Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
2 Criminal Justice Research on Police T he research evaluated in this report was generated during a relatively short but extraordinarily productive period of intellectual effort. Be- fore publication of the report of the President's Commission on Law Enforcement and the Administration of Justice, The Challenge of Crime in a Free Society (Black and Reiss, 1967a), there was hardly any scientific research on the police. Today there is so much that scholars and police find it difficult to keep up, let alone to evaluate its qualitative merits and practi- cal utility. This chapter describes the scientific enterprise of police research that has developed since 1967--its scale, substantive coverage, methods, and the auspices under which it has been conducted. Special attention is paid to the body of research funded by the 1994 crime act in the final section. It is not an intellectual history of the field or a detailed portrait of its develop- ment, but rather a description of its scale and scope.1 The research described here does not exhaust all that has been written about the police. We focused on studies that follow the scientific method, were subjected to peer review, and were publicly disseminated. Our assess- 1Important examples of this kind of intellectual history include L.W. Sherman (1973) "Soci- ology and Social Reform of the American Police 1950-1973" in Journal of Police Science and Administration; Bittner and Rumbaut (1979) "Changing Conceptions of the Police Role-A Sociological Review" in Crime and Justice: An Annual Review,Volume 1, edited by M. Tonry and N. Morris (Chicago: University of Chicago Press); R. Reiner (1994) "Policing and the Police" in M. Maguire, R. Morgan, and R. Reiner (eds.), The Oxford Handbook of Criminol- ogy. London and New York: Oxford University Press, 705-772. 20
CRIMINAL JUSTICE RESEARCH ON POLICE 21 ment does not include studies done by police departments for in-house pur- poses, by management consultants on contract, by investigating commis- sions, or by experts involved in civil and criminal proceedings. SCALE OF POLICE RESEARCH The publication of the President's Commission on Law Enforcement and the Administration of Justice was a watershed in the development of research on the police. Before 1967 only a handful of scholarly books had been published, notably American Police Systems by Raymond Fosdick (1920), Police Systems in the United States by Bruce Smith (1949), and Violence and the Police by William Westley (1953).2 Today, one of the most complete collections of books on the police in the United States, North- western University's Transportation Library, lists 2,934 books on the po- lice published since 1967. In the history of Dissertation Abstracts Interna- tional, which lists Ph.D. dissertations going back to 1861, over 1,300 with the word "police" in the title have been written: 69 before 1967 and just over 1,250 after 1967. Today the National Criminal Justice Reference Service of the U.S. De- partment of Justice lists approximately 31,000 references under the heading "police and law enforcement." These documents, which constitute 20 per- cent of its total holdings, include federal, state, and local government re- ports, books, journal articles, and published and unpublished research re- ports. The cascade of peer-reviewed research on the police since 1967 has been part of the development of criminal justice studies generally. Of the 12 most highly regarded journals in criminology and police studies published today, only 3 existed before 1967.3 Altogether these journals have pub- lished over 6,900 articles dealing with the police and law enforcement, the larger part after 1967.4 Sociological Abstracts, which covers 2,500 journals and periodicals, lists 6,929 citations to material published between 1963 and 2001. 2Westley's book was originally entitled The Police: A Sociological Study of Law, Custom, and Morality. It was republished by M.I.T. Press as Violence and the Police in 1970. 3These are Crime and Delinquency (1960), Criminology (1963, known before 1970 as Criminologica), and the Journal of Research on Crime and Delinquency (1964). Those pub- lished after 1967 are Criminal Justice History (1980), the Journal of Crime and Justice (1981), the Journal of Criminal Justice (1973), Justice Quarterly (1984), the Journal of Quantitative Criminology (1985), the Journal of Police Science and Administration (1973-1990), Police and Society (1990), Police Studies (1978), and the American Journal of Police (1981). 4The figure is actually somewhat larger because indexes of titles available for a keyword search are not available for all the journals for the entire post-1967 period.
