potential to magnify the effects of that single event many-fold because of the minister’s access to large numbers of the public and his or her status among them. As community policing continues to grow in popularity, many police agencies are in the business of strengthening the potential influence of this kind of secondary impression of the police. Under community policing, neighborhood and community groups are encouraged to pay attention to police policies and practices, to provide guidance as to community preferences, and even to participate in some police programs. This approach is designed to create more public forums in which members of the public share their past experiences and impressions of the police (Skogan and Hartnett, 1997). Where this occurs, results – both positive and negative – will be magnified by these secondary sources of public impression.
Mass Media Portrayals of Police and Crime
We found a small body of research on the impressions of police that are formed from mass media presentations. At the outset we stress that the number of such studies is small and the number of unanswered questions is large, so conclusions about mass media influences on the public’s image of police are necessarily quite tentative. We divide our discussion into two parts: news media and entertainment media.
Before discussing evidence on each, we briefly describe three theoretical approaches for explaining mass media effects on public attitudes about institutions such as the police (Fox and Van Sickel, 2001:6-8). The “hypodermic needle” theory assumes that the public takes in media presentations like a drug, which produces powerful and long-term effects on their views of institutions such as the police. Members of the public are viewed as independent consumers of these media presentations, which they use to answer questions about the police and from which they formulate attitudes and perceptions of the police. The “limited effects” theory also assumes that the public uses the media for information, but it argues that individuals evaluate that information in the context of what they know from other sources – such as direct contact, family, friends, etc. These pre-existing and more-or-less independent impressions are believed to constitute powerful influences with which media images must contend in the competition for influencing the public’s views of the police. Under these circumstances, the effects of the mass media are expected to be present, but limited. The “subtle/minimal effects” theory falls in between the “hypodermic needle” and “subtle/minimal effects” theories. Here, the hypothesized media effects are neither overwhelming nor minimal, but rather work in special ways by: (a) agenda setting – instructing the public in what to think about as the most important issues (e.g., whether policing is an important issue at a given time and what aspects are important), (b) priming – associating people or institutions with particular issues (e.g., associating the police with crime fighting), and (c) framing – shaping how to think about a given issue by either identifying general trends or covering specific events (e.g., how often the police use excessive force in dealing with suspects). Thus, all three theories posit that the mass media influence the public’s views, although in different ways and to different degrees. Evidence relevant to all three approaches can be found in research, but evidence is mixed for all three models.
News Media Influences. The prominence of crime in the news and the importance of the police as a source of news about crime inevitably focuses public attention on the role of the police as crime fighters. Indeed, surveys of the public indicate that up to 95 percent of the public consider the mass media as their main source of information about crime (Surette, 1998:197). Most news stories about police are focused on a specific crime, crime trends, or crime problems, and the police are rarely the focus of the report. Thus, a lot of news places police in the background of the story, mentioning them only inasmuch as they describe what they are doing about a crime event or larger crime issue (Surette, 1998:69-70). These news stories rarely provide a larger historical, sociological, or political context or interpretation that would place the story’s topic in a broader perspective. Thus, for example, most stories about the police shooting of Diallo in New York failed to provide a broader discussion of long-term trends and cross-city comparisons in the police use of lethal force (which was declining at the time in New York and which placed New York below many other large American cities in the rate of police shootings). Consumers of the news are thus left to form their own impressions, and according to one researcher, the implicit message stresses the inability to catch offenders – but police are portrayed as doing at least a fair job, while courts and corrections do poorly (Graber, 1980:74-83). If the agenda-setting properties of news media influence are valid under the “framing” theory, then one would expect that the public generally would evaluate police primarily according to their ability to fight crime. However, as we will later show, a number of studies suggest that – at least in the 1990s – the public appears to give the processes of policing (e.g., how police treat citizens they encounter) much greater priority than crime control in calculating their overall evaluation of the police.
Given the luke-warm image of police as crime fighters provided by the news media, it is remarkable that support for the police is as high as it is, leading one scholar to speculate that although the media message is that the system does not work well, it is still presented as “the best hope against crime” (Surette, 1998:226). By way of offering a speculative explanation, we suggest that a careful reading of news stories about the police would show that although the over all media image of police may reflect poorly on police ability to achieve crime control, these stories – either explicitly or implicitly – present the police as well-motivated and trying to “do the right thing.” It may be the inferences the public makes about police objectives and motivations that produce the positive general assessments about police that routinely appear in national surveys.
Recently some researchers have identified a trend in news media coverage of the criminal justice process that they characterize as “tabloid justice” (Fox and Van Sickel, 2001). They argue that mass media have entered a time when they concentrate on the “sensationalistic, personal, lurid, and tawdry details of unusual and high-profile trials and investigations” (p. 3). This is characterized by three trends observed in both mainstream and emerging news media formats: (a) treating news as entertainment, (b) a proliferation or “frenzy” of media coverage of specific cases, and (c) an increasingly attentive public that uses this information to understand and evaluate the criminal justice system (that is, the hypodermic needle model of mass media influence). Although much of the empirical research focuses on the legal system generally and especially on cases at the stage of trial, the researchers do present some results relevant to police.
First the researchers tested the effects of “priming” respondents to a national survey to certain “tabloid” cases of the 1990s by asking them questions about their familiarity with those cases.[6] The researchers assumed that virtually all of the respondents’ knowledge of these cases would come from the news media. The researchers found that respondents who were first questioned about these cases revealed in a subsequent question that they expressed lower levels of confidence in the police than those who were asked about tabloid questions after being asked about their confidence in the police (Fox and Van Sickel, 2001:132). Next the researchers directly asked the survey respondents about whether these cases had influenced their confidence in the police. In four of the five tabloid cases analyzed in this part of the study, the pattern of responses indicated that there was a net reduction in confidence. The O. J. Simpson case was the most extreme, 62 percent reporting a loss of confidence, 5 percent reporting an increase and 30 percent reporting no change. The over all impact of all of the tabloid cases discussed with survey respondents showed a net decline in confidence in the police of 23 percent attributed to knowledge of tabloid cases.
The researchers concluded that exposure to these cases (presumably as a result of news coverage) reduced the legitimacy of the police, and they attribute the negative coverage of tabloid-style journalism. Actually, it seems more appropriate to say that when people were reminded of these cases and asked to consider their impact, those very people tended to believe that it reduced their confidence in police. It is quite possible that this research overstates the scope of this effect – for two reasons. First, the survey mentioned only tabloid stories about police to the respondents and did not mention any other kinds of media coverage that might have predisposed the respondents to feel more positive about the police. That is, the survey itself probably does not accurately replicate how people actually consume and process news stories, which presumably would include a wider variety of stories. Second, the respondents themselves may not have an accurate view of how these stories actually affected their own confidence in the police. Simply by asking the question, the researchers may have created in the respondents’ minds an inclination to make a judgment one way or the other, when in fact the effect was small, nonexistent, or even contrary to what they believe about themselves. In the final analysis, it is not unreasonable to suppose that people exposed to negative mass media images will have lower evaluations of the police, but it remains to be demonstrated just how extensive this exposure is, and how influential media images are compared to other sources of information.
Entertainment Media Influences. We consider entertainment media to include accounts of police work through various communications media (print, broadcast, recording) that are advertised as fiction or explicit entertainment (e.g., music, video games). Also included are books, film, and television shows that claim to present the reality of police work in an entertainment context. These are sometimes called “reality TV” or “info-tainment.” We have not found an empirical literature on the effects of entertainment media on public attitudes toward the police, although there is a larger literature on the consumption of entertainment media (especially television) on such things as fear of crime. The general conclusion is that television consumption is associated with greater fear of crime and violence and greater cynicism and distrust in social attitudes (Surette, 1998:212). The effects are not uniform, however, and are especially related to the credibility of the information source (suggesting the importance of examining the effects of the many “reality” police shows now available). Print media have a greater influence on people’s knowledge about crime and their adoption of crime prevention actions, likely due to differences in content and style of presentation associated with each type of medium and perhaps also differences in who tends to consume each type of medium.
Although we do not present findings on the relationship of entertainment media consumption on police specifically, we are able to report some general findings on the nature of entertainment media content. Studies of entertainment media content repeatedly demonstrate the obvious: the entertainment media present an extremely distorted view of the nature of police work, one that stresses crime fighting, police violence, and individualism (as exemplified by Dirty Harry) (Surette, 1998: 40-43). The entertainment media present police as effective in solving, but not preventing, crime and as doing incident driven, not community- and problem-oriented work. Effective law enforcement officers in the entertainment media are those who eschew routine methods, often violating rules and laws, and take exceptional measures to solve cases (involving weapons and highly sophisticated technology). Private investigators and amateur private citizens prove far more successful in solving television crimes than the police. Television police are more positively presented than television attorneys, judges, and corrections officers, but for every heroic officer portrayal, two others are incompetent and two others break the law (Lichter and Lichter, 1983). Because of the focus on the exceptional and spectacular, one crime and media expert concluded,
Whatever the [entertainment] media show is the opposite of what is true. In every subject category – crimes, criminals, crime fighters, the investigation of crimes, arrests, the processing and disposition of cases – the entertainment media present a world of crime and justice that is not found in reality. Whatever the truth about crime and violence and the criminal justice system in America, the entertainment media seem determined to project the opposite. Their wildly inaccurate and inevitably fragmentary images provide a distorted reflection of crime within society and an equally distorted reflection of the criminal justice system’s response to crime. The lack of realistic information further mystifies the criminal justice system, exacerbating the public’s lack of understanding of it while constructing a perverse topsy-turvy reality of it (Surette, 1998:47).
Police researchers have devoted little attention to the second-hand accounts that may influence the police image. For example, it would be valuable to know just how much and in what ways second-hand accounts of policing affect the general impressions of citizens who have infrequent contact with the police as opposed to those who have frequent contact. A working hypothesis is that citizens who have frequent direct contact – or who have friends and acquaintances who report frequent direct contact – will be much less affected by both positive and negative entertainment media portrayals than those who experience such direct or second-hand contact infrequently. The underlying assumption of this hypothesis is that different sources of information compete with each other for influence. An alternative hypothesis is that consumers of the media selectively perceive events they observe in both the entertainment media and their personal experiences. Under these conditions, the hypothesis would be that citizens would focus on information in both the media and their experiences that tend to reinforce already existing preconceptions.
Of course, entertainment media may play an important role in framing citizens’ impressions and expectations about their police. Movies, television shows, and novels that present police in a positive light can be expected to predispose citizens more positively toward police generally – especially those with little direct personal contact with police. The proliferation of “reality TV” programs that show video clips of police-citizen interactions may play an especially important role in shaping the public’s view of police in recent years (Surette, 1998). However, it is not clear whether these “reality TV” portrayals tend to produce more positive or negative images of the police.
VI. Police Image Compared to Other Major Social Institutions
If attitudes toward the police are enmeshed in a larger complex of attitudes toward government, law, politics, and other social institutions, how well do the police fare in comparison with these other institutions? Polling organizations have been asking Americans about their confidence in various social institutions for a number of years. Only in 1993 were the police added to this list, which now contains 13 institutions. Since 1993, the police have inspired either the second or third most confidence of all the institutions listed. During that period, the public expressed the most confidence in the military every year, with almost 65% of respondents on average reporting either a great deal or quite a lot of confidence. Out of the eight years with available data, the police came in third place four times, second place three times, and tied for second place once. In terms of public confidence, the police are in a neck-and-neck race for second place with the church and organized religion; during the previous eight years, 56.5% of respondents, on average, report having a great deal or quite a lot of confidence in the police, compared with 56.25% for the church and organized religion, 1999).
One observation about these data is noteworthy. While more than 56% of respondents expressed confidence in the police in 2000, only 24% expressed confidence in the criminal justice system. This was the lowest rating in the list, tied only with Congress. It is unknown why the police, a major component of the criminal justice system, fare much better in the eyes of the public than the criminal justice system as a whole. One possibility is that respondents may associate the criminal justice system with lawyers. As we will show later, lawyers are viewed by the public as among the least honest and ethical professionals, generating levels of confidence similar to those who sell cars or insurance. One possibility is that the public is responding to the mission and motivations they attribute to police. If the police mission is seen as bringing wrongdoers to justice and helping those who are wronged, then that is a simpler, more easily conceived mission than one for the criminal justice system. The courts in particular, operate in theory at least as an adversarial system in which one side tries to convict wrongdoers and the other attempts to get them acquitted or minimize their punishment. Such a construction has a zero-sum quality, where the more one side wins, the more the other loses. Faced with assessing a more complex role, perhaps many citizens select one aspect or the other, and invariably find the courts wanting when they attempt to accomplish both simultaneously.
Finally, a recent Roper survey suggests that the American people find that the police are among the top five values for services received from their tax dollars (Roper, 2001). Fifty-seven percent of the survey respondents rated police/law enforcement as an “excellent” or “good” value. The services ranking higher than police (from top to bottom) were military defense (63), medical and technological research (61), public television (58).[7]
VII. Police Image from Community Surveys
Numerous municipal police agencies around the country conduct community surveys to assess citizen attitudes toward issues related to crime and justice. Items asking respondents about their overall image of their local police are almost always included in these surveys. It is not uncommon for local police departments to collaborate with criminal justice researchers with university affiliations when conducting community surveys. These working relationships are usually mutually beneficial. On the one hand, police executives are provided technical assistance in constructing and administering the survey, as well as aid in data analysis and in interpreting results. On the other, criminal justice researchers gain access to the survey data to assess research hypotheses that may eventually appear in academic journals. Research articles investigating citizen attitudes toward the police are fairly common. For the most part, these surveys have consistently revealed a high level of support for local police. As one might imagine, however, the quality of these research reports varies considerably.
Instead of wading through a vast number of research articles of varying quality, we can gain a good understanding of how citizens feel about the police in their cities by looking to a recent study conducted by the Census Bureau under the auspices of the Bureau of Justice Statistics (BJS) and the Office of Community Oriented Policing Services (COPS)(Smith et al., 1999). Using rigorous survey research methods, the Census Bureau conducted a large number of telephone surveys in 1998. The sample consisted of 12 cities, which were located throughout the United States, whose police departments community policing initiatives were at different stages of development. Among other questions, citizens were asked, “In general, how satisfied are you with the police who serve your neighborhood?” Results from the survey showed quite clearly that a large majority of citizens residing in each city were either “very satisfied” or “satisfied” with their local police department (see Exhibit 6). The percentage of citizens responding positively to the question ranged from a high of 97% (Madison, WI) to a low of 78% (Washington, D.C.). Given the diversity of communities in this sample, the relatively small range in satisfaction levels is remarkable. Eight of the twelve departments fall within the 84-89 percent satisfaction range. The average satisfaction score across the 12 cities was 85% (Smith et al., 1999:25). In sum, the most reliable available evidence suggests that a considerable majority of citizens from city to city evaluate their local police in favorable terms.