22 FAIRNESS AND EFFECTIVENESS IN POLICING Police research is now published overwhelmingly in specialty criminal justice journals rather than the mainline social science journals. For ex- ample, in the 33 years before 1967, the American Sociological Review, the American Journal of Sociology, and the American Political Science Review published 12 articles having to do with "police." In the 33 years following, they published 18. During their entire publishing histories, these journals have published only 33 articles with "police" in the title. In addition to criminal justice journals, police research is also reported in three periodicals intended for police professionals: the FBI Law Enforce- ment Bulletin, Police Chief Magazine, and Law Enforcement News. All but the latter predate the creation of the academic journals. Their articles fre- quently summarize reports of peer-reviewed research. Assuming that most of the people who do scientific research on the police probably belong to either the American Society of Criminology or the Academy of Criminal Justice Sciences, they are estimated to number some- where between 300 and 400. In 2000, 307 members of the American Society of Criminology (11 percent of the membership) belonged to its special- interest police section, while 187 members of the Academy of Criminal Justice Sciences identified themselves as police researchers (5 percent of the membership) (personal communications with these associations). Although the Academy of Criminal Justice Sciences is generally regarded as the more practitioner-oriented of the two, a larger proportion of members of the American Society of Criminology identified themselves as police researchers. In sum, police research has become a substantial industry in 35 years, with a dedicated core of scholars, a large body of published work, several specialized journals, many publicly accessible data sets, and regular profes- sional meetings. SUBJECT MATTER OF POLICE RESEARCH The scientific study of the police in the United States arose out of con- cern with the fairness of police actions, especially discriminatory treatment of black Americans. It reflected growing sensitivity to the unequal treat- ment of minorities generally, whether formally through discriminatory laws--voting, education, employment, and housing--or informally through the pervasive exercise of prejudice, especially by government officials. Beginning in the late 1950s, the American Bar Foundation undertook a series of studies designed to describe the way in which the institutions of criminal justice actually worked, going beyond the traditional descriptions of their formal organization or legal empowerment. The key discovery was that officials, among them the police, possessed enormous discretion in the way they applied the law (LaFave, 1965; Davis, 1969). Two classic studies confirmed this insight: Justice Without Trial by Jerome Skolnick (1966),
CRIMINAL JUSTICE RESEARCH ON POLICE 23 about criminal investigations, and The Functions of the Police in Modern Society by Egon Bittner (1970), about police patrolling. This discovery also led to the first attempt to describe police behavior through systematic ob- servation of police encounters with the public (Reiss, 1971). The data, which were collected in Boston, MA; Washington, DC; and Chicago, IL, were used to explore several hypotheses about factors shaping police actions. Systematic observation was extended by the Police Services Study in 1976 to Rochester, NY; St. Louis, MO; St. Petersburg and Dade County, FL. Research on police behavior and its determinants declined after the 1970s. With the exception of two smaller scale studies in Denver and New York, writing about the determinants of police behavior relied on informa- tion collected in the 1960s and 1970s (Bayley and Bittner, 1984; Bayley and Garofalo, 1989). This situation persisted until the early 1990s, when the National Institute of Justice (NIJ) sponsored a major new effort at system- atic observation in Indianapolis, IN, and St. Petersburg, FL (the Policing Neighborhoods Project). The 1960s also saw the beginning of scientific surveys of public as well as police opinion about police matters (Biderman et al., 1967; Bayley and Mendelson, 1969; Kerner Commission, 1968; Jacob, 1971; Smith and Hawkins, 1973). These studies have remained a staple of police research ever since. They also provided a tool for broadening the criteria used to judge police performance, from a narrow focus on crime and arrest rates to consideration of public satisfaction, respect, legitimacy, and perceptions of bias. These criteria are now explicitly included in lists of police perfor- mance indicators. Neglecting the precedent set by Skolnick, studies of the exercise of po- lice discretion have focused almost exclusively on patrol. Only a handful of scholars have studied criminal investigation, and most of that has been, like the work on patrol, during the 1970s (Greenwood et al., 1977; Wilson, 1977; Manning, 1977; Eck, 1983). Another precedent of the 1960s that has been neglected was James Q. Wilson's attempt to account for variations in police behavior at the organizational level (1968). In Varieties of Police Behavior, he described three styles of police interaction with communities and explored their contextual determinants. Although scholars continually deplore the absence of research about the behavior of police agencies as a whole, most analysis continues to focus on the behavior of individual police officers. In addition to discretion and its determinants, police research focused initially on evaluating the effectiveness of the standard strategies of polic- ing, notably motorized and foot patrolling and rapid response to calls for service (Kelling et al., 1974; Trojanowicz, 1986). In general, this research found that police were not getting the results expected in terms of crime prevention and public satisfaction. Curiously, although these evaluations
24 FAIRNESS AND EFFECTIVENESS IN POLICING cast serious doubt on the efficacy of contemporary police practices and were, as a result, hotly contested by the police, very few replications were undertaken. The research findings were accepted as true, despite method- ological criticisms by other scholars, and became the basis for a profound rethinking of the police role in crime control and prevention in the 1980s. The new strategy of policing that emerged in the 1980s is called com- munity-oriented policing. Very closely related to it is problem-oriented po- licing. Although there are important philosophical differences between the two, their operational practices tend to overlap. The lesson that the propo- nents of community-oriented policing drew from the research on the effi- cacy of customary police practices was that the police were unable to con- trol crime on their own without the cooperation and support of the public. This could involve reporting crime, identifying criminal suspects, calling attention to conditions that breed disorder and crime, and taking self-pro- tecting actions. The other major intellectual inputs to the reform movement were Herman Goldstein's (1979) insight that police needed to involve com- munities in cooperative problem solving and James Q. Wilson's and George Kelling's (1982) proposal, captured in the metaphor of broken windows, that police should help communities create a crime-deterrent environment by minimizing public disorder. With the advent of community policing in the 1980s, police scholarship underwent a dramatic change and became explicitly prescriptive. Scholars became advocates as well as analysts. In the 1960s, by contrast, police schol- ars undertook research on police behavior, notably the exercise of discre- tion, in order to serve the implicit agenda of providing information about its fairness, but they were reluctant to prescribe programs of remedy. Fur- thermore, the advocacy of the 1980s focused on issues of community safety, not on the exercise of police powers in individual encounters with the pub- lic. A vast literature grew up that outlined, elaborated, and encouraged the philosophy of community policing. Training seminars were held and in- structional manuals prepared. Although a few pilot projects were under- taken, only a few systematic attempts were made to test the efficacy of community policing. Indeed, police as well as scholars lamented through- out the 1990s that both the extent and the effectiveness of community po- licing remained largely unknown (Rosenbaum, 1994; Roth et al., 2000). The community policing movement reflected not only the willingness of scholars to work with, rather than simply on, the police, but also the acceptance by police of the value of such collaboration. Viewed with in- tense suspicion and often outright rejection in the 1960s, research by pro- fessional scholars was now viewed as a tool of progressive management and innovation. Along with efforts to reform the core strategies of policing in the 1980s, empirical research continued but was more narrowly focused than previ-
CRIMINAL JUSTICE RESEARCH ON POLICE 25 ously. Descriptive studies of the use of deadly force, reminiscent of research about police discretion in the 1960s, prompted police agencies to adopt general policies about its use. Research showed that police behavior could successfully be changed through this mechanism. Larry Sherman and his colleagues undertook a series of studies that evaluated the deterrent effect of arrest on men accused of spouse assault (Sherman, 1992). The findings of the Minneapolis Spouse Assault Project, which was the initial study, were almost immediately incorporated into legislation and police practice. Although its findings were significantly modified by later research, the Min- neapolis Spouse Assault Project is probably the single most influential re- search undertaken since 1967. During the 1990s, evaluations of the impact of police strategies on community safety expanded without a dominant focus--drug crackdowns, community crime prevention, DARE, beat patrols, crime prevention edu- cation, and coordinated interagency crime prevention. While community policing dominated the headlines, due largely to the federal government's investment of $8.8 billion in the Community Oriented Policing Program in 1994, researchers pursued a more diverse agenda. The record suggests, in fact, that during the 1990s researchers recognized that in order to be evalu- ated successfully, police strategies, including community-oriented policing, had to be disaggregated, broken down into specific crime and disorder programs. By the end of the decade, the evaluation literature about crime control and prevention was so extensive, as well as contradictory, that the Univer- sity of Maryland, supported by the National Institute of Justice, attempted to summarize what research evidence had shown to work, not to work, and to have promise (Sherman et al., 1997). In the process, the Maryland team developed a scale for judging the scientific rigor of evaluation research based on the control of variables, measurement error, and statistical power. Although the range and sophistication of evaluative research increased during the 1990s, the contribution of the police to the decade's remarkable decline in crime was undetermined and controversial. This may have been due to the unexpectedness of the decline, which prevented data about police activities, as well as contextual social correlates, from being collected early enough so that a connection could be explored. Also during the 1990s, a research theme from the 1960s emerged with renewed force--the unequal and sometimes abusive treatment of blacks at the hands of the police. It began with concern about police brutality (Skolnick and Fyfe, 1993; Toch and Geller, 1996) and grew to encompass racial profiling by the police (Fridell et al., 2001). Although interest in police misbehavior had never entirely died out, the events of the 1990s rekindled interest in police ac- countability and discipline (Walker, 2001).
26 FAIRNESS AND EFFECTIVENESS IN POLICING Taking the period as a whole (1967 to 2000), the 10 most important topics that police researchers studied were: 1. Organization and management. 2. Crime. 3. Strategies, including community-oriented policing. 4. Drugs. 5. Women. 6. Discrimination. 7. Evaluation. 8. Ethics/accountability/discipline. 9. International/comparative policing. 10. Patrol. This finding is based on a keyword search of four citation indexes: Criminal Justice Abstracts, Sociological Abstracts, the 12 leading criminology/crimi- nal justice journals (listed earlier), and the three professional police periodi- cals (mentioned above). The 10 topics were most frequently coupled with the words "police," "policing," and "law enforcement" in the titles, identi- fiers, or abstracts of articles that appeared in the serials covered by these indexes. Details about this study and its methodology appear in the chapter appendix. There was remarkable agreement across the three social science indexes with respect to these topics, even to the rank order. At the same time, the popularity of the topics, indicated by the frequency of their citation, varied enormously from the top to the bottom of the list. For example, articles about organization and management, depending on the index, were from 8 to 18 times more likely to be written about than patrol. The most popular topics selected for discussion in the three police- oriented journals also appeared at the top of the social science lists: crime, organization and management, drugs, and strategies, including commu- nity-oriented policing. Only two topics were unique to the practitioner list: traffic and criminal investigation. Interestingly, practitioners were as inter- ested in ethics/accountability/discipline as scholars. And international/com- parative articles about police occurred more frequently in the practitioner journals. From negligible beginnings, scientific research on the police has become a large and diverse enterprise by the beginning of the 21st century. It has sought both to explain police activity--the social science perspective--and to evaluate its effects--the public policy perspective. Although the topics studied have varied in popularity over time, no intellectual thread disap- pears entirely in any period. Police research has consistently described po- lice behavior, analyzed its determinants, evaluated its efficacy, and judged