To interpret these highly favorable ratings of local police service to the respondent’s own neighborhood, we should keep several things in mind. First, it is a well established pattern that Americans tend to evaluate specific individuals who serve or represent them more positively than they do the institutions in which those public officials work. For example, while Congress routinely receives low marks of respect and satisfaction, survey respondents also routinely rate their own congressperson much higher. Second, the phrasing of the particular question used in the twelve-city survey (focusing on “satisfaction”) is, as we have shown earlier, more likely to elicit a positive response than a question phrased about “confidence in” the police.
Of course, we would expect more variation with a larger sample that included a wider range of departments by size, region, and other characteristics. Even so, the modest amount of variation across this rather diverse sample is surprising if one expects that public opinion reflects differences in objective levels of performance. If, as many police professionals assume, there is a great difference in the quality of policing among communities, why is this not reflected in these surveys? Two possibilities occur to us. One possible answer, which we explore below, is that the critical differences in police performance occur, not from department to department, but from neighborhood-to-neighborhood within departments. Although this particular survey asked respondents to assess police service to their own neighborhood, the survey did not construct its sample around neighborhoods, so we are unable to compare satisfaction levels from neighborhood to neighborhood. We also caution that over-all satisfaction scores can be very misleading if dissatisfaction tends to be concentrated in certain groups of citizens (for example, by race or wealth). A better way to make cross-department comparisons would be to provide a breakdown by such factors as race and wealth – to make it possible to see how each department compares to others in dealing with the groups of citizens who have traditionally shown more reluctance to support the police. In 10 of the 12 cities white respondents were more satisfied with their local police than were African-Americans. Overall, 90% of whites reported to be either “very satisfied” or “satisfied” with the police. In contrast, 76% of African-Americans and 78% of “other” minorities (e.g., Asians and Pacific Islanders) were either “very satisfied” or “satisfied.” Despite these differences, it should be noted that the majority of citizens from all racial groups rated the police favorably.
A second possible reason for the low level of variation in citizen satisfaction across departments is that these departments differ in their styles of policing (as reflected in what officers do and accomplish), but that the particular style a department displays is closely matched to the preferences of most citizens in each city. It may be that different styles produce similar levels of satisfaction because local police departments are customizing their efforts to local preferences.
Although informative, the BJS/COPS survey results also limited in certain respects. In particular, questions about community policing and one’s neighborhood were asked only to residents aged 16 or older (Smith et al., 1999:v). Accordingly, results from the BJS/COPS survey do not provide an accurate assessment of how juveniles view their local police because their responses were lumped together with adults. A recently conducted study that surveyed public school children in Cincinnati, Ohio sheds some light on how children view their local police. Hurst and Frank’s (2000) sample consisted of 852 children from grades nine through twelve from three schools in the greater-Cincinnati area. Results from their survey revealed that a majority of students report that they trust the police (60%), like the police (50.1%), were satisfied with the police in their neighborhood (54%), and thought police officers were doing a good job (61%)(Hurst and Frank, 2000:196)(see Exhibit 7). Using more sophisticated multivariate statistical techniques, the authors found that differences in evaluations of local police existed between racial groups. In particular, non-white students reported less favorable evaluations of the police when compared to whites (Hurst and Frank, 2000:197).
VIII. Conclusion
Police in America enjoy relatively high levels of satisfaction, support, confidence, and esteem from the public, but this is not necessarily cause for complacency. There are two reasons for this. First, at least by a couple of indicators (public confidence in the police generally and respect for the police in the citizen’s own area), the image of the police appears to have been declining since the mid-to-late 1960s. Second, although American police still evoke a positive image for the majority of the public, one might argue that Americans – regardless of background and status – are socialized to expect high levels of performance from their police, something that the police themselves encourage. A business that experiences even 15-20 percent levels of luke-warm-to-negative assessments from its customers would have cause for grave concern. Of course, the police “business” is in some respects like few others because many “clients” are involuntary, but police leaders of today often present themselves as adopting the perspective and methods of successful private sector enterprises. The public may begin to hold police leaders accountable for their promises about making their organizations “results-oriented” and “customer-friendly.” This may well change expectations faster than performance, which can ultimately result in a less positive image.
From a practical perspective, when even modest levels of negativity and low esteem are concentrated disproportionately in certain groups of the public, this can become highly problematic for the police – especially when members of these groups begin to coalesce and organize around these feelings. They usually mobilize to call attention to their unhappiness and may undertake to persuade others that things need to change. When they are successful, nearly always the first thing to change is who heads the police organization. Social scientists have not systematically examined how well patterns in the general public image of the police predict various forms of support and resistance for police and police leaders. This would be a valuable topic for future research.
CHAPTER 3
PUBLIC PERCEPTIONS OF THE OUTCOMES OF POLICING
I. Introduction
In this chapter we review research on outcome-oriented elements. “Outcome-oriented elements” refer to goals citizens expect the police to achieve. These elements include goals that have long been part of the police mission, such as solving crimes. Outcome-oriented elements also fall under the rubric of community policing, such as reducing fear of crime and developing a sense of community among neighborhood residents. These outcome measures are frequently assessed using community surveys that ask citizens their feelings and perceptions of crime and justice issues.
Are outcome-oriented elements linked to citizens’ overall image of the police? Contemporary criminal justice researchers believe they are related. What is the nature of the relationship? Research shows that citizens who evaluate the quality of life in their neighborhood in favorable terms are more likely to report a positive overall image of the police. Citizens appear to hold the police accountable for neighborhood quality of life. Research also shows that citizens believe their neighbors share responsibility for preventing crime and should work with police to address neighborhood issues. In other words, citizens who believe they can trust their neighbors also perceive neighborhood conditions positively and also hold the police in higher regard. Before exploring this issue in greater detail, we will identify some variables that contemporary police agencies commonly use to measure outcomes (or performance).
II. Different Ways to Measure Police Outcomes
As the movement toward community policing and away from a pure “crime-fighting” model (i.e., reduce crime and victimization) continues, police departments have looked for new measures of police performance. Traditional performance measures, such as crime rates, arrests, clearance rates, and response times, are still used regularly. Contemporary researchers argue that these measures of performance are limited because they “fail to capture the many important contributions that police make to the quality of life” (Alpert and Moore, 1993:110). Community policing calls upon officers to build meaningful relationships with the communities they serve and work with residents to address neighborhood problems. So, many police executives regularly supplement traditional performance indicators with a host of measures that reflect police efforts to improve citizen quality of life.
David Bayley (1994) calls these outcome-oriented elements “direct-soft” performance indicators. “Direct” measures refer to what police have achieved over time, and “soft” indicators focus on “subjective perceptions of change” (Bayley, 1994:97-98). To understand the different types of outcome-oriented elements better, we make two important distinctions. First, we can identify outcomes related to traditional police efforts concerning crime, such as providing protection, solving cases, and prevention. We term this the “police-crime” dimension. The second distinction concerns the community. Here, we take into account two factors: citizens’ perceptions of the social conditions of their neighborhood and residents’ perceptions of crime. The former, which we call the “community-social” dimension, includes measures that gauge police attempts to strengthen a neighborhood’s social fabric. Among these measures include social cohesion (e.g., trust among neighbors; see Sampson et al., 1997), sense of community (e.g., mutual feelings of belongingness; see Chavis and Wandersman, 1990), collective security (e.g., watching each other’s home; see Cao et al., 1996), and integration (e.g., neighbors help on another; see McGarrell et al., 1997). Citizens’ perceptions of neighborhood crime (or the “community-crime” dimension) include such measures as fear of crime, perceived crime, perceived social disorder and physical decay, and risk of victimization. These dimensions along with the corresponding measures are presented in Exhibit 8.
Some outcomes reflect traditional police practices, such as solving crimes, while others tap into police initiatives consistent with community policing that address quality of life issues, such as neighborhood sense of community and fear of crime.
III. Police-crime Outcome-oriented Elements: Results from National Polling Data
National polls have used outcome-oriented items tapping into traditional goals of policing, such as preventing crime and protecting citizens. Shaw and colleagues (1998) provide the results from six national polls asking citizens to rate “police protection” in their locality. When averaged together, results show that 60% of citizens believe police protection in their area is either “excellent” or “good” (see Exhibit 9). Specifically, results from the 1983, 1986, and 1992 Roper Poll show that a majority of citizens favorably rate police ability to provide protection (61%, 56%, and 67%, respectively). The poll conducted in 1993 by the Princeton Survey Research Associates revealed that 73% of adult Americans rate police protection as either “excellent” or “good.” The 1994 Harris poll reported similar results. The results from the Gordon Black Corporation poll appear inconsistent with the other five polls in that less than half (35%) of respondents rated police protection favorably. One must use caution when interpreting these results because, in comparative terms, the number of persons sampled was quite low (sample size = 482). Taking this into account, public ratings of the quality of police protection appear to have declined in the 1980s but had increased in the 1990s. Without knowing what made citizens feel more and less safe during these time periods, we can only speculate about the causes of this trend. Perhaps the increase in police efforts in a highly publicized “war against drugs” and increased efforts to work closely with the community through community policing encouraged more people to rate police protection higher.
Whatever the cause, by the mid-1990s, public confidence in the police capacity to deal with crime had achieved impressive levels. Using data from the 1995 National Opinion Survey on Crime and Justice (NOSCJ), Huang and Vaughn (1996:38) assessed public opinion of three traditional police outcomes. In particular, they asked respondents “how much confidence” they had in the “ability of the police” to (a) provide protection from crime, (b) to solve crime, and (c) to prevent crime (see Exhibit 10). Results from the NOSCJ showed that a sizable majority of respondents expressed confidence in the police to protect citizens from crime (74%), solve crime (74%), and prevent crime (65%) (Huang and Vaughn, 1996:38). The public rated the reactive capacity of the police to deal with crime (that is, to “solve” it) higher than their capacity to prevent it. It is not obvious to us what it means to “protect citizens from crime.” On the face, it could be both reactive and preventive, but given that its level is identical to the reactive measure, it would appear that the public views it as a predominantly reactive enterprise.
Although the results from the NOSCJ were quite positive, Huang and Vaughn (1996) did observe differences across social groups. The authors found that African-Americans reported lower levels of confidence in the police to provide protection from crime (60%), solve crime (61%), and prevent crime (56%) when compared with whites (77%, 79%, and 68%, respectively) and Hispanics (83%, 74%, and 73%, respectively) (Huang and Vaughn, 1996:40). They did not find gender differences in confidence for the three police outcomes. In sum, both men and women expressed a high level of confidence in police to provide protection from crime, solve crime, and prevent crime.
Worrall (1999) used more sophisticated statistical techniques to assess differences across different groups using the NOSCJ data. Worrall did not find differences in confidence in the police to protect citizens from crime, solve crime, and prevent crime among respondents with different incomes and educational backgrounds, between men and women, nor among respondents of different ages (Worrall, 1999:59). In contrast to Huang and Vaughn (1996), Worrall (1999) observed no differences regarding race across the three outcome-oriented variables. How can we explain this inconsistency? The lack of observed differences across racial groups was likely due to Worrall’s (1999) coding of the race variable. Specifically, Worrall (1999:55-56) lumped African-Americans and Hispanics into one category (termed “nonwhite”). Huang and Vaughn’s (1996) analysis shows that Hispanics and African-American citizens express drastically different levels of confidence in the ability of police to produce preferred outcomes. We can reasonably conclude, then, that Worrall’s (1999) combining African-Americans and Hispanics into one category nullified the differences observed by Huang and Vaughn (1996).
IV. Community Outcome-oriented Elements and the Public’s General Image of the Police
Research investigating the link between community outcome-oriented elements and the public’s general image of the police are emerging in the literature. These recent efforts provide considerable support for the notion that community outcome-oriented elements, such as citizens perceptions of neighborhood conditions, affect the public’s general image of the police.
Using community survey data from Indianapolis, Indiana and St. Petersburg, Florida – collected as part of the Project on Policing Neighborhoods – Reisig and Parks (2000) found that several community outcome-oriented elements influenced citizens’ overall satisfaction with police. In particular, citizens who perceived crime as more problematic, believed that safety was an issue, perceived incivilities (e.g., social disorder and physical decay) as widespread in their neighborhood, and rated their neighborhood poorly, also expressed lower levels of satisfaction with their local police departments. Reisig and Parks’s (2000) statistical techniques allowed them to adjust their findings for several factors that also influence that public’s overall image of the police, such as prior contact with the police, perceptions of police resource distribution, familiarity with individual police officers, neighborhood poverty, and official neighborhood crime rates. Results from their analysis led Reisig and Parks (2000) to conclude that citizen community outcome-oriented elements are the most important factors influencing the public’s image of local police. They did not include most of the process measures described in Chapter 6 (e.g., police fairness, manners, responsiveness), so we are unable to use this study to compare the effects of outcome and process image.
Additional research supports Reisig and Parks’s (2000) finding. Using community survey data collected in Cincinnati, Ohio, Cao and his colleagues (1996) found that citizens who believed that neighborhood disorder (e.g., noisy neighbors, litter, and loitering teenagers) was a problem also reported less confidence in the police. What is more, Cao and colleagues (1996:11) found that citizens who believed their neighbors were willing to help provide protection from crime (termed “informal collective security”) reported higher levels of confidence in police. Similarly, Percy (1986:80) found that citizens in Fort Worth, Texas who perceived that crime was getting worse reported lower levels of satisfaction with police service in their neighborhood. Hurst and Frank (2000:197) found that juveniles who reported that their neighborhood had higher crime rates than other residential settings also reported less general support for their local police. Smith et al. (1999:26) analysis of the BJS/COPS 12 city survey data also revealed that respondents who feared crime were also less satisfied with the police who serve their neighborhoods.
The evidence from available research reports is clear: community outcome-oriented elements influence citizens’ overall image of the police in their communities. We provide a summary of the research studies investigating the link between community outcome-oriented elements and overall police image in Exhibit 11.[8]
V. Responsibility for Crime Control: Neighborhood- and Citizen-level Differences
As best we can tell, systematic comparisons of outcome-oriented elements between neighborhoods do not exist. Neighborhood-level studies have focused on whether citizens residing in different neighborhoods share the view that controlling neighborhood crime is the responsibility of police and citizens. Questions of this type are relevant. After all, many community policing programs involve an active role for citizens. That is, citizens are encouraged to work with police to control crime and social disorder.