CRIMINAL JUSTICE RESEARCH ON POLICE 27 its rectitude. Police research is more diverse than any particular set of sub- stantive fashions. POLICE RESEARCH METHODOLOGIES The early police researchers studied the police by "walking around" (Bittner, 1970). Their research was qualitative and ethnographic, providing subtle understanding of the working life of police officers (Skolnick, 1966; Van Maanen, 1974; Manning, 1977; Muir, 1977; Brown, 1981). Descrip- tive research by scholars working alone remained fashionable through the 1970s, but it tended to die out during the 1980s and 1990s. Qualitative research by individuals was quickly overtaken by controlled observations by multiple observers recording the behavior of large samples of police officers. Both sorts of studies focused almost exclusively on the behavior of patrol officers. Selective and often impressionistic observation by one person is referred to as participant observation; observation by many people using a predesigned protocol is called systematic social observation (Skolnick, 1965; Reiss, 1979). Participant observation is somewhat of a misnomer. On one hand, participant-observers are carefully instructed not to become actively involved in what they witness--indeed, not to partici- pate. On the other hand, ethnographic observers participate in the same sense as participant-observers, in that they accompany police officers at work. It is not participation that distinguishes the two observational meth- odologies, but the systematization of observations, which reflects differ- ences in the scale of research. Systematic social observation is costly and requires considerable orga- nizational skill, and for these reasons it has been rare: Black and Reiss (1967a, b), Bayley and Garofalo (1987), and Mastrofski et al., 1998). The data collected by these efforts have been analyzed and reanalyzed as schol- ars explore hypotheses about the determinants of police behavior (Sherman, 1983; Chermak and Riksheim, 1993). Descriptive studies of the police have also utilized two sorts of surveys: (1) surveys of the attitudes and behaviors of individual police officers, often stratified by rank, and of the public (Flanagan and Longmire, 1996; Bureau of Justice Statistics, 1998; Weisburd, 2000) and (2) "establishment sur- veys" of police organizations to determine structures and practices (Law Enforcement Management and Administrative Survey or LEMAS). With respect to evaluation studies that assess the social consequences of police activity, especially their contribution to public safety, the major meth- odological distinction is between experiments and quasi-experiments. Both, it should be noted, may employ ethnographic as well as systemic observa- tion. In the well-known Kansas City Preventive Patrol Experiment (Kelling et al., 1974), for example, researchers varied the strength of patrol deploy-
28 FAIRNESS AND EFFECTIVENESS IN POLICING ment in demographically similar areas. Despite the difficulties in control- ling police behavior and the reluctance of police forces to adhere to experi- mental requirements, this methodology has been employed extensively: in Newark, NJ, to study foot patrol (Brown, 1981); in Flint, MI, for commu- nity policing (Trojanowicz, 1986); in Oakland, CA, for "beat health" (Green-Mazerolle, 1999); in Jersey City, NJ, for drug market interdiction (Weisburd and Green, 1995); and in Detroit, MI, for drug crackdowns (Bynum and Worden, 1996). In a quasi-experimental research design, the researchers monitor natu- rally occurring changes in police practices and the social factors that they are assumed to affect. Little is required of the police beyond supplying in- formation. The problem with such designs is that it is extraordinarily diffi- cult to infer causal impact due to the multiplicity of variables that generally need to be controlled. For this reason, studies undertaken to determine so fundamental a matter as whether increasing or decreasing the number of police hired or deployed affects crime, while of enduring interest, continue to be problematic and controversial (Loftin and McDowall, 1982; Levitt, 1997; Marvel and Moody, 1996). The emphasis on careful evaluation of the impact of police strategies led to the making of a theoretical distinction with important methodologi- cal consequences. In order to judge the effectiveness of any police practice, it is critical to describe accurately the practice itself, especially its quality and quantity. In the parlance of research, evaluation of social outcomes from police activity requires careful delineation of the outputs that consti- tute that activity. Evaluation requires the study of both process and result. This distinction has informed the recent movement in policing to develop performance indicators to measure what police do. Police find it easier to report outputs than outcomes, while people outside the police want evalua- tion of outcomes. Both, it turns out, are essential to sound evaluation. And both are essential to public accountability. Although ethnographic research is probably less popular today than it was in the 1960s and 1970s, the methodologies adopted by police research- ers have been relatively stable over time. More attention is given today to random assignment in experiments, as well as to problems of displacement. Understanding of the difficulties of causal inference from statistical correla- tions has also increased. As a result, the requirements of statistical analysis have become more demanding. These developments represent changes in the sophistication with which methods are used and not the development of new methods. Police research has profited from the fact that it has devel- oped along with the social sciences. It has not had to invent new methods, but rather has been able to employ appropriate research practices from all the social sciences.
CRIMINAL JUSTICE RESEARCH ON POLICE 29 At the same time, it would be naive to expect that all police research, even all research that is responsive to the canons of scientific method, has produced equally valid and conclusive results. The scientific quality of re- search varies widely. The University Maryland team developed a scale for rating the scientific merit of research studies evaluating the effectiveness of crime prevention initiatives, including many undertaken by police (Sherman et al., 1997). The scale takes into consideration the control of other vari- ables, measurement error, and statistical power (Sherman et al., 1998). The average score on a 5-point scale was 3.0, with 5 being the highest score. Only 13.4 percent of the studies could be considered to have no serious threats to internal validity. The rest had some defect, mostly commonly failing to control for causal direction and context. AUSPICES OF POLICE RESEARCH Scientific, peer-reviewed research on the police in the United States is currently carried out by and large by people who are not employed by either the police or the government. They are professional researchers work- ing in colleges and universities, nonprofit think tanks, or nongovernmental organizations.5 The police themselves do very little genuine scientific re- search, although since 1967 they have become more accessible to it and more understanding of it. The research and development units of most po- lice agencies serve instead as all-purpose staff for senior executives. Although few people who do scientific research on the police are gov- ernment employees, the bulk of funding for police research comes from government. This was not always so. The American Bar Foundation's pio- neering descriptions of criminal justice practices in the late 1950s and early 1960s were supported by private foundations. So, too, was the Reiss and Black systematic observation of patrol behavior for the President's Com- mission on Law Enforcement and the Administration of Justice (1967). In 1970 the Ford Foundation created the Police Foundation with a grant of $30 million for the purpose of inducing progressive change through the scientific description and analysis of police activities (Lewis and Kelling, 1979). The Police Foundation later created the Police Executive Research Forum, whose purpose was to study policy issues important to the country's largest municipal police departments. Despite the precedent set by Ford and a handful of other foundations, the dominant source of financial support for research on the police has been 5Scholars working for state universities are technically government employees, but the stan- dards of academic freedom with respect to research are so well developed that they should be considered independent of government.