Research shows that a good portion of citizens share the view that controlling crime is a shared responsibility. Dunham and Alpert’s (1988) Miami study asked citizens whether “only the police can control crime.” The analysis revealed that residents from a middle-class white neighborhood, a low-income black neighborhood, and an upper-middle-class black neighborhood disagreed with this statement. However, residents from two Cuban neighborhoods were generally supportive of the traditional view that controlling crime was the responsibility of the police. These findings suggest that differences in attitudes toward crime control responsibility are similar across race and class, but differences exist along ethnic lines.
Another neighborhood study, which used community survey data from a small university town in a northwestern state, revealed that residents from four neighborhoods all disagreed that “only the police can control crime” (Reisig and Giacomazzi, 1998). What is more, these neighborhoods differed significantly in family income, home ownership, and occupation. These findings provide support for Dunham and Alpert’s (1988) conclusion that attitudes toward crime control responsibility do not vary across neighborhoods characterized by different levels of socioeconomic status. In addition, the neighborhoods included in Reisig and Giacomazzi’s (1998) study also differed in perceptions of neighborhood disorder, fear of crime, and integration. In other words, no matter the socioeconomic status of the neighborhoods, and despite differences in levels of perceived ills across neighborhoods, citizens from these different neighborhoods viewed crime control as the responsibility of both the police and citizens (Reisig and Giacomazzi, 1998:554, 560-1).
Webb and Marhsall (1995) also addressed the question of responsibility for crime control, but adopted the citizen as a unit of analysis as opposed to neighborhoods. Their multivariate analysis revealed that Hispanic residents were more likely than whites to support the notion that only the police can control crime. Nevertheless, they also found that differences did not exist between men and women, across different age groups, nor between blacks and whites (Webb and Marshall, 1995:57). Results from Webb and Marshall’s (1995) study also supported Dunham and Alpert’s (1988) earlier finding that attitudes toward crime control responsibility vary across ethnic lines. Specifically, Hispanics are more likely to hold a view that is consistent with the professional model of policing.
VI. Conclusions
Most of the research on the public’s views of the capacity of police to produce desired outcomes has focused on crime-related outcomes. We found considerable fluctuation in public perceptions, depending upon the year and the polling firm. In general, substantial majorities of the public offer positive assessments of the police ability to achieve crime-focused outcomes. Less favorable assessments were more heavily concentrated among African Americans compared to white and Hispanic citizens. Interestingly, citizens appear to view crime control as a jointly held responsibility between the police and the public.
CHAPTER 4
PUBLIC PERCEPTIONS OF POLICING PROCESSES
I. Introduction
There are many ways to measure service quality. A waiter might be friendly and helpful but record the order incorrectly. A clerk at the Department of Motor Vehicles might process paperwork quickly and efficiently, but ignore the hellos, goodbyes, and other basic niceties of a personal transaction. Generally as citizens and consumers, we are pleased by service transactions that are successful on all dimensions, we ignore those that are successful in some regards but not others, and we are outraged at those featuring service that is poor all-around. Research on service quality in the private sector has been helpful for delineating some of the dimensions that citizens, consumers, and clients associate with quality service (Parasuraman, et al., 1988). Based on this research, Mastrofski (1999) has outlined six characteristics that Americans associate with quality service delivery from their police: attentiveness, reliability, responsiveness, competence, manners, and fairness. He characterizes these dimensions as constituting a style of policing known as “policing for people.”
We do our best to organize this chapter according to the six dimensions of service quality outlined by Mastrofski. That is difficult for at least two reasons. First, in research, as in life, not everything fits neatly within the little boxes or categories we would like to use. Terminology varies from study to study: one study may examine the “manners” of the police, while others look at their “politeness” or their “friendliness.” And, as we learned in previous chapters, though these differences in terminology may appear subtle, they often produce important differences in how we view the public image of the police. Despite some difficulty in slotting all of the previous research findings about the public image of police processes, these six dimensions are a useful tool for making sense out of a large body of research over the past several decades. Second, there is an uneven amount of research on these categories. For instance, fairness has probably produced the greatest amount of research, but very little nationally is known about the public image of police in some of the other categories like reliability or competence. In addition, we add a seventh dimension that we view as generic to all occupations and not specific to police: integrity. Although integrity is not an issue unique to policing, it is a crucial issue for the public image of the police worldwide.
II. Generic Dimensions of the Quality of Service
Attentiveness
We were unable to locate any studies on how the public feels about the attentiveness of the police. Perhaps the closest we can come in addressing attentiveness is looking at surveys of crime victims. For instance, Homant, Kennedy, and Fleming (1984) found, like many other studies, that crime victims have less satisfaction with the police. However, unlike many of the studies linking victimization experience to satisfaction with police, they also found that "a policy of providing the citizen with some crime prevention counseling can go a long way toward undoing the negative effects of victimization on attitudes toward the police" (p. 331). In other words, while research has demonstrated clearly that victims are less satisfied than non-victims with police, when the police go out of their way to help victims prevent future victimization, they are able to mitigate the effect of the incident on their public image. For this reason, the authors conclude that victimization represents not only a crime problem, but a community relations problem. While this one study does not provide definitive evidence that attentiveness influences the public image of the police, it does suggest that this hypothesis deserves further research.
Reliability
According to Mastrofski (1999:2), “People expect a degree of predictability in what police do. They want service that is timely and error-free.” Once again, we are not aware of any studies that have directly examined the impact of reliability on the public image of the police. However, there is some indirect evidence that might provide some clues about the role of reliability. When service is reliable, it is predictable. As Mastrofski argues, “McDonald’s succeeds not because the cuisine is superb but because the food is predictable and more-or-less error free” (p. 2). A McDonald’s hamburger is pretty much the same whether you order it in Boston or Budapest, Chicago or Shanghai. Like a McDonald’s customer, citizens have expectations about how the police will behave.
Researchers have examined citizen satisfaction with police from several theoretical perspectives that take into account these expectations. For instance, Reisig and Chandek (2001) applied a concept from consumer psychology called “expectancy disconfirmation” to the study of citizen satisfaction in police-citizen encounters. They found that citizen satisfaction with police in both voluntary (such as seeking assistance) and non-voluntary (such as traffic stop) encounters is influenced by the citizen’s initial expectations about how the police would behave. Not surprisingly, they found that when police performance exceeded the citizen’s prior expectations, the citizen tended to be more satisfied with how the officer handled that event. When police performance fell below prior expectations, then the citizens tended to give police performance lower ratings for that event.
Erez (1984) found something a little different. She applied the concept of “distributive injustice... [which] occurs when a person does not get the amount of reward he expects to get in comparison with the reward another person gets" (p. 1281). Erez (1994) found that offenders who were arrested were not less satisfied with police than non-offenders because the behavior of the police met their expectations. She surmised that offenders "become accustomed to harassment and thus become desensitized and indifferent" (p. 1297). Even though the evidence on reliability is weak, taken together, these studies suggest that citizens accept service from the police that meets or exceeds their expectations.
Responsiveness
Another element of police service that citizens presumably value is responsiveness. Huang and Vaughn (1996) found that 78% of respondents had favorable attitudes about the tendency for police in their community to respond quickly to calls for help. However, Hispanics and African Americans reported less favorable attitudes than whites. Brown and Coulter’s study of police responsiveness in Tuscaloosa, Alabama found that citizen satisfaction with police response times was affected by prior victimization experience, age, race, income, and participation in the political process. Citizens who had been victimized were significantly less satisfied with response times, as were those who were younger, nonwhite, poorer, who felt less safe, or who participated less in the political process. Prior contact with police, gender, perceived crime rates in relation to neighboring jurisdictions, and education all had no effect on assessments of response time. Percy (1980) found that while actual response time had an influence on overall satisfaction with police, expected response time also played a major role. The policy implications are similar to our discussion in the previous section on citizen expectations: the public image of the police can be improved if police provide citizens with a realistic estimate of the time it will take police to arrive.
Competence
Another element of good service delivery is simply getting the job done right (Mastrofski, 1999:2). For example, if crime prevention is an appropriate goal for police, then perhaps it is fair to ask the public whether the police can give the public in general and victims in particular useful tips on avoiding crime. And given the pervasiveness of handling domestic disputes and the growing expectation that the police should play a key role in reducing repeat problems from the same household, this would seem to reflect some expectation that the police should know how to resolve these disputes so that they do not recur. Citizens may also judge competence in resolving disputes according to such criteria as “taking charge” or using the least coercion necessary to get the job done (standards that are widely endorsed among police officers). Or, they may define competence in terms of the mode of resolution selected (e.g., getting both parties to agree on a course of action, banishing one party, arresting one or both disputants, etc.). At this point we can only speculate on how the public defines competence in resolving disputes. Indeed, we are unaware of any studies that attempt to illuminate how the public actually defines police competence in any of a variety of situations and problems the police routinely confront.
There are a number of reasons why police should be concerned with the definitions citizens use to define competence. First, where professional definitions of competence are strikingly different from those used by the public, the police should carefully examine both to see whether either should be changed. If, for example, the public believes that competent crime scene investigations should always involve dusting for prints, the police department might consider strategies to inform the public about when this activity is unlikely to produce useful evidence. Or suppose that large numbers of the public expect that a competent police response to a reported burglary includes taking the time to educate the victim on how to reduce the risk of future victimizations. This may suggest the need for the department to require officers to provide this information routinely. It would be especially valuable to know whether the public tends to define police competence in terms of what officers do or what they accomplish. Because it is often difficult for the public to observe directly many of the outcomes they care about (e.g., reduced crime and disorder in the neighborhood), they may tend to rely on “proxy measures” – things they can observe directly. Thus, “competent police” may be viewed as those who take a detailed report, because the complainant assumes that the chances of apprehending the offender and recovering stolen property are much greater when all of the information they have is available for follow-up investigators. We suspect that citizens’ expectations about competence are heavily influenced by mass media portrayals of the police, which we previously noted tend to present a highly distorted, and unrealistic picture of police work.
Another reason to be concerned about the public’s image of police competence is that it may have a strong influence on citizen’s general satisfaction with and support for the police. In section V of this chapter we review the findings of one study that indicates that citizens’ views of police competence do indeed have a significant impact on the citizen’s confidence in and support for the police (Tyler, 2001a).
Manners
Citizens also want police officers who are polite and friendly, who are well mannered, who have a good temperament, and who treat them well. According to Huang and Vaughn (1996), the vast majority of Americans (78%) hold favorable attitudes on the friendliness of the police. However, like other elements of police image, the public is not uniform in its assessment, since those respondents who were younger, unmarried, male, and black expressed the least favorable opinions. Huang and Vaughn’s (1996) thorough review of the research concluded: "research suggests that police temperament, politeness, and deportment are important contributors in citizen opinions of police friendliness" (p. 39; also see Brandl and Horvath, 1991; Garofalo, 1977; Hindelang, 1974; Lasley, 1994; Murty et al., 1990). This finding is not limited to the United States. Research in Britain also shows that manners matter (Bradley, 1998; Skogan, 1994). For instance, Stone and Pettigrew (2000) found in their study of police stop-and-frisk practices in England that negative feelings from the stops resulted when officers were "patronising, arrogant, aggressive or intimidating" (page iv).
Three Harris polls have asked random samples of American adults to rate the police in their community on “being helpful and friendly.” In 1992, 73% of respondents rated police as “excellent” or “very good” on this dimension. By 1999, this figure had dropped to 68%, and by 2000 it had risen to 74%. Perhaps more importantly, the 2000 Harris Poll also provides some evidence about how citizens view the friendliness and helpfulness of police in relationship to other elements of policing. Exhibit 12 shows that citizens view the police as more successful in being “helpful and friendly” than in responding quickly, refraining from using force, treating people fairly, preventing crime, and solving crime. This is both good and bad news. On one hand, the number of citizens viewing the police as helpful and friendly is quite high. On the other hand, it is curious that the public views the police as more helpful and friendly than responsive, fair, or effective. We now explore the public image of the fairness of the police.
Fairness
Of the seven dimensions of service quality discussed in this section, fairness is the one that has generated the greatest amount of research. Therefore we will spend substantially more space sorting through the findings of this research. Research on juveniles by Leiber, Nalla, and Farnworth (1998) has identified several attributes thought to contribute to the perception that police are unfair, including: being nonwhite, having delinquent attitudes, prior contact with police, low socio-economic status, and having been wrongly accused. Smith and Hawkins (1973) found that education, income, occupation, and sex had no effect on citizen ratings of police fairness. Both studies, as well as others that we will discuss shortly, found that race has an important effect on perceptions of fairness. Stone and Pettigrew (2000) found in their British study that "public trust and confidence is primarily based on being treated fairly and with respect and being given a good reason for the stop, rather than on changes in procedure" (page iii).
In one important way, the findings here parallel the findings we have reviewed in other sections of this report, with race exhibiting the strongest impact on perceptions of fairness. In Huang and Vaughn’s (1996) national survey, 78% of respondents expressed favorable attitudes about police fairness, with African Americans, younger, single, and low-income respondents reporting less favorable views. The effect of race on perceptions of police fairness has been consistently documented in the research, with African Americans expressing the belief that they are treated less fairly than whites by police (Biderman, et al., 1967; Huang and Vaughn, 1996; Smith and Hawkins, 1973; Sullivan et al., 1987). The most recent evidence confirms that this longstanding trend still holds in the United States. A Harris poll administered in 2000 found that while 69% of whites thought that police in their community treated all races fairly, only 36% of blacks believed this to be true (Pastore and Maguire, 1999, p. 110).
As in other sections of the report, the most important question here is whether it is the race of the individual that is most important, or whether the race effects found in this body of research are a product of larger social forces, such as a racial subculture or composition of the neighborhood where the individual lives. Once again, there is a small body of research to draw on in answering this question, though there is not enough evidence to draw firm conclusions at this point. For instance, Smith and Hawkins (1973) interpret their race effect as evidence of a"sub-cultural phenomenon of uniform hostility" for the police (p. 138). Evidence from a study of juveniles by Leiber, Nalla, and Farnworth (1998) found that minorities and those with less economic security perceive lower levels of fairness by police, but number of parents in the home, and the desirability of the neighborhood both had no effect. Since disadvantaged neighborhoods often feature single parent households and are perceived as being less desirable, these findings are contrary to the notion that neighborhood is important. On the other hand, other findings in the same study "imply that, quite aside from the nature of police encounters with juveniles, the imposition of legal authority and social control in certain neighborhoods engenders a pervasive resentment and resistance, and that youthful residents of those neighborhoods harbor a general disrespect for the law itself" (p. 169).
Erez (1984) found that police did not chase, question, or warn blacks more than whites, though they did have these contacts with offenders more than non-offenders. Thus, the differences between black and white assessments of police treatment cannot be explained by blacks having had more of these contacts with the police. Blacks were searched more than whites however. Based on these findings, she concluded that “something other than actual experience may account for blacks ' negative assessment of the police” (p. 1297). Some of the reasons she mentions include (a) police are seen as part of an oppressive white regime, (b) their attitudes may be shaped by observing, rather than experiencing, police misconduct, or (c) that blacks expect more from police than they get.