30 FAIRNESS AND EFFECTIVENESS IN POLICING government. Moreover, government funding has also become increasingly concentrated. First, the federal government substantially outspends state and local governments for peer-reviewed research--by the Law Enforce- ment Assistance Administration from the late 1960s to 1978 and by the National Institute of Justice after 1978. Second, the U.S. Department of Justice accounts for most of the federal government's investment in scien- tific police research, through the National Institute of Justice, the Office of Justice Programs, and the Office of Community Oriented Police Services. Over the past two decades, other federal research agencies, for example the National Science Foundation, have increasingly deferred to the Department of Justice with respect to criminal justice research. It is very difficult to estimate changes in financial support for police research since the 1967 crime commission, even the proportion of invest- ment coming from the federal government. The federal government does not maintain a central register of the cost of programs supporting criminal justice, nor is it possible to isolate the research component of such invest- ments without examining each program separately. The report of the Uni- versity of Maryland team concluded that federal investment across the board for crime prevention was small, in the range of $8 million in 1996, and probably had not changed much since the beginning of federal funding for this purpose in 1969 (Sherman et al., 1997). This impression is supported by figures from the National Research Council (1993) that showed that in the early 1990s, the U.S. government spent approximately $800 per year of potential life lost to cancer, almost $700 per life from AIDS, $40 per life from heart, lung, and blood diseases, but only $31 for potential lives lost due to interpersonal violence (cited in Travis, 1995). Paradoxically, then, police research in the United States today is independent of government in its practice but dependent on it for funding. RESEARCH IMPACT OF THE 1994 CRIME ACT The committee is well aware of a federally funded body of research that resulted from the Violent Crime Control and Law Enforcement Act of 1994, which mandated evaluation of existing police programs as well as new ones funded directly by the act itself. The Department of Justice implemented this mandate by transferring money internally from the Community Ori- ented Policing Services (COPS) Office to the National Institute of Justice. The COPS-funded research allocation to the National Institute of Justice totaled $46,639,165, the largest single investment in police research under- taken by this, or any other, government (Figure 2-1 shows yearly expendi- tures). In managing this enormous investment in police research, the National Institute of Justice set as its overarching goal: "to explore practices that will
CRIMINAL JUSTICE RESEARCH ON POLICE 31 20,000 18,000 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 0 1995 1996 1997 1998 1999 2000 2001 FIGURE 2-1 COPS-funded police research expenditures by the National Institute of Justice. improve the quality of police services." It specified several strategies to achieve this goal: "one, identifying effective tactics and technologies; two, highlighting efficient organizational processes and structures; three, improv- ing mechanisms by which police identify public needs or expectations; and, four, informing the public debate regarding what the police do on the job" (Samuels, 2000). From 1995 through 2001, 177 grants were given by the National Insti- tute of Justice using COPS money. Half (88) were in the range of $150,000 to $399,999; 38 percent (67) were $150,000 and under. Only 12 percent (22) were $400,000 and over, and only 4 of those were over $900,000 (see Figure 2-1). The largest research grant from 1994 crime act monies was awarded to the Urban Institute--$3,356,156--to evaluate the success of the COPS of- fice in encouraging the adoption of community policing. Using nationwide phone surveys, site visits, and case studies, the report (Roth et al., 2000) found: 1. High crime areas applied for and received 31 percent of all funds awarded by the COPS office. 2. COPS grants increased the number of police officers working on the street. 3. COPS initiatives lead to the spread of community policing practices throughout the country, although the content and quality of those initia- tives varied considerably.
32 FAIRNESS AND EFFECTIVENESS IN POLICING 4. The impact of technology purchased with COPS money was also uneven across the country. The Urban Institute did not evaluate the link between COPS invest- ments and crime rates, although others have attempted to do so (Eck and Maguire, 2000; Zhao and Thurman, 2001; Muhlhausen, 2002). With COPS funds the National Institute of Justice also sponsored the largest study of police patrol activity since the 1970s--the Project on Polic- ing Neighborhoods in Indianapolis and St. Petersburg, 1996-1997--at a cost of $1,969,701. The study produced five major findings with respect to patrol activity and community police: 1. Patrol officers specializing in community policing spend more time problem solving than do patrol generalists (Parks et al., 1999). 2. Community police officers tend to interact more with people of higher socioeconomic status and less with those with severe problems than do police generalists. 3. Police officers' views toward community policing bear little relation- ship to their behavior, suggesting that imbuing officers with the philosophy of community policing will do little to advance its implementation. 4. Officers' views about community policing have no effect on their success in getting citizens to comply with police commands (McCluskey et al., 1999), the amount of coercion used against suspects (Terrill, 1997), or the amount of time engaged in problem-solving activities (DeJong et al., 2001). 5. Officers with a positive predisposition toward community policing were more likely to comply with requests to control other citizens (Mastrof- ski et al., 2000). Within the COPS portfolio, the National Institute of Justice has also devoted significant resources to research on firearm violence, especially fire- arm violence by young people. The largest of these grants was for the Na- tional Evaluation of Youth Firearms Violence Initiative ($1,314,787). The COPS office itself created the Youth Firearms Violence Initiative in 1995, which involved spending approximately $1 million to assist 10 police de- partments to develop and implement programs to reduce youth gun vio- lence. Evaluations of these efforts showed that traditional police practices are not effective in achieving this goal. Instead, success was achieved when departments worked cooperatively with other government and community organizations, initiated aggressive arrest policies for gun crime, and created special units to target gun crime (Dunworth, 2000). The studies also showed that when federal money for such programs was no longer available, de- partments returned to traditional practices.