Koenig (1978) found that the strongest declines in evaluation of the police were seen among those who "had experienced, or personally observed, what they perceived to be improper field practices" (p. 246). These included: impolite or rude treatment, unfair treatment when arrested or suspected of a crime, physical mistreatment, a police officer covering up another officer 's wrong-doing, a police officer taking sides in an argument between citizens, and an officer not performing required duties. 13% of the respondents claim to have personally observed one or more of these instances of misconduct, while 26% "claimed secondhand knowledge of situations involving these practices" (p. 246).
Taken together, these findings suggest that racial disparities in assessments of police fairness may be a function of the areas where whites and nonwhites, particularly African-Americans, live and spend most of their time. Although African-Americans may not have experienced contact with the police themselves, they may be exposed secondarily to unfair police treatment by witnessing such events in their neighborhoods or hearing about it from friends. If racial disparities in the public image of police fairness are shaped vicariously through stories of police misconduct endemic in African-American subcultures or in neighborhoods primarily occupied by African-Americans, then police agencies will need to focus their image-repair efforts in these “hot spots” of discontent. The implication is that a generalized public relations effort may be less effective at improving the image of police among African Americans than a more concentrated effort in African American neighborhoods to change both the behavior of the police and the perceptions of the citizens.
What other kinds of evidence exist about how Americans view the fairness of the police? Exhibit 13 shows the results of three Harris polls asking respondents to rate the fairness of the police in their community. In 1992, 63% of respondents rated the fairness of the police as either excellent or very good. By 2000, this figure had risen to 67%.
Another more extreme indicator of how Americans perceive the fairness of the police is the percentage who fear being arrested when innocent. Exhibit 14 shows the results of two Harris polls asking this question. 1999, 22% of respondents said they feared being arrested though they were in fact innocent; by 2000, this figure dropped to 17%.
Integrity
Although researchers have spent a substantial amount of time probing police integrity, it is not an issue that has received much attention in the public opinion literature. Nonetheless, some research findings are useful for drawing limited conclusions. Since 1976, the Gallup Organization has done several national polls on the “honesty and ethical standards” of a number of occupations (20 in 1977 and 32 in 2000). In 1977, the first year that police were included in the poll, 37% of respondents rated the honesty and ethical standards of police as “high” or “very high.” The police ranked sixth of the 20 occupations included in 1977, finishing behind clergy (61%), pharmacists (59%), medical doctors (51%), college teachers (46%), and bankers (39%). Exhibit 15 shows the trend for all of the years in which data were available between 1977 and 2000. Clearly, the public image of the honesty and ethical standards of police has improved during this period, albeit with considerable fluctuation. By the year 2000, 55% of respondents rated the honesty and ethical standards of police as “high” or “very high.” Several new occupations were added to the list over the years; by 2000, the police ranked tenth of the 32 occupations included in the list.[9] One noteworthy feature of the trend in Exhibit 15 is that some of the year-to-year fluctuations are fairly dramatic. It is not known what drives these yearly fluctuations, but as we will examine shortly, one explanation may be major public image crises like the Rodney King incident in Los Angeles. Understanding the role of such incidents on the public image of the police is one of the most important unanswered questions in the body of research reviewed in this report.
Recall that in this section we have been examining qualities that are presumably generic across the service industries, from public to private. These seven generic dimensions are related to more specific processes that are unique to each industry. For instance, there are a number of practices in policing, like use of force or stop and search procedures, that span across these categories (such as fairness and competence), but are not found in other service industries like restaurants, welfare offices, or auto repair shops. The seven generic dimensions of service quality that we have reviewed in this section are important, but it is also important to look at those aspects of service quality that are unique to policing. That is the task of the next section.
III. Police-specific Dimensions of the Quality of Service
Policing is one of a handful of service professions that asks service providers to deliver services to clients who both request their assistance (voluntary clients) and who do not want their attentions (involuntary clients). This dual nature of policing has been likened to forging an iron fist inside a velvet glove, where punitive actions toward (usually) involuntary “clients” represent the iron fist and assistance aspects to voluntary clients represent the velvet glove (Mastrofski, 1988). Naturally, individual citizens who feel the “velvet glove” aspects will have more positive feelings about their contact with police than those who experience the iron fist. Conversely, most of the public relations problems which the police have experienced come from the “iron fist” activities. In this section, we examine the public image of two realms of coercive police behavior: stops and searches, and use of force.
Stops and Searches
Racial differences in citizen perceptions of police stop and search behavior have been identified in several studies. A 1999 Gallup poll of 2,006 adults found that only 11% of respondents felt that they had been stopped by police because of their racial or ethnic background. There were pronounced differences by age, sex, and race however. For instance, only 6% of whites reported having been stopped because of their race or ethnicity, compared with 42% of blacks. Males were more likely than females, and younger respondents were more likely than older respondents to report having been stopped due to race or ethnicity, (Pastore and Maguire, 1999:111).
In a 1999 national study of contacts between the police and the public, Langan and his colleagues found that most drivers (84%) thought they had been stopped for a legitimate reason. Age had no influence on perceived legitimacy of the stop, but there were significant differences by race and gender. Following are the percentage of each category rating the stop as legitimate: Men (82%), Women (87%), Whites (86%), Blacks (74%), and Hispanics (82%). Of the estimated 19.3 million drivers pulled over by police, 90% thought the police had behaved properly during the stop. Once again, females gave higher ratings than males, and whites gave higher ratings than blacks and Hispanics. Only two age categories differed significantly from the others: teenagers and drivers in their twenties gave lower ratings than other drivers, particularly drivers in their fifties.
In the same study, nearly 1.3 million drivers were estimated to have experienced a search of their person, their vehicle, or both. An estimated 34% felt that the search was legitimate, while the other 66% thought the search was not legitimate. There were no significant differences by gender or age. Once again, fewer blacks (17%) thought the search was legitimate than whites (39%) or Hispanics (33%). 35% of the estimated 829,000 experiencing a physical search of their person viewed the search as legitimate. There were no significant differences by gender or age. Once again, fewer blacks (22%) thought the search was legitimate than whites (40%) or Hispanics (28%). Of the approximately 1 million drivers experiencing a vehicle search, 34% viewed the search as legitimate. There were no significant differences by gender. Older drivers viewed searches as more legitimate than younger drivers. Fewer blacks (15%) thought the search was legitimate than whites (39%) or Hispanics (31%).
Homant and Kennedy (1994) found that more than 75% of the respondents in their study of Detroit area voters thought police use good judgment in deciding whether to pursue suspects in vehicles, though 40% thought that there should be more restrictions placed on the police with regards to pursuits. Attitudes about pursuits were unrelated to age, gender, and education. Support for police pursuits was found to be related to political orientation, however, with those who were more conservative reporting stronger support for police pursuits.
Stone and Pettigrew’s (2000:iii) study of stop and frisk behavior in England found that "public trust and confidence is primarily based on being treated fairly and with respect and being given a good reason for the stop, rather than on changes in procedure." Based on their focus group findings, they concluded that positive feelings emerging from the stops and searches resulted from the following: (1) being given an acceptable reason for the stop, (2) the officer was polite, (3) the stop was brief, and (4) people did not feel unfairly targeted (page iv). As we reported earlier, negative feelings from the stops resulted when officers were "patronising, arrogant, aggressive or intimidating" (page iv).
The public image of police stop and search behavior has become a major public policy issue in recent years with the massive media attention focused on racial profiling. We identified two national polls on public perceptions of racial profiling, a Gallup poll conducted in December 1999, and another poll conducted in June 2000 by Penn, Schoen, and Berland (PS&B) Associates. In the 1999 Gallup poll, 59% of respondents agreed that racial profiling was “widespread.” Consistent with other findings in this report, 56% of whites viewed the practice as widespread, compared with 77% of blacks. Racial differences in approval or disapproval for racial profiling are much less dramatic: 80% of whites disapproved of racial profiling, compared with 87% of blacks. In the 2000 PS&B poll shown in Exhibit 16, 75% of respondents viewed racial profiling as a problem. The results were not available by race, but were broken down by political orientation: 82% of Democrats viewed racial profiling as a problem, compared with only 63% of Republicans. Furthermore, 69% of respondents thought police should be banned from taking race into account when targeting people as suspects; 77% of Democrats agreed, compares with 60% of Republicans. These findings on the relationship between political orientation and attitudes toward police stop and search behavior are consistent with those on police pursuits that we reported earlier by Homant and Kennedy (1994).
One of the most recent pieces of evidence on public opinion toward racial profiling, though not national, comes from an April 2001 poll of 802 New Jersey adults. The study found that 61% viewed racial profiling as a “big problem” or “somewhat of a problem.” Once again, however, disaggregating the findings by race showed that 31% of whites viewed it as a big problem compared with 82% of blacks (Star-Ledger/Eagleton-Rutgers Poll, 2001).
We note that views on racial profiling may be strongly influenced by highly publicized national events. In the wake of the September 11, 2001 terrorist attacks on New York City and Washington, DC, the public may see threats to national security as so severe that racial profiling to catch terrorists is justified. A national Gallup poll found that 54 percent of blacks favored singling out Arab-Americans for special scrutiny at airport check-ins, while 53 percent of whites and 63 percent of Hispanics opposed this action (Scales, 2001). A poll of Arab Americans in the Detroit metropolitan area found that 61 percent felt that extra questioning or inspections of people with Middle Eastern features or accents by law enforcement officers was justified (Niemiec and Windsor, 2001). Whether this remarkable pattern of views will be maintained over several years may well depend upon the extent to which the news features two kinds of stories: the threat of terrorism to public safety and the harm done to innocent persons screened by profiling practices. Regardless of the long-term effects of the terrorist attacks on views toward racial profiling, the two surveys described above show how profoundly major events can alter public opinion results in the short run.
Use of Force
Nowhere is the public’s discomfort with the “iron fist” of police felt more strongly than in the issue of police use of force, particularly deadly force. Echoing the distinction we made at the beginning of this section between voluntary and involuntary clients, a recent national study of contacts between the police and the public found that the majority (91.9%) of citizens who had force used against them by police viewed the police behavior as improper (Langan, et al., 2001). The differences by race are not as pronounced as in other findings in this study, with 94.4% of blacks viewing the force as improper, compared with 88.5% of whites. Less than 20% of those viewing the force as improper took any kind of formal action such as filing a complaint or a lawsuit. Racial differences in citizen perceptions of police use of force behavior have been identified in several studies (Williams, Thomas, and Singh, 1983; Flanagan and Vaughn, 1995). In addition, Huang and Vaughn (1996) report that "public perceptions of police use of force varied with age, income, community type, and political ideology" (p. 41). Respondents who were older, conservative, rural, and in the middle income bracket had more favorable perceptions about police use of force.
Exhibit 17 provides some evidence about how the public views the police use of force. The National Opinion Research Center has asked respondents annually (with some years missing) since 1973 whether they would approve of a policeman striking a citizen who: (1) said vulgar and obscene things to the policeman, (2) attacked the policeman with his fists, or (3) was attempting to escape from custody? While the majority of respondents endorse the need for police to use force when a citizen attacks the officer with his fists (90% in 1998) or when a suspect attempts to escape from custody (68%), few respondents see the need for force when a citizen says vulgar and obscene things to the officer (7%). Furthermore, as the trend lines in Exhibit 17 clearly show, since 1973 the public has grown less tolerant of the use of force in all three situations. For instance, while 22% of respondents in 1973 said that they would approve of a policeman using force against a citizen who said vulgar or obscene things to him, by 1998 this figure had dropped to only 7%. The public’s taste for police use of force has clearly diminished. This topic has received no research attention; therefore we have no ready explanation for why this occurred. We note, however, that it may reflect a continuation of expanding the application of middle-class values about coercion to police that have been part of a long-term historical process (Bittner, 1970; Fogelson, 1977:ch. 11). That is, the public increasingly expects that police accomplish their work by resorting less frequently to the more physically coercive aspects of their authority.
A series of Gallup polls starting in the 1960s asked respondents whether there was any police brutality in their area. Surveys were completed in 1965 and 1967, and then were not resumed again until 1991, therefore it is difficult to draw any conclusions about trends that occurred in between, during the 1970s and 1980s. Nonetheless, the pattern revealed in these surveys, shown in Exhibit 18, is striking. Averaging the responses in the 1960s and the 1990s reveals a stark contrast in perceptions of police brutality. In the 1960s, approximately
7.5% of respondents believed that police brutality existed in their area; by the 1990s, this figure had risen to approximately 37%. The 2000 poll revealed a slight downturn, with 32% of respondents expressing the opinion that police brutality existed in their area. Once again, there is no research on the factors that led to these changes in public opinion.
Public opinion data cannot tell us whether in fact police have become more brutal since the 1960s, but it clearly there has been a substantial shift in public perceptions that they have. Based on the pattern in Exhibit 17, we might conjecture that at least part of the increased perception in brutality can be attributed to a shift in the public’s standards about what constitutes brutality. Increasing numbers of the public are lowering the threshold of what they classify as brutal.
The 2000 poll also reveals the same patterns outlined in other parts of this report, with important differences in perceptions of police brutality based on the characteristics of the respondent. Although there were no tests of statistical significance done, it appears that males, blacks, conservatives, and those from urban areas were the most likely to report police brutality in their area. Fifty-three percent of black respondents perceived police brutality in their area, compared with only twenty-eight percent of whites. Once again, we do not know whether the race of the individual is producing this effect or whether larger social forces are at work.
Race and the Image of Police
Throughout this report we have made reference to the differences between racial groups in their attitudes toward and assessments of the police. In this section we attempt to address this issue across the broadest possible range of image indicators, concentrating especially on process measures. Exhibit 19 presents racial breakdowns for a survey that was previously presented in Exhibit 12 as reported by Huang and Vaughn (1996:44). The first four attitudes concern police processes – how they do their work. For all four attitudes whites show more favorable attitudes than do Hispanics and, especially, blacks. The difference is most pronounced on use of force. The pattern is repeated for the three measures of police outcomes – with the exception of crime prevention and crime protection, where Hispanics outscored whites.
The researchers also examined the effects of race for the subset of citizens in this sample who had contact with the police within the previous two years. Their analysis controlled for the effects of a variety of personal background characteristics and attitudes of the respondents, as well as their satisfaction with the particular police contact. They found that satisfaction with the police contact was the only significant influence on the citizen’s overall view of the police for the three outcome measures (crime protection, crime solving, and crime prevention) plus two process measures (promptness, and friendliness). Race no longer proved a significant predictor of the citizen’s overall assessment, once the citizen’s view of the specific contact was taken into account. However, race remained a significant influence for the citizen’s overall attitude toward police for fairness and use of force. Blacks were significantly less likely to rate police favorably on fairness, even after taking their satisfaction with their most recent contact into account. The same held for both blacks and Hispanics when considering the use of force.
The persistence of racial effects for fairness and use of force reinforces the view discussed earlier about the power of the belief among minority groups that they receive disparate treatment from police. Stated another way, a single favorable contact with the police will have much less of a positive effect on their beliefs about fairness and use of force for minority citizens than whites. On the other hand, satisfactory police contacts do seem to dissipate the effects of race when considering citizen views of promptness and friendliness. A possible reason for this difference is that it is much easier to demonstrate promptness and friendliness in readily observable ways. Fairness is perhaps more in the eye of the beholder, and therefore more difficult to demonstrate. And officers so rarely use force with minorities, as well as whites, that a given contact would be unlikely to reveal opportune circumstances for the citizen to perceive its relevance. Therefore they are not inclined to generalize from most of these situations to their view about the police use of force in general.
V. The Relationship between Police Processes and the General Image of the Police
One of the most important questions for national bodies like the IACP is how to improve the public image of the police. Are there certain “levers” which the American police can use to improve their image? Are there certain areas in which they can focus their efforts? One of the most valuable pieces of information from a policy perspective is how citizens develop their image of the police, whether through personal experience, through the media, through vicarious experiences of others, through the subcultures in which they are immersed, the neighborhoods where they live, or most likely, through some combination of these channels. One of the important questions is how much citizens’ perceptions of police processes affect their overall view of the police – the kinds of measures we discussed in Chapter 2. In the public’s eye, which of the various aspects of police process are the most and least influential? How much influence do public perceptions of police process exert compared to public perceptions of policing outcomes? The answers to these questions are quite important to police, because they indicate what aspects of their image are most important to the public. Knowing this can help police be more effective and efficient when they wish to improve their overall standing with the public.
Unfortunately, there is not a great deal of research that answers these questions, but the good news is that researchers are beginning to address them. At this stage we are able to report on a handful of studies. Some examine the views of citizens generally – without regard to specific experiences they may have had with police. Others ask citizens to reflect on specific encounters they have had with police.
Benson (1981:59) found that respondents who believe that police lack integrity provide lower ratings of police performance. Those who believe that police lack integrity and who think that crime is increasing "are particularly apt to view police performance poorly." Similarly, Brown and Coulter (1983) found that more than half of the variation in satisfaction with police could be explained by three factors:: satisfaction with response time, satisfaction with the way the police treat people, and the perceived equity of police protection (which means whether people think police protection in their neighborhood is better than that in other neighborhoods). All these effects were positive, meaning that those who saw the police as responsive, fair, and equitable were most satisfied with the police overall.
Tom Tyler has conducted a series of research projects that look at the impact of citizens’ assessments of police processes on the over all image of the police. One research project focused only on citizens who had recent contact with the police or courts (Tyler, 2001b). The vast majority of contacts reported were with police, not the courts. Tyler conducted a telephone survey of residents of Los Angeles and Oakland, California., which yielded a sample containing roughly equal numbers of whites, Hispanics, and African Americans. He created a scale of legitimacy of law and legal authority based on a variety of questions tapping into the respondents’ feelings of obligation to obey authorities, trust in the police/courts, cynicism about the law, and feelings about legal authorities as a group (p. 386).[10] Tyler found that by far the most powerful predictor of high legitimacy scores was a feeling of trust in the motives of the authority with whom the citizens dealt in their recent contact. Trust in the motives of the authority was a much more powerful predictor of legitimacy than the person’s view of the fairness of the outcome and the favorability of the outcome to that person. This result was uniform across all three racial/ethnic groups. Further research showed that trust in the motives of authorities is strongly influenced by people’s evaluations of the quality of the process and only weakly linked to people’s assessments of outcome fairness and favorability (p. 378). These findings too are identical across racial/ethnic groups. The researcher concluded,
...the findings support a model of process based policing and process based problem solving by the courts. In the case of both types of legal authority, people’s willingness to trust authorities and to defer to their decisions is rooted in people’s judgments about the fairness of the processes through which those authorities exercise their authority. Both the quality of decision making and the quality of treatment are found to influence overall procedural justice judgments and trust in the authorities (Tyler, 2001b: 382).
Tyler (2001a) also analyzed survey responses of a general sample of Chicago residents, looking for what accounts for variation in the public’s confidence in and support for the local police and courts. In this study Tyler focused not on the impact of personal experiences with the police, but rather the impact of respondents’ general assessments of police competence/performance[11] and fair treatment.[12] The analysis showed that both competence and fairness assessments were significant predictors of the public’s general confidence and support for the police. However, the fairness assessment exerted considerably more influence than the competence assessment. These findings were the same for both white and minority respondents.
Finally, Tyler (2001a) conducted additional analyses of the Oakland data described earlier. This mail-return survey focused on the residents of high crime and predominantly minority areas of Oakland. The response rate was low, so considerable caution must be used due to the increased risk of a biased sample. The survey was conducted during a period of aggressive policing to suppress gangs and control gun-related crimes. The survey asked respondents to give an over-all evaluation of job performance and to indicate their willingness to pay taxes to support intensified police activity targeted at street-level drug dealing. The multivariate analysis showed that, once again, evaluations of the quality of police treatment (especially whether the police harass people)[13] exerted far more influence (by a factor of six) on the public’s overall evaluation of police job performance than did people’s sense of police performance in dealing with crime.[14] The quality of treatment factors showed about five times the level of influence that crime performance levels showed. Citizens’ assessment of the quality of treatment also showed the strongest influence on their willingness to pay more taxes, although the strongest influence here was on whether police were trying to control crime. Tyler (2001a:223) concluded, “...support is primarily linked to judgments about how the police treat people, not to whether they are effective in controlling crime.”
The limited number of empirical studies on the impact of police processes versus outcomes all point to the same conclusion: process matters more than outcome in shaping the overall image of American police. Because the studies are few and are in only a handful of communities, it would be premature to regard this pattern as fully validated. A great deal more research needs to be done in a broader range of communities and under a variety of different conditions (e.g., low crime v. high crime, rising crime, v. declining crime v. stable crime rate).
VI. High Visibility Events and the Police Image Regarding Processes
High visibility events – those that are widely and repeatedly covered in the press and other mass media – undoubtedly shape the public’s view of the police. Regrettably, we know very little about how and how much these events influence the police image. Clearly negative events (Rodney King, Abner Louima, Amadou Diallo) undoubtedly tarnish the image of the specific department, but do they also hurt the police image generally, and if so, how much? What about the impact of positive events, such as dramatic rescues, saving lives, and capturing dangerous criminals? We were unable to find studies that answered these questions, but we can present some data that offer intriguing possibilities for further study.
Exhibit 20 tracks citizens’ negative ratings of the honesty and ethics of both police and lawyers from 1977-2000.[15] If the police image is tarnished by widely publicized events, one might expect to see sharp increases in the negative ratings of police following those events. No negative event has been more highly publicized than the Rodney King arrest on March 3, 1991. Other events later in the decade also received intense national coverage. In New York City the Abner Louima and Amadou Diallo use of force cases occurred in 1997 and 1999 respectively in New York City. The Rampart police corruption and abuse of force allegations emerged in late 1999 and 2000.
The chart shows a fairly sharp increase in the percentage of citizens rating police honesty and ethics as “low” or “very low” between 1990 and 1991 (going from 9 to 13 percent).[16] It reached a high point of 14 percent in 1992. This is consistent with the argument – though it does not prove it conclusively – that the King incident had a negative affect on the police image nationally. However, the subsequent fluctuations do not appear consistent with the argument that the Louima, Diallo, and Ramparts events also had a similar impact. In 1993 it declined to 10 percent and then began to rise over the next two years back to 14 percent. During that period there was no negative publicity that received anything like the national attention of the cases considered here. The absence of such highly publicized events should have produced a decline or stability in the public’s negativity about police honesty and ethics. That did not occur until the 1996-1999 period, during which the Louima, Diallo, and Ramparts incidents surfaced. Thus, with the exception of the immediate aftermath of the King incident, the pattern in negative public opinion about police honesty and ethics does not appear to respond to the timing of major, nationally-publicized events that place police integrity in a bad light. Further, the fluctuations in negative views of police are relatively small, indicating that if local events publicized nationally do tarnish the image of police across the nation, they do not have a profound effect.[17] Even though we live in a time of instantaneous global communication, it may be that respondents’ views on ethics and honesty are more influenced by how they see their local police performing. We caution, however, that this analysis is hardly complete. Given the sensitivity of citizen response patterns to question wording, a thorough test requires using a much wider variety of questions, covering, for example, views on the use of excessive force.
We included the data on lawyers to see if their trend matched that of the police. If it did, it would suggest that specific events that reflect badly on the police are not the cause of changes in the police image, but may be due to some larger force that affects both police and lawyers (for example, distrust in professions or government generally). Over this time period, the public’s negativity about lawyers’ honesty and ethics steadily increased through 1995 from 26-45 percent, then declining over the next two years and holding fairly steady at about 40 percent. As the chart shows, the trends for lawyers and police are not very similar, suggesting that at least to some extent, the forces that affect images of the parts of the justice system in which lawyers participate (the courts) are different from those affecting the police image. Thus, while police might well be influenced by shifts in the legitimacy of American political institutions generally and criminal justice institutions more specifically (LaFree, 1998), it would be unwise to treat them as the same (see also Chapter 2, Section VI).
It seems remarkable that public views of police honesty and ethics would not appear to be more responsive to the emergence of nationally publicized police “scandals” in the 1990s, especially in light of the findings we described earlier on the impact of “tabloid justice” news media portrayals of police. That research suggested that exposure to these types of stories reduces public trust and confidence in police. The apparent contradiction may be reconciled if we recognize that citizens may have different levels of exposure to tabloid justice coverage in the news media, and they also may vary in the susceptibility of their general views of police to change that would come from this kind of news (see section on news media portrayals in Chapter 2). Some citizens may pay little attention to negative news or give it little credence – even if they are repeatedly exposed to it. Others may simply interpret the news in such a way that it reinforces their previously held views – such as a citizen who interprets what the press calls police brutality as an instance of the police giving a citizen the justice he deserves. This kind of variation in the habits and mind sets of members of the public may be sufficient to blunt the effects of highly publicized events sufficiently to make it impossible to see strong and consistent effects on general public views in the wake of highly publicized police events.
It would be useful to consider the “positive” side of news coverage of police in the context of the police response to terrorist attacks on the World Trade Center. At the time of this writing, news coverage of police in this event appears to be overwhelmingly positive, portraying rescue efforts as courageous, self-sacrificing, and heroic. What impact has this coverage had on the police image? As yet we have not located a survey that allows us to answer this question comprehensively, but we have found one survey that provides some indication of the impact of this event on one kind of public support for police. A Harris Poll conducted in the wake of the attack examined the public’s support for increasing a broad range of police and FBI powers to increase security, identify, and catch terrorists (Taylor, 2001). In the immediate aftermath, an “overwhelming majority” of the surveyed public supported expansion of police powers, which by itself would suggest a high level of trust and support for American law enforcement. However, the poll also showed that substantial portions of the public expressed concern that these powers might be abused. Between 68-79 percent of respondents had at least “moderate” concern about abuses (roughly half of those expressing “high” concern). On the other hand, 87 percent of respondents were at least “somewhat confident” that these new police powers would be used in a proper way. But only 39 percent of those expressing at least some confidence said they were “very confident” that the powers would not be abused. The researcher concluded that despite these concerns, the public was more willing than previously to accept tough surveillance measures and the risks of abuse. This suggests that the public as a whole conveys a more complex view of police in a time of crisis. It is not so much that large numbers of the public trust the police not to abuse their power, but that they are willing to tolerate the higher risk of abuse in light of the need to deal with the crisis. If this is so, when the crisis has passed, one would expect the public to hold the police accountable for perceived abuses, and this would presumably be reflected in their assessments of police processes and the general image of the police.
Over all, the available evidence suggests that highly publicized events can have significant short-term national impact on the public’s image of police, but it is difficult to discern which type of events will have that impact, and it is even more difficult to track long-term effects. What is most noteworthy about research on this issue is the lack of it.
VII. The Consequences of the Police Image
Throughout this report we have operated under the assumption that the public’s image of the police has important consequences for the police and society. Early in Chapter 2 we argued that the public’s image of the police influences the public’s behaviors that are important to both the police and society. Presumably a positive image encourages the public to support police organizations by voting for political candidates who support police organizations and their leaders. A positive image should also encourage citizens to work directly with the police department by participating in police-sponsored programs and engaging in personal activities that police encourage (e.g., responsible supervision of one’s children, reporting suspicious circumstances, and taking responsibility for keeping the neighborhood attractive) and avoiding activities that police discourage (disruptive and illegal behavior). At the very least, a positive image should produce a contented, complacent public that does not oppose or object to or resist police policies and practices. A highly discontented public may decline to behave in ways that police prefer. They may instead behave in ways that make police work much more challenging. Indeed, they may in extreme circumstances rebel against police authority and even seek to have police leaders replaced and perhaps even make drastic changes to the structure and makeup of the police organization.
We have reviewed some evidence that suggests that a positive image, especially one that pictures police processes positively, reduces the likelihood of citizen resistance to police authority and increases the probability that citizens will obey the law. But these findings are based on what happens in specific police-citizen contacts. We have virtually no evidence on what surveys of the general public can predict about larger patterns of the public’s behavior. For example, at what level of public dissatisfaction with police is there a high risk that the chief will lose his or her job? How negative must the public’s perception be to create an unacceptable risk of massive civil disorder? If one third of the public believes that the police are often – rather than occasionally – brutal, will that foment resistance to police authority that manifests itself in a high risk of assaults on police? Or will it lead to an outcry for placing more constraints on the police use of their coercive authority – such as civilian review boards or federal oversight of local police practices? We found no studies that answered these questions. The answers to those questions are undoubtedly complex, in part because – even when the public’s image is sufficiently strong to motivate them to act, how people act depends in part on what recourse is available to them. A poor person who feels aggrieved may fight the police on the street; a wealthy person is more inclined to hire an attorney to do it in court.
This lack of evidence on the consequences of the police image means that we cannot say how high the positive image must be for a police department to rest easy, nor can we venture how high the negative image must be for the police to be uneasy. We suspect that for the purposes of predicting problem behaviors from the public, tracking the level of the negative image among the public will be a better predictor than tracking the level of the positive image. When public discontent reaches a sufficiently high level among a segment of a community or the larger society, leaders soon arise who work to create a group identity around those grievances, which leads to collective action of one sort or another, be it legal or illegal. While we cannot prove what those thresholds are, we suggest that by the time they reach 20-25 percent of a segment of the public that identifies itself as a victim of the police, a reasonable police chief in a democracy would feel uneasy.