CRIMINAL JUSTICE RESEARCH ON POLICE 33 Although it is important to assess the quality of COPS-funded research through NIJ, the committee decided not to do so for two reasons. First, many committee members have been recipients of these grants and have worked at universities or research organizations that have financially ben- efited from them, creating a potential conflict of interest. Second, by sum- mer 2002, when this report was written, only 50 percent of COPS-sup- ported research projects had produced written results that appeared in the National Criminal Justice Research Service data base. A total of 40 projects, representing 27 percent of the NIJ portfolio, were officially closed without a publication of any kind, while 41 grants (23 percent) were still active but have not produced a publication (see Figure 2-2). It should be noted that a significant proportion of COPS-supported research was not designed to produce peer-reviewed publications, although some nonetheless did. These constitute the Locally Initiated Research Project, which represented 23 percent of the COPS-supported grants awarded (41 of 177) and 11.4 percent of the money awarded. The purpose of these grants was to create partnerships between police agencies and re- search groups for the development and evaluation of new strategies for the delivery of police services. Despite the large investment in research on community policing, out- come evaluations and other research still cannot comprehensively and de- finitively provide a guide to the usefulness of this strategy. Therefore, the committee recommends more research on problem-oriented and commu- nity policing. These are two of the most widely discussed innovations in 180 160 Total Number of Grants 140 Active Grants without 120 Publication 100 Active Grants with Publication thousands) 80 (In Closed Grants without 60 Publication 40 Closed Grants with 20 Publication 0 FIGURE 2-2 Total number of COPS-funded police research expenditures by grant and publication status.
34 FAIRNESS AND EFFECTIVENESS IN POLICING policing today, each involving complex packages of programs and organi- zational adaptations, and each calling for efficiency and effectiveness mea- sures that currently are not well captured by most police information sys- tems. The organizational structures and practices that comprise these innovations are highly varied, making it difficult to form general conclu- sions about their effectiveness. Police employ tactics under each of these strategies that can be rigorously evaluated, but we found that there gener- ally was not yet enough evidence to document their successes or failures. Future research should do more to advance knowledge about the effective- ness of these innovations by focusing on their key elements. The organiza- tional arrangements that foster effective community and problem-oriented policing, and the effectiveness of police training in this area, are also not well understood. CONCLUSION The enterprise of police research that has developed in the United States since 1967 is unique in the world in several respects. First, no other country has made a more concerted effort to harness the rigor of social science to the study of policing. The scale of this effort is greater than that of any other country, perhaps of the rest of the world put together. Second, police research in the United States is behavioral rather than jurisprudential. It is based more on accurate description of the police in action, whether by individuals or agencies, and less on its legal authoriza- tion. Police researchers in the United States are trained primarily in social science rather than law. Third, rigorous police research is done mostly by people who are em- ployed outside of government. In the United States there is a division of labor between those who do policing and those who study it. This is often deplored, because it deprives police of in-house research capacity, but it has helped to ensure more independent inquiry. Fourth, over the past 35 years, private philanthropy has largely with- drawn from the support of police research. Today, government dominates the funding of police research. In this connection, one must ask whether the growing dependence of police research on the financial support of government has affected what has been studied or how studies are conducted. In our view, it undoubtedly has. Governments are more interested in evaluation that helps to reform police activity than in description that can be used to explain it. Further- more, governments are more inclined to accept what police do (outputs) as a measure of effectiveness than what they achieve (outcomes). Police are, after all, part of government, and therefore they share an interest in shaping
CRIMINAL JUSTICE RESEARCH ON POLICE 35 appearances. It is also easier and cheaper to document output activity than outcome accomplishments. Ironically, the emphasis on developing "perfor- mance indicators" in the Western democracies in recent years has encour- aged this tendency. Although the impulse to demonstrate usefulness to the public is laudable, outcome evaluations, which take years to complete, do not satisfy the political need for instantaneous accountability. Governments are also more likely to support research that evaluates the effectiveness of the police in terms of instrumental goals, such as crime control, drug interdiction, and traffic safety, than research documenting the normative and legal rectitude of police in achieving these objectives. Al- though governments generally embrace performance assessment for the police more than they did a generation ago, they are still reluctant to moni- tor normative performance. This accounts in part for the rapid develop- ment of civilian review in the United States over the last decade (Walker, 2001). As a general rule, governments fund research on the normative as- pects of police conduct only when driven to it by public outrage over revela- tions of alleged misbehavior. Since the President's Commission on Law Enforcement and the Admin- istration of Justice (1967), the United States has developed a research enter- prise that is substantial, empirical, expert, and independent of direct gov- ernmental, including police, control. But police research in the United States is distinctive in more than its generation of knowledge, extensive though that has been. Research has become part of the institutional practice of American policing. State and local police agencies, which account for the bulk of American policing, are the most open bureaucracies in the land, second only, perhaps, to schools. Few other American institutions, public or private, allow outsiders to observe routine operations or share in-house information as freely as the police do. Understanding research and incorporating its insights into policy have become hallmarks of police management in the United States. Facts matter in a way that they didn't a generation ago. Research--along with higher educational standards for recruits, encouragement of postgraduate degrees for promotion, and specialized in-service courses for managers--has made American policing considerably more intelligent and less insular than it was when the president's crime commission reported in 1967. Police research has also made American police more accountable by opening them to outside opinions, by creating partnerships with indepen- dent observers, and by helping the public and its representatives to judge whether the police are fulfilling their mandate to provide security in accor- dance with the principles of a democratic society. These indirect benefits from police research, which arise as much out of the conduct as the findings of research, may be even more important than its substantive contributions to police policy.