VIII. Conclusions
This chapter has probed the public image of policing processes. We have attempted to place our review of the police within a broader framework of service organizations. All service organizations have an interest in delivering services that satisfy their clients or their constituents. We examined the public image of the police according to seven generic categories on the quality of service: attentiveness, reliability, responsiveness, competence, manners, fairness, and integrity. As we showed throughout the chapter, the amount of research on each of these categories ranges from almost none (such as with competence) to a great deal (such as with fairness). Furthermore, it is apparent that the public has different expectations of the police within each of these categories. For instance, police need not be prompt (responsive) in responding to non-emergency calls, as long as they are attentive in notifying citizens how long it will take to respond, and then respond reliably at that time. Despite the utility of viewing the police among a larger class of service institutions, there remain other elements of service quality which are unique to policing. We reviewed two of those in this chapter: stops and searches and use of force. The public has clear views on both of these topics, and each presumably plays an important role in structuring the public image of the police.
From the many findings we have discussed, we find four points that seem especially noteworthy. First, the majority of the American public appears to hold generally positive views about police processes – that is, the way officers go about the business of policing the public. These vary, of course, according to the aspect of process considered and how the question is asked, and there are some interesting contrasts. The public seems to evaluate the police more positively in terms of their manners toward the public (being polite or friendly) than their fairness. While less than half of the public rates police honesty and ethical standards as “very high” or “high,” the portion of the public rating police “very low” or “low” has fluctuated between only 10-15 percent since 1977. The police rank in the top third of professions on honesty and ethics and well above attorneys. Thus, in a comparative sense, police are doing pretty well in the integrity area, but a substantial portion of the public seems to see considerable room for improvement. The area in which the public seems to show the greatest concern at present is police use of force. Police brutality is seen as present in the communities of almost a third of the public.
Second, race makes a difference when one characterizes the public image of police. Although racial minorities tend to have a positive view of police processes, their views are nonetheless consistently more negative than those of whites. Especially according to the criterion of “fairness,” police suffer a more negative image among minorities. And it is worth noting that on the specific issue of racial profiling, fully three quarters of the entire American public views this as a problem in the United States.
The prior expectations that the public brings to its evaluations seem to have an important, if not always straightforward, affect on the public’s assessments and images of the police. Although some research suggests that the public can become inured to inattentive or coercive policing (thus having little influence on the day-to-day evaluations of police service), other research suggests that meeting or exceeding prior expectations is required if the police are to sustain a successful campaign to improve the public’s evaluations and over all view of the police. Tracking public standards of police use of force over several decades suggests that current levels of public concern about police brutality issue largely from increasingly restrictive standards about when use of physical force is appropriate.
Finally, a small but growing body of research shows that – at least in recent years – the public places greater weight on the processes of policing. Trust in the motives of police to “do the right thing” does more to promote a view of the police as legitimate than even the citizen’s assessment that the outcome was favorable or fair. A sense of fair treatment at the hands of the police is a more powerful influence on public confidence in police than an assessment that the police were competent. And fair and lawful police treatment trumps many times the effects of police performance in reducing crime and making neighborhoods safer.
We must stress that these findings are tentative. In most cases, the number of studies is few. Many draw from local surveys, making it risky to generalize broadly across the great diversity of communities around the nation. And there are a variety of limitations to the data and methods used that make it difficult to rule out alternative hypotheses about what influences citizens’ assessments of police processes and what influences those assessments have on the general image of the police. This, of course, suggests the need for further research, which we will explore in the report’s concluding chapter.
CHAPTER 5
IMPROVING PUBLIC PERCEPTIONS OF THE POLICE IMAGE: THE IMPACT OF COMMUNITY POLICING
An ever-present concern of a competent police leader is how to improve or sustain the public image of his or her agency. We found no studies that evaluated specific departmental efforts to do this. This is a remarkable hole in the evaluation research literature, given the amount of effort and resources devoted by police leaders to grooming their agency’s image (Chermak, 1995; Mastrofski, 2001). One area where there has been some research in recent years is the impact of community policing on the public’s assessments of police and police service. Community policing advocates make public support an essential means, as well as an end unto itself, of police performance. In this chapter we review research on the impact of community policing on the image of police, focusing on the neighborhood development aspects of community policing.
I. The Impact of Community Policing on General, Outcome, and Process Measures
The policy implications of the research reviewed on police outcomes concerning quality of life issues appear straightforward: police should build cooperative relationships with community members (i.e., improve the sense of community among neighbors) to address neighborhood problems, such as public drinking. And, in so doing, the police can improve their overall image and maybe even prevent predatory crimes (Wilson and Kelling, 1982). Unfortunately, recently conducted research suggests it is not that simple.
Using survey data from St. Paul, Minnesota, Ralph Taylor (1997) found a high level of disagreement among individuals from the same neighborhoods concerning neighborhood conditions. In other words, citizens residing in close proximity to one another expressed different levels of fear, risk of victimization, neighborhood attachment, and informal social control (also see Reisig and Parks, 2000:625-6). Taylor’s (1997) findings suggest, then, that survey items tapping into quality of life assessments are largely psychologically-based, emotional reactions to neighborhood ills as opposed to objective measures of actual conditions. How does this affect the overall police image within neighborhoods? Reisig and Parks (2000:626) argue that police strategies to improve neighborhood cohesion and address social disorder will likely only have a modest impact on levels of satisfaction with the police in general.
It is important to note that these research findings do not categorically dismiss the importance of police-citizen collaborative efforts to improve quality of life. But such efforts should be evaluated in a relative sense, neighborhood-by-neighborhood, before and after a particular initiative is implemented. There is research to suggest that police departments in the United States have been successful at improving outcome-oriented elements (e.g., fear of crime) when targeting their efforts toward smaller geographic units, and comparing outcomes with similar units that did not receive similar police attention.
Wesley Skogan’s (1990) book, Disorder and Decline, represents one of the most rigorous assessments of the link between perceived neighborhood conditions, community policing activities, and the public’s image of police. In two particular cases (i.e., Houston and Newark), Skogan (1990:95-122) was able to demonstrate that changes in police strategy (i.e., the adoption of community policing initiatives) was associated with reductions in perceived social disorder and physical decay, as well as improvements in the public’s image of police. For example, in Houston, three community policing initiatives were selected for implementation: police-community stations (i.e., storefront stations), a community organizing response team (CORT) to create neighborhood organizations, and a citizen contact patrol that was designed to increase the number of non-emergency contacts between citizens and the police. Through pre- and post-program survey interviews, Skogan (1990:105) found that the inclusion of a storefront office in the experimental neighborhoods significantly reduced perceptions of physical disorder, and reduced levels of fear. Other encouraging results were also reported for the other two community policing activities. For example, in neighborhoods were CORT concentrated their efforts, perceived physical disorder and social disorder declined, and citizens’ satisfaction with their neighborhood as well as with the police improved. Similar results were found in neighborhoods where citizen contact patrols were implemented. Skogan cautions, however, that the effects of community policing initiatives were not consistent across different groups of citizens. As Skogan (1990:106) succinctly puts it, “In general, those at the bottom of the social ladder were not helped at all.”
In Newark, experimental (i.e., neighborhoods receiving police programs) and comparison neighborhoods were closely matched. Skogan (1990:110) notes that these neighborhoods were all “plagued with physical and social disorder, but at moderate levels” that were believed by police officials to be “treatable” over the study period (i.e., one year). Using before-and-after panel surveys, Skogan (1990:117) found that citizens residing in neighborhoods that received community policing (i.e., community station, citizen contact patrol, and neighborhood police newsletter), reported that physical and social disorder declined after the study period. Additionally, levels of fear decreased, and citizens’ satisfaction with their neighborhood and with the police significantly increased.
Skogan and Hartnett’s (1997) evaluation of Chicago’s Alternative Policing Strategy (CAPS) also provides support for the notion that community policing activities designed to improve citizen quality of life can influence the public’s image of police. The CAPS program was instituted in April, 1993 in five of Chicago’s twenty-five police districts. CAPS brought about a number of changes, including the permanent assignment of patrol officers to specific beats, training in problem-solving strategies, regularly conducting neighborhood meetings, and the formation of citizen advisory committees. The study districts represented a cross-section of Chicago, and were matched with similar districts where CAPS was not instituted. Prior to CAPS, a random sample of citizens residing in the five districts where CAPS was implemented and the five matched districts were surveyed and asked about neighborhood problems, as well as their image of the police. A follow-up survey, which was conducted after CAPS was well underway (i.e., a time period between fourteen and seventeen months) indicated that problems associated with physical decay in three of the five experimental districts, as well as gang and drug problems in two districts, were less frequently reported in citizen surveys (see Hartnett and Skogan, 1999:5). Results from the surveys also revealed that the percentage of surveyed residents responding that their image of the police had become more favorable increased across all race categories. Specifically, when asked how “good a job are the police doing in dealing with the problems that really concern people in your neighborhood,” residents who observed officers involved in community policing activities were more satisfied with the police (Skogan and Hartnett, 1997:205-6). Results from evaluation also showed that citizens who observed community policing activities also expressed more satisfaction with police response to crime, and also reported to feel safer (Skogan and Hartnett, 1997:207-9).
II. Conclusions
The lessons learned from existing evaluation research of community policing activities is that alterations toward a community policing strategy of neighborhood development can influence citizen perceptions of neighborhood problems, as well as improve the overall image of the local police. The research also indicates, however, that such reforms do not provide a quick fix, but entail a long-term commitment on behalf of the police to work with citizens to address neighborhood ills.
CHAPTER 6
CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH
I. Review of Findings and Implications
In this section we summarize in an abbreviated format the major findings discussed in previous chapters. Then we conclude with some general interpretations and observations designed to provide an overarching perspective for our study’s results.
General Police Image
How one measures image makes a difference.
The more general the question, the more positive the response tends to be.
Slight changes in the wording of the question or the response options can make a big difference in how positive the image appears to be.
Questions that measure the “favorableness” of the police image tend to generate more positive responses than questions that ask about “confidence” in the police.
The majority of the public has a substantial degree of confidence in the police as a general institution.
Only a small percentage of polled citizens reports having very little or no confidence in the police in their community.
The proportion of people with confidence in the police can change several percentage points from year to year, but it often changes only 2-3 points per year.
Confidence in police has been declining slowly since 1996 (from 60 to 54 percent).
The trend in respect for the police is that the level of positive views of the police have been declining since the mid-to-late 1960s.
A more positive general image of the police is associated with the following characteristics of the public:
Being older. National samples of high school seniors consistently rate the job police (generally) are doing as substantially lower than do national samples of older persons.
Being of higher wealth or socio-economic status
Living in suburban (as opposed to urban) areas
Being white (as opposed to black)
Having positive attitudes about one’s own neighborhood
One major study of Chicago suggests that there is no difference between blacks and whites when socio-economic disadvantage of the neighborhood is taken into account. That is, blacks’ negativity toward police appears to be due to their concentration in disadvantaged areas.
Negative attitudes about the police by disadvantaged persons appears to be part of a more diffuse alienation from government, law, and the political process generally.
Citizens’ experiences with the police influence their general image of the police.
One study indicates that police courtesy/friendliness toward the citizen in a recent contact with police exerts the most powerful influence on the citizen’s general evaluation of the police. This holds for situations where the citizen’s contact was involuntary (traffic stops) and voluntary (breaking-and-entering complaints).
However, two studies have indicated that people’s prior general views of police have stronger influence on their evaluation of a subsequent specific contact than their evaluation of a specific contact has on subsequent general views of police.
The vast majority of the American public has not had face-to-face contact with a police officer in the previous twelve months.
Most citizens regard the mass media as their prime source of information about crime, and crime news is the context for most mass media accounts of police work.
The implicit message of much crime news is the inability to catch offenders.
There is an increasing trend in the news media to concentrate their coverage on a few sensational cases in a tabloid style of journalism. The net impact of tabloid-style coverage appears to be a decline in confidence in the police.
Entertainment media present images of police that distort the realities of every-day police work. Although more positively presented than attorneys and judges, police are more often than not presented as incompetent rule-breakers.
Since 1993 the police consistently rate among the top three institutions out of thirteen in public confidence (with the military clearly at the top and churches/organized religion and the police vying for second place). Police rate much higher than the rest of the criminal justice system.
Large majorities of adult citizens are satisfied or very satisfied with police service in their neighborhoods.
Although there is variation in satisfaction levels across urban jurisdictions, most fall within the 80-90 percent range.
The majority of school-age children in Cincinnati trust their local police, but a large portion do not, and this distrust is particularly strong among nonwhite students.
Perceptions of the Outcomes of Policing
From the 1980s to the mid-1990s, increasing numbers of the public have offered positive assessments of the quality of police protection.
Confidence in the ability of the police to achieve traditional crime-focused goals appears to be high based on a 1995 survey. 74 percent expressed confidence in the police to protect citizens from crime, 74 percent to solve crime, and 65 percent to prevent crime.
African Americans reported lower ratings than whites and Hispanics.
Residents appear to hold police at least partially responsible for outcomes at the neighborhood level.
There is a relationship between satisfaction levels of police service in the neighborhood and residents’ ratings of crime, disorder, and physical decay in the neighborhood.
Research at the community level suggests that residents believe that citizens share responsibility for controlling crime.
This relationship holds across neighborhoods with different levels of family income, home ownership, and length of occupancy – differing only according to ethnicity.
Hispanics appear more likely to hold police solely responsible for controlling crime.
Public Perceptions of Policing Processes
Police attentiveness to crime victims (providing them counseling on how to undo the negative effects of victimization) has a positive effect on their attitude toward the police.
Citizens’ expectations about how the police will perform affects their evaluation of how they perform in a specific situation.
Positive evaluations are associated with perceived performance that meets or exceeds prior expectations.
A substantial majority of the public rates their police as doing an excellent or pretty good job of being “helpful and friendly.”
A substantial majority of the public express positive attitudes about the fairness of the police, but a significant portion rate them as “only fair or poor.”
African Americans, younger people, singles, and low-income respondents tend to offer less positive evaluations of police fairness. The difference is particularly striking between African Americans and whites.
Racial disparities in assessments of police fairness may be caused in part by indirect exposure to unfair treatment by receiving second-hand accounts from others in their neighborhood.
Almost one in five respondents fear that the police will stop and arrest them when they are completely innocent.
The public image of the honesty and ethical standards of police has fluctuated considerably, but has improved substantially from 1977 to 2000.
– In 2000 the police ranked 10th out of 32 occupations rated on honesty and ethical standards.
Approximately one in ten respondents in a national survey reported that they had been stopped by police because of their racial or ethnic background.
Blacks were seven times more likely to report this than whites.
Most vehicle operators stopped by police felt that they had been stopped legitimately, but there were significant differences by race and gender.
Men and blacks were less likely to feel that their stops were legitimate
Most vehicle operators who are searched by police feel that the search was not legitimate.
Blacks were substantially more likely to view the search as illegitimate than whites.
A study in England found that citizens were more likely to feel fairly treated when officers gave a good reason for the stop.
In 1999 59 percent of the American public perceived racial profiling by the police as “widespread,” and in 2000 75 percent viewed it as a problem in the United States.
Blacks are much more likely than whites to perceive that racial profiling by police is widespread.
Nearly all citizens who experience police force view the police behavior as improper.
· Nearly all citizens view police force as appropriate when a citizen attacks the officer, and the majority approve when a suspect escapes from custody, but few citizens approve when a citizen says vulgar or obscene things to a police officer.
The public’s acceptance of police force is declining over time, especially for suspect escapes and the use of vulgar or obscene language.
Lowered thresholds of what constitutes brutality in the public’s mind may account for some of the significant increase in the public’s perception of brutality.
The citizen’s race is a significant influence on the citizen’s assessment of the quality of the police process.
Hispanics, and especially African Americans, evaluate police less favorably on the use of force, fairness, friendliness, and promptness.
For citizens who have had contact with the police within the previous two years, when the level of satisfaction with that contact is taken into account, race is no longer an significant influence on citizens’ assessments of police friendliness and promptness. However, race remains a significant influence on assessments of use of force and fairness.
A small, but growing number of studies indicates that citizens’ assessments of police processes have a powerful influence on their view of police legitimacy.
Trust in the motives of legal authorities, such as the police, has more impact on police legitimacy than the citizen’s view of the fairness or favorability of the outcome for that person. This is uniform across race/ethnic groups.
Citizens’ assessments of police competence and fairness are both significant predictors of the public’s general confidence in and support for the police, but the fairness assessment is by far the more powerful of the two predictors. This holds for both whites and racial minorities.
A study in Oakland during a time when police were using aggressive tactics to suppress gangs and gun-related crimes showed that public support for the police was influenced far more by how police interact with the public than whether crime was reduced.
The Rodney King incident may have had a nationwide effect on the public’s view of police honesty and integrity in the few years following the event, but the effect was modest and not enduring. Other high visibility events (Louima, Diallo, and Ramparts) showed no readily discernable effect.
In the immediate aftermath of the September 11, 2001 terrorist attacks on New York City and the Pentagon, the overwhelming majoirty of the public was willing to give additional powers to police to conduct surveillance, identify, and apprehend terrorists.
However, substantial proportions of the public were concerned about the possibility of police abuses of these powers.
A growing body of research suggests that how the public feels about the way police treat them affects the public’s behavior (obeying the law and obeying the police).
Most of this research is based on studies of citizen contacts with the police.
Improving the Public Perception of the Police
Community policing may have some modest positive influence on citizens’ satisfaction with the police.
Some evidence suggests that adoption of community policing programs is associated with perceptions of improved quality of neighborhood life and improvement in the image of the police.
Community policing reforms are unlikely to provide a quick fix, but entail a long-term commitment by police to work with citizens to address neighborhood ills.
A Perspective on the Findings
Here we offer a broad perspective to assist in the interpretation of our findings. First, and by now most obviously, the public image of the police is complex, which makes it difficult to make broad-reaching statements. There are many dimensions to the police image, and while the police may score well in one dimension, they may not do so well in another. And any given attribute of the police image (for example, the support the public shows for the police) may be measured in many different ways. Selecting one way, as opposed to another, can have a profound impact on the conclusions drawn, and we caution, parenthetically, that a cynical selection of measures on which police tend to do well to promote a positive image is professionally reprehensible and is likely to be detected by an increasingly sophisticated group of researchers, commentators, and interest groups. The image of the police can fluctuate over time and can also vary among communities at any given time. And the events and forces that affect that image are numerous, and their influence can wax and wane as well. Thus, the image of the police is a slippery thing to measure.
Having noted the complexity of the police image and the hazards of generalization, we find that at the beginning of the 21st century, the image of American police can be construed as positive from one perspective and mixed from another. First, consider the positive interpretation of our results. For most of the indicators of general image, outcomes, and processes, the majority of the American public gives the police a positive score, and relatively few offer a strongly negative rating. Over all, the public’s esteem for the police is sufficiently high that they rank police rank among the more admired institutions and occupations in American society. But consider the “mixed” interpretation. In some cases, the majority of positive responses is not overwhelming, meaning that there are plenty of people who see ample room for improvement. If, as some people argue, effective policing in a democracy requires high levels of trust and confidence in police, then we really need to ask what proportion of the public needs to give the police good ratings to satisfy the contemporary standards of American society. Right now we do not have much scientific evidence on what the thresholds of high and low ratings of police must be. We know that most American businesses would view with alarm a report that even as few as twenty percent of their customers were less than pleased with their product or service. But the analogy does not go too far for police, since most police agencies are not in competition for the public’s business with other agencies, and many police “clients” would rather not receive any attention from the police. Yet any competent, experienced street officer knows that he or she cannot be effective without the confidence and trust of the public, and that each alienated citizen makes daily police work more difficult, more challenging.
We conclude that complacency in the face of our results is a reaction that carries high risks. Indeed, many current public figures, including some law enforcement leaders, argue that American police are now experiencing a crisis in confidence. Some researchers note that the current legitimacy crisis for police is not recent, but rather has been a long time coming as part of a broad and long-term erosion in the public’s trust in all of its great institutions of social stability and control, such as schools, families, and government generally (LaFree, 1998). Racial profiling, brutality, and corruption received a great deal of attention in the 1990s, and these issues seem to dominate the news and public debate early in the 21st century, when crime rates continue to decline. We should keep in mind, however, that the 1990s were preceded by two decades of public policy obsession with the seeming inability of the police to curb soaring crime rates. It is worth noting that there has hardly been a time in the history of modern police forces in America when the police were not characterized as in some form of crisis (Fogelson, 1977).
In this historical context, perhaps we should not be too surprised to find reason for both alarm and celebration in the state of the police image in America. There are many indicators that American police are among the most trusted and admired institutions of contemporary society, while there are also many indicators that the American public – especially the young and disadvantaged members of that public – are wary of the police and see plenty of room for improvement. The seemingly high overall levels of confidence and satisfaction with American police – even among the disadvantaged in our society – suggest that the police continue to enjoy widespread legitimacy. However, police leaders should be careful not to succumb to complacency just because their agencies score well over all in national and local public opinion surveys. Police work is done predominantly to and for the disadvantaged – those segments of the public who have the greatest alienation from the police (Bittner, 1970). That the majority of these people express confidence in their police should not hide the other reality that large proportions of these groups do not share that view. And these are precisely the people whose cooperation and good will are required if community policing is to fulfill its promise of increasing effective crime and disorder control. To put it simply, even relatively low levels of public dissatisfaction with the police are problematic for a democracy when they are concentrated among groups who develop a self-identity as “victims” of policing.
Where public dissatisfaction and negativity about the police image exist, the reasons for that perspective are undoubtedly complex. There seem to be at least three sources, about which we know relatively little. First is the state of objective police performance. Dissatisfaction and negativity may exist in part because the police simply fail to perform well at a given time or place. If the police lay claim to crime control as their mandate, and they argue that their efforts can and do reduce crime, then it is logical that the public will hold them liable when and where crime fails to go down or does not decline as far or fast as desired. If lawfulness, fairness, professionalism, integrity, and service are keys to the legitimacy of police agencies, the police may simply fail to live up to those standards often enough to cast doubt in the public mind.
A second cause of public dissatisfaction with police may be thought of as the way the police are presented to the public through the press and entertainment media. Such presentations can never be comprehensive nor precisely accurate, because those are not the purposes of the mass media. They tend to feature what is new, unique, and entertaining. We note that the police are not without considerable influence in news stories about them and their work (Chermak, 1995).
A third force behind negative views of the police may be heightened expectations and standards that the public brings to their evaluations of the police. It is entirely possible that over the latter half of the 20th century police performance in the outcomes and processes of policing have improved, but the public’s standards and expectations about police performance have increased even more rapidly, and this may be especially so for the disadvantaged groups. Rising expectations are hard to meet, especially if they rise fastest among precisely those people who are worst off.
The police have advanced and sustained their legitimacy in America by embracing two maxims: egalitarian service according to law and efficacy and efficiency by means of science and technological innovation. These are, of course, ideals that resonate with larger cultural trends, and which over the last 150 years have been accelerated for police by a public willing to rely increasingly on government to improve the quality of life. The American public has become accustomed to increasingly higher standards of health, medicine, transportation, and communications – to name but a few of the domains affected tremendously by scientific and technological advance. By tying their legitimacy to science and technology, police are implicitly promising the same kinds of advances in reducing crime, increasing order, neighborhood quality of life, and community cohesion. By relying heavily on laws, rules, education, and training to advance their occupational status – especially with regard to principles of lawful egalitarianism, police have set high standards for the processes of policing. And the community policing movement adds to this expectations of responsiveness and participation. Reform movements, such as community policing, thus cut both ways – increasing the public’s faith in the motivations and capabilities of their police, but also increasing their expectations in the results. A “problem-oriented policing” promises to solve a wide range of problems, not just to process paper about those problems (Mastrofski and Ritti, 1999). Such promises have consequences for what citizens come to expect police can accomplish. And, if the police do improve their performance in an objective sense, there is always the risk of the public “raising the bar”after becoming accustomed to higher levels of performance, driven by a “what-have-you-done-for-me-lately?” attitude.
Any attempt to explain variations and fluctuations in the public support of the police should take all three possible explanations into account: objective police performance, mass media presentations of the police, and public expectations of the police. The studies reviewed for this report do not attempt to test these propositions, but that should be an important issue for future research, a topic of our next section.
II. Recommendations for Future Research
In the report’s concluding section we identify priority issues and questions for future research and then propose an agenda for collecting data that would help answer those questions.
Research Questions
The following represent questions that are unanswered by the available research. They are important topics for future research on the public’s image of the police.
Measurement Issues. We have demonstrated that a particular aspect of the police image can be measured in many different ways, and that the particular measure chosen to represent a given concept (such as support for police) can have profound consequences for the conclusions drawn. Much more research should be done to identify the best measures to use for particular purposes. Measurement issues include not only question wording, but also question ordering and how the survey itself is framed for the respondents.
Explaining Variation in the Police Image. What is the relative importance of the following in accounting for variation among members of the general public in their view of the police: the personal experiences of those citizens, what they learn second-hand from friends and acquaintances, and what they learn from the mass media? What is the role of the citizen’s prior view and expectation regarding the police? How readily is this changed, and what kinds of experiences and mass media presentations are most likely to change it?
What is the importance of highly publicized events on the police image? Does the public pay as much attention to positive portrayals (e.g., acts of police heroism) as it does negative ones (e.g., police corruption and brutality)? How long do these effects last? How much influence do highly publicized events occurring in one community affect the image of police in other communities?
Most survey research has focused on explaining differences among individual members of the general public, comparing one type of person to another, using such characteristics as race, sex, and age. While this is important, other levels of analysis have received insufficient attention. We need cross-jurisdiction studies that compare the police image in a diverse set of communities. Is the level of public satisfaction with the police fairly uniform – as found in the 12-city study reported by Smith and colleagues (1999), or is there really much greater diversity? What characteristics distinguish departments that receive high evaluations from those that receive low ones? For example, do police departments with higher proportions of minority officers receive higher evaluations from minority residents than departments with lower proportions – that is, does the racial representativeness of the police work force make a difference? Do departments with more college-educated officers receive more favorable ratings from the public than those with fewer college-educated officers? Do departments that make more arrests receive higher evaluations than departments that make fewer arrests?
We also need more studies that compare citizen support for police from neighborhood to neighborhood. Some departments may do a much better job of sustaining support from neighborhoods that are traditionally alienated from the police. If so, is this due to outreach efforts that create a greater potential for neighborhood collective action (e.g., police agencies creating and supporting neighborhood organizations), or is it more a matter of delivering higher quality service to individuals within the neighborhood?
The Influence of Context on the Police Image. Most of the research on the police image fails to take into account the larger social, economic, demographic, and cultural context in which the survey data were obtained. For example, national surveys that may produce reliable representations of how the nation as a whole views their police fail to capture the great diversity among communities in the United States. We know, for example, that suburban respondents report higher levels of support for police than do urban residents, but we do not know the pattern for communities according to their level of crime and disorder, the nature of their legal system, their economic vitality, and so on. Are the patterns of influence on public support for the police the same in each of these different community environments? Is Mayberry the same as Minneapolis? When researchers have more knowledge of the effects of community context, their research will be more useful to police administrators who want to fashion programs that will be most effective for their jurisdictions.
Of course, context can change over time, and some contextual effects may take decades to distinguish. For example, we still do not know the implications for police legitimacy in times of rising crime (and rising fear of crime) compared to times of declining crime (and declining fear of crime). Our evidence on the priority citizens give to police processes over outcomes is based almost entirely on research done in the 1990s during a time when crime was declining. Would this hold during a time when crime rates are skyrocketing? And what about the influence of external threats (economic or national security)? Do citizens change their priorities or standards for evaluating the police? During a time of greatly heightened concern about the nation’s vulnerability to terrorism, will the public downgrade its priority for how the police treat them and upgrade its concern about the police capacity to produce safe and secure communities?
Police Efforts to Shape their Image. We know that police agencies exert much effort to shape their images, but we need to know a lot more about what kinds of strategies are most effective. Do public image campaigns that rely on mass communications have a greater impact than efforts to transform the day-to-day experiences of citizens by changing officer practices? Are public image campaigns most effective at changing citizens’ beliefs about what police are doing? Or are they more effective in changing citizens’ standards and priorities for police performance? What influence do the courts and high visibility public officials have in shaping public expectations (such as due process standards and expectations about crime control)?