36 FAIRNESS AND EFFECTIVENESS IN POLICING APPENDIX 2A: METHODOLOGY OF KEYWORD SEARCH The data used to document the popularity of topics that have been studied by police researchers since 1967 are based on a keyword search of the following four bibliographic indexes: 1. Sociological Abstracts, covering 2,500 periodicals. 2. Criminal Justice Abstracts, covering 549 periodicals. 3. Twelve criminal justice journals: American Journal of Police (1981) Crime and Delinquency (1960) Criminal Justice History (1980) Criminology (1963) Journal of Crime and Justice (1981) Journal of Criminal Justice (1973) Journal of Quantitative Criminology (1985) Journal of Research on Crime and Delinquency (1964) Justice Quarterly (1984) Police and Society (1990) Police Science and Administration (1973) Police Studies (1978). 4. Three professional police journals: FBI Law Enforcement Bulletin Police Chief Magazine Law Enforcement News. The titles, abstracts, and delimiters were searched to determine how frequently the keywords listed below were found with "police," "polic- ing," and "law enforcement." The search determines the number of sepa- rate citations in which words are associated and not the number of times these associations appear in titles, abstracts, and delimiters. Because the indexes do not cover the life of all the journals indexed, the number of citations for particular keywords found understates slightly the number actually published. A total of 68 keywords were searched for their association with "po- lice," "policing," and "law enforcement." Tables 2A-1 through 2A-6 dis- play the results.
CRIMINAL JUSTICE RESEARCH ON POLICE 37 Accountability Drug(s) Police community Affirmative action Effectiveness relations Battered women Ethics Prejudice Broken windows Evaluation Problem-oriented Brutality Fear of crime policing Budget(s) Firearm(s) Public opinion Civil rights Gender Quality of life Community policing Gender bias Race Comparative Grievance Rank(s) Corruption Homosexuality Recruitment Cost(s) Human rights Riot(s) Crime Inequality Road safety Crime prevention Information technology Spouse abuse Criminal investigation Integrity Spouse assault Criminal justice International Strategy Criminal proceedings Leadership Structure Crowd control Management Supervision Deadly force Morale Tactic(s) Discipline Narcotics Technology Discretion Order maintenance Traffic Discrimination Organization Training Disorder Patrol Victimology Diversity Penology Women Domestic violence
38 FAIRNESS AND EFFECTIVENESS IN POLICING TABLE 2A-1 Keyword Frequency by Index Sociology Criminal Justice Abstracts Abstracts 12 Research 3 Professional Keyword Data Base Data Base Journals Journals Accountability 102 396 21 6 Affirmative action 25 57 7 9 Battered women 129 92 8 3 Broken windows 3 17 1 2 Brutality 52 163 24 178 Budget or budgets 34 152 6 38 Civil rights 98 138 27 110 Community policing 177 570 68 75 Comparative 149 356 86 11 Corruption 124 463 70 192 Cost or costs 157 821 80 77 Crime 1,503 5,998 1,384 1,790 Crime prevention 101 1,133 182 677 Criminal investigation 29 80 61 375 Criminal justice 512 2,588 695 639 Criminal proceedings 55 45 7 0 Crowd control 7 22 10 45 Deadly force 17 170 19 54 Discipline 119 398 60 124 Discretion 107 468 96 42 Discrimination 169 353 100 296 Disorder 87 300 35 4 Diversity 54 119 10 8 Domestic violence 163 440 81 82 Drug(s) 409 1,433 254 726 Effectiveness 217 925 85 32 Ethics 63 283 82 106 Evaluation 234 1,055 221 207 Fear of crime 80 321 43 18 Firearm(s) 56 367 29 248 Gender 230 257 133 1 Gender bias 5 8 8 0 Grievance 29 79 10 2 Homosexuality 53 73 6 31
CRIMINAL JUSTICE RESEARCH ON POLICE 39 TABLE 2A-1 Continued Sociology Criminal Justice Abstracts Abstracts 12 Research 3 Professional Keyword Data Base Data Base Journals Journals Human rights 74 111 5 5 Inequality 94 78 46 3 Information technology 30 21 1 6 Integrity 16 108 13 18 International 167 593 35 556 Leadership 91 282 23 45 Management 213 1,146 118 351 Morale 20 104 6 19 Narcotics 40 349 65 597 Order maintenance 31 64 8 0 Organization 850 2,091 236 134 Patrol 138 1,449 108 550 Penology 15 27 34 9 Police community relations 455 600 282 890 Prejudice 49 108 13 4 Problem-oriented policing 10 76 6 4 Public opinion 85 208 115 46 Quality of life 24 87 6 4 Race 371 651 151 19 Rank or ranks 84 336 13 19 Recruitment 79 338 15 18 Riot(s) 142 266 28 60 Road safety 1 8 3 0 Highway safety 2 14 0 22 Traffic safety 7 25 2 124 Spouse abuse 23 77 13 14 Spouse assault 4 16 3 0 Strategy 159 440 53 23 Structure 670 1,343 137 13 Supervision 54 377 66 25 Tactic(s) 98 379 15 89 Technology 92 283 24 321 Traffic 73 497 28 468 Training 294 2,053 190 959 Victimology 8 34 11 5 Women or woman 505 1,003 244 149