Police Image v. Objective Measures of Police Performance. The image of the police is determined by learning the subjective views of the public. Although these subjective measures are important to police agencies for a variety of reasons, police leaders rely heavily on “objective” measures of police performance – indicators that attempt to determine how well police are doing independent of the viewpoint of members of the public (e.g., crime and victimization rates, response times, cases cleared, complaints against police sustained). Competent police management calls for a balanced system of assessment that includes both subjective and objective measures (Bayley, 1994; Mastrofski and Wadman, 1991). But it is not enough just to use both subjective and objective measures, police managers need to know what to make of those occasions when the results differ between the two types of measures. Thus, we need research that tells us where objective and subjective measures are congruent, and where they are not, we need to know the reason for those differences. For example, if a police department is viewed by a substantial portion of its clientele as engaged in brutality while at the same time the department’s records show use of force to be at an all-time low and also low in comparison to other comparable police agencies, this calls for an explanation. There may be flaws in one or both measures, or they may be really measuring different things. Researchers should attempt to answer these questions, which are of critical importance to police executives.
Implications of the Police Image. What are the consequences of the police image for the public’s actions? Public opinion about candidates for elective office only matter insomuch as they influence voting behavior. To what extent does the public’s view of the police influence such actions as reporting crime and other forms of cooperating with police? Are citizens who are more positively disposed to the police also more likely to participate in collective action designed to improve neighborhood quality of life? Obeying the law? Voting and other forms of political expression in support or opposition to the police leadership’s agenda? What are the consequences of the police image for the tenure of the police leadership? How often do police chiefs lose their jobs in a climate of declining public confidence in the police? How long or how far can public confidence decline before chiefs lose their jobs? Under what circumstances do negative images of the police enable chiefs to launch successful reform campaigns?
Agenda for Future Data Collection
Based on the important research questions that need to be answered, we suggest a program that the membership of the IACP could pursue to advance knowledge. Although the proposed program would not address all important issues, it would provide for an efficient way for individual police agencies to answer questions about their own image while at the same time contributing to broader knowledge about policing across the United States.
Currently survey research on the public’s views of the police is conducted on a haphazard basis. Some national polling organizations periodically question samples of the American public about certain aspects of police, but there is not always consistency in the survey questions and methodology across different polls, and the surveys are nearly always designed to capture a sample representative of the nation as a whole, rather than specific jurisdictions. The national samples drawn do not have enough respondents in any jurisdiction to make reliable estimates of public opinion within jurisdictions. Also, surveys conducted for a national sample – or even a particular jurisdiction – will often fail to focus on issues that are of great importance to other jurisdictions. Inasmuch as policing is overwhelmingly a local enterprise, it makes sense to develop a program of survey research that provides local jurisdictions with important information about public opinions and perceptions that are relevant to each community.
Just as the Uniform Crime Reports provide a basis for cross-jurisdiction comparison, so could a national system of uniform public opinion surveys on policing provide a basis for learning about the public’s views and assessments of police agency performance. Aside from the type of data collected, the major difference with a national program of survey research on policing from the Uniform Crime Reports would be to build in sufficient flexibility so that local jurisdictions could also ask questions that focus on local concerns that may not be shared nationwide. This calls for a program that melds uniformity and flexibility, one that mirrors the federal system of government in which American police operate. Just as the IACP played a key role in the development of the UCRs in the 20th century, it could also play a key role in the development of a program of national public opinion surveys in the 21st century. A provisional title for such a program (for the purposes of our discussion) is the Uniform Public Opinion Poll on Policing (UPOPP).
In developing UPOPP, the IACP would serve four important program functions. First, it would commission the development of a standard survey instrument and methodology. Second, it would underwrite the field testing and revision of this instrument and methodology. Third, it would solicit participation from its members. Fourth, it would select and monitor a research organization that would conduct a number of activities. The research organization would be responsible for keeping abreast of studies and survey research developments relevant to the UPOPP mission. It would assist participating members in developing survey items that would deal with local issues of concern that were not dealt with in the standardized part of the instrument that all jurisdictions would use. The research firm would also provide limited technical assistance to participating police agencies regarding survey methodology, such as sampling, data collection, and contract arrangements with local survey research firms for collecting data. The participating police agencies would be responsible for conducting the local surveys with the advice and assistance of the UPOPP research organization. The research organization would receive copies of the instruments and survey data collected by participating police organizations and would archive those data. The UPOPP research organization would establish and enforce technical standards of computerized data entry formats so that data across participating agencies could be readily merged and archived. And the research organization would issue an annual report that summarized analysis from the data archive for the current year, as well as showing comparisons and trends for previous years. This report would be made available to the IACP to release and distribute.
The proposed program raises a host of questions and issues, not all of which can be answered here, but we will at least raise the most important ones to facilitate future discussion.
Developing the Standardized Research Instrument. Uniformity in survey questions is essential for comparison of jurisdictions at any given time and to track their progress over time. Consequently, the standardized section of the survey instrument must be carefully constructed so that it is relevant to the widest possible range of local police agencies at any given time and over many years (so that trends can be meaningfully assessed). Once established, certain items in the standardized section should rarely, if ever, be changed – so that trends in certain public views can be tracked reliably over time. However, some parts of the standardized section might be changed annually to facilitate obtaining nation-wide measures of public opinion on new and emerging issues. And some standardized items might not be asked every year, but only every few years to provide an efficient method of tracking over time.
The research organization should have a high level of expertise on the substantive issues in policing and up-to-date- knowledge of the state of relevant research literature. However, the research organization should work closely with a special committee of IACP members to establish the initial standardized portion of the instrument and to monitor it for necessary improvements and changes. This would provide for a high level of technical excellence while ensuring that UPOPP is responsive to the needs of the police profession.
Developing the Local-Option Portion of the Instrument. Participating members would have the option of supplementing the standardized part of the survey with questions unique to their own jurisdiction. The UPOPP research organization would maintain a “bank” of pretested survey items from which a participating agency might choose to address one or more issues of local concern. If there were no items in the data bank appropriate to the participating agency’s concerns, the UPOPP research organization would assist in the development of questions that would meet the agency’s needs.
Participation. We envision a program of voluntary participation by IACP members. We anticipate that interest would be widespread among police agencies. Many around the country already conduct their own survey research on an occasional, and even routine, basis. Small police organizations may find it difficult to plan and pay for an annual survey, so it may make sense for several of them in a given area to work together to develop and conduct a common survey.
Assistance to Participating Agencies. Some police agencies already have extensive experience with survey research of this sort. They have either conducted such surveys themselves or have worked with research organizations to design and conduct surveys and analyze data. Many, however, have limited or no experience and require support. We think it impractical to have a single research organization actually conduct all of the surveys for all participating agencies. The “production” tasks associated with actually conducting the surveys can be performed by the local police organization – or as will be likely in most cases – contracted out to local survey research organizations (universities, survey research firms, and think tanks). However, many police organizations will need preliminary assistance in planning the survey: survey design, sampling, and mode of data collection. The UPOPP research organization would provide this sort of planning assistance and would also offer advice on selecting and contracting with a local survey research organization to collect the data. As necessary, the UPOPP research organization would also provide advice on how best to analyze and interpret data, although it would not be responsible for actually conducting the data analysis for the participating organization. Except in rare circumstances, the UPOPP research organization would provide this assistance via telephone, fax, and internet (both email and web site) – to avoid the high cost of long-distance travel.
The UPOPP Annual Report. The UPOPP research organization would publish annually a report that does the following:
Reviews important research findings on topics relevant to the police image during the previous year.
Provides an overview of national (and international, if available) patterns and trends in public views of police – based on survey data archived by participating members.
Provides data analysis on one or more focused issues. These issues might include both high-visibility issues with interest that is of national or international scope. They might also include data analysis on emerging issues or matters that are more localized in scope (drawing on the local-option portion of the survey).
One important issue to resolve in the future is whether to identify individual jurisdictions in the report, as do the Uniform Crime Reports. The advantages to this are increased ability for participating agencies and outside researchers to learn more about the causes and consequences of the police image. Tremendous advances in our knowledge of crime and the effectiveness of crime control methods have come from identifying results by jurisdiction in the Uniform Crime Reports. However, some agencies may be understandably wary of such visibility at the beginning of such a data system. They may worry about the potential for abuse and misinterpretation of results. Although we are inclined to recommend listing survey results by jurisdiction, we recognize that this is an issue that the IACP must resolve with its members.
Archiving the Data. Although individual agencies will reap tremendous benefits by knowing the results of surveys in their own jurisdictions, far greater advantages can be obtained when it is possible to compare results across jurisdictions. To this end, the UPOPP research firm would serve as the archive for UPOPP data. Participating agencies would be required to submit a copy of the research instrument and a description of research methodology (e.g., sampling design) to the UPOPP research organization upon completion of the survey. The data would be submitted in a standardized format to facilitate archiving. The UPOPP research organization would analyze these data to issue its annual report, and it would fulfill data analysis requests from participating members. For example, if a department wanted to know how it compared to similar agencies on the public’s view of its ability to reduce gang violence, UPOPP would conduct the analysis for the agency.
As with the annual report, an important issue to resolve is whether to release the archived data to other research organizations or data archives, such as the ICPSR at the University of Michigan, which houses many data sets on crime and justice issues. This would require a decision by the IACP.
Funding the UPOPP. As proposed, the largest share of the costs for this program are assigned to the participating local agencies, who would be required to fund the costs of conducting the surveys. The IACP would be responsible for contracting with the UPOPP research organization, which will provide important, but much less costly services that can be effectively and efficiently centralized in one organization. Nonetheless, such costs can be substantial, so the IACP may wish to consider charging participating members a fee for participating. This fee (which could be adjusted according to the amount of UPOPP assistance required) would cover the costs of the UPOPP research organization. Given the potential benefit of UPOPP to the improvement of policing and the advancement of the profession, IACP might wish to consider seeking external support for this program – especially during a start-up and early stages.
REFERENCES
Albrecht, S. and M. Green. 1977. “Attitudes Toward the Police and the Larger Attitude Complex: Implications for Police-Community Relationships.” Criminology: An Interdiscipilnary Journal 15:67-86.
Alpert, Geoffrey P. and Mark H. Moore. 1993. “Measuring Police Performance in the New Paradigm of Policing.” Pp. 109-140 in Performance Measures for the Criminal Justice System, edited by the Bureau of Justice Statistics. Washington, D.C.: U.S. Department of Justice. Apple, Nancy. and David J. O 'Brien. 1983. "Neighborhood Racial Composition and Residents ' Evaluation of Police Performance." Journal of Police Science and Administration 11:76-83.
Bayley, David H. 1994. Police for the Future. New York: Oxford University Press.
Bayley, David H. and Harold Mendelsohn. 1969. Minorities and the Police. New York: Free Press.
Benson, P. 1981. "Political Alienation and Public Satisfaction with Police Service." Pacific Sociological Review 24:45-64.
Bittner, Egon. 1970. The Functions of the Police in Modern Society. Bethesda, MD: National Institute of Mental Health.
Bradley, Russell. 1998. Public Expectations and Perceptions of Policing. Police Research Series Paper 96, London: Home Office.
Brandl, S.G., J. Frank, J. Wooldredge, R.C. Watkins. 1997. "On the Measurement of Public Support for the Police: A Research Note." Policing: An International Journal of Police Strategies & Management 20:473-480.
Brandl, S.G., J. Frank, R.E. Worden, and T.S. Bynum. 1994. "Global and Specific Attitudes toward the Police: Disentangling the Relationship." Justice Quarterly 11:119-134.
Brandl, S. and Horvath, F. 1991. "Crime Victim Evaluation of Police Investigative Performance." Journal of Criminal Justice 19:293-305.
Bratton, William. 1998. Turnaround: How America’s Top Cop Reversed the Crime Epidemic. New York: Random House.
Brown, Karin, and Philip B. Coulter (1983). "Subjective and Objective Measures of Police Service Delivery." Public Administration Review. January-February: 50-58.
Bureau of Justice Statistics and the Office of Community Oriented Policing Services. 1999. Criminal Victimization and Perceptions of Community Policing in 12 Cities, 1998. NCJ-173940.
Cao, Liqun, James Frank, and Francis T. Cullen. 1996. “Race, Community Context, and Confidence in the Police.” American Journal of Police 15(1):3-22.
Chavis, David M. and Abraham Wandersman. 1990. “Sense of Community in the Urban Environment: A Catalyst for Participation and Community Development.” American Journal of Community Psychology 18:55-81.
Chermak, Steven. 1995. “Image Control: How Police Affect the Presentation of Crime News.” American Journal of Police 14:21-44.
Dean, Deby. 1980. “Citizen Ratings of the Police: The Difference Contact Makes.” Law and Policy Quarterly 2:445-71.
Decker, Scott. 1981. "Citizen Attitudes toward the Police: A Review of Past Findings and Suggestions for Future Policy." Journal of Police Science and Administration 9:80-87.
Dunham, Roger G. and Geoffrey P. Alpert. 1988. “Neighborhood Differences in Attitudes Toward the Police: Evidence for a Mixed-Strategy Model of Policing in a Multi-Ethnic Setting.” The Journal of Criminal Law & Criminology 79(2):504-523.
Erez, Edna (1984). "Self Defined 'Desert ' and Citizens ' Assessment of the Police." The Journal of Criminal Law and Criminology 75(4): 1276-1299.
Flanagan, T. J. and M. S. Vaughn. 1995. “Public Opinion About Police Abuse of Force.” In W. A. Geller and Hans Toch, eds., And Justice for All: Understanding and Controlling Police Abuse of Force, pp. 113-131. Washington, DC: Police Executive Research Forum.
Fogelson, Robert M. 1977. Big City Police. Cambridge: Harvard University Press.
Fox, Richard L. and Robert W. Van Sickel. 2001. Tabloid Justice: Criminal Justice in an Age of Media Frenzy. Boulder, CO: Lynne Rienner Publishers.
Garofalo, James. 1977. Public Opinion about Crime: The Attitudes of Victims and Nonvictims in Selected Cities. Washington, DC: Government Printing Office.
Gibson, J. L. 1989. “Understanding of Justice: Institutional Legitimacy, Procedural Justice, and Political Tolerance.” Law and Society Review 23.
Gilmore, Susan. 1999. “WTO Poll: Police Did Well; Mayor Fumbled.” December 19: seattletimes.nwsource.com/news/local/html98/poll_19991219html.
Goldstein, Herman. 1977. Policing a Free Society. Cambridge: Ballinger.
Graber, Doris. 1980. Crime News and the Public. New York: Praeger.
Hartnett, Susan M. and Wesley G. Skogan. 1999. “Community Policing: Chicago’s Experience.” National Institute of Justice Journal (April):3-11.
Hindelang, M.J. 1974. "Public Opinion Regarding Crime, Criminal Justice, and Related Topics." Journal of Research in Crime and Delinquency 11:101-116.
Homant, Robert J. 1994. "Citizen Preferences and Perceptions Concerning Police Pursuit Policies." Journal of Criminal Justice 22:425-435.
- See more at: http://www.theiacp.org/PoliceServices/ProfessionalAssistance/ThePublicImageofthePolice/tabid/198/Default.aspx#ch1