40 FAIRNESS AND EFFECTIVENESS IN POLICING TABLE 2A-2 Citation Frequency by Topics Sociology Criminal Justice Abstracts Abstracts 12 Research 3 Professional Topic Data Base Data Base Journals Journals Crime 1,503 5,998 1,384 1,790 Disorder 87 300 35 4 Broken windows 3 17 1 6 Fear of crime 80 321 43 18 1,673 6,636 1,463 1,808 (If drugs, domestic violence, and crime prevention are added the totals are as follow) (2,519) (10,099) (2,056) (3,239) Drugs 409 1,433 254 732 Narcotics 40 349 65 597 449 1,782 319 1,379 Domestic violence 163 440 81 82 Spouse assault 4 16 3 0 Battered women 129 92 8 3 296 548 92 85 Women 505 1,003 244 149 Gender 230 257 133 1 735 1,260 377 150 Police community relations 455 600 282 890 Public opinion 85 208 115 46 540 808 397 936 Evaluation 234 1,055 221 207 Effectiveness 217 925 85 32 451 1,980 306 239 Organization 850 2,091 236 134 Structure 670 1,343 137 13 Management 213 1,146 118 351 Leadership 91 282 23 35 Supervision 54 377 66 25 Budget(s) 34 152 6 38 Cost(s) 157 821 80 77 Grievance 29 79 10 2 Morale 20 104 6 19 Rank(s) 84 336 13 19 Recruitment 79 338 15 18 Training 294 2,053 190 959 2,383 11,505 690 1,709
CRIMINAL JUSTICE RESEARCH ON POLICE 41 TABLE 2A-2 Continued Sociology Criminal Justice Abstracts Abstracts 12 Research 3 Professional Topic Data Base Data Base Journals Journals (If accountability/discipline are added, then) (2,605) (12,299) (771) (1,839) International 167 593 35 556 Comparative 149 356 86 11 316 628 121 566 Community policing 177 570 68 75 POP 10 76 6 4 Order maintenance 31 64 8 0 Quality of life 24 87 6 4 242 797 88 83 (If police-community relations are added, then) (782) (1,605) (485) (973) Patrol 138 1,449 108 550 Discrimination 169 353 100 296 Prejudice 49 108 13 4 Race 371 651 151 19 Diversity 54 119 10 8 Gender bias 5 8 9 0 Inequality 94 78 46 3 Homosexuality 53 73 6 31 Affirmative action 25 57 7 9 820 1,447 342 370 Strategy 159 440 53 23 Tactics 98 379 15 89 257 819 68 112 Traffic 73 497 28 468 Road safety 1 8 3 0 Highway safety 2 14 0 22 Traffic safety 7 25 2 124 83 544 33 614 Technology 92 283 24 321 Information 30 21 1 6 technology 122 304 25 327 continued
42 FAIRNESS AND EFFECTIVENESS IN POLICING TABLE 2A-2 Continued Sociology Criminal Justice Abstracts Abstracts 12 Research 3 Professional Keyword Data Base Data Base Journals Journals Crime prevention 101 1,133 182 677 Disorder 87 300 35 4 Broken Windows 3 17 1 2 90 317 36 6 Ethics 63 283 82 102 Integrity 16 108 13 18 Corruption 124 463 70 192 Brutality 52 163 24 178 255 1,017 189 490 Human rights 74 111 5 5 Civil rights 98 138 27 110 172 249 33 115 Riot 142 266 28 60 Crowd control 7 22 10 45 149 288 38 105 Accountability 102 396 21 6 Discipline 119 398 60 124 221 794 81 130 Discretion 107 468 96 42 Firearms 56 367 29 248 Deadly force 17 170 19 54 73 537 48 302 Criminal investigation 29 80 61 375 NOTE: Keywords have been grouped into topics--subject matter that involves more than one keyword.
CRIMINAL JUSTICE RESEARCH ON POLICE 43 TABLE 2A-3 Topics Rank-Ordered in Sociology Abstracts Topics Citations 1 Organization 2,383 (with accountability/discipline) (2,605) 2 Crime 1,673 (with drugs, domestic (2,519) violence, and crime prevention) 2a See strategies (1,382) 3 Discrimination 820 4 Women 735 5 Police community relations 540 6 Strategies, including COP 499 (with COP at 242 COP with c-p rels 782 Crime prevention 101) (1,382) 6a See ethics/acct/discipline (477) 7 Evaluation 451 8 Narcotics 449 9 International/comparative 316 10 Domestic violence 296 11 Ethics 255 12 Accountability/discipline 222 (with ethics) (477) 13 Human rights 172 14 Riots 149 15 Patrol 138 16 Technology 122 17 Discretion 107 18 Crime prevention 101 19 Disorder 90 20 Traffic 83 21 Firearms 73 22 Criminal investigation 29
44 FAIRNESS AND EFFECTIVENESS IN POLICING TABLE 2A-4 Topics Rank-Ordered in Criminal Justice Abstracts Topics Citations 1 Organization 11,505 (with accountability/discipline) (12,299) 2 Crime 6,636 (with drugs, domestic (10,099) violence, and crime prevention) 3 Evaluation 1,980 4 Drugs 1,782 4a Strategy with COP (1,616) 4b COP with community relations (1,605) 5 Patrol 1,449 6 Discrimination 1,447 7 Women 1,260 8 Crime prevention 1,133 9 Ethics 1,017 (with acct/discipline) (1,239) 10 Strategy 819 11 Police-community relations 808 12 COP 797 13 Accountability/discipline 794 14 International/comparative 628 15 Domestic violence 548 16 Traffic 544 17 Firearms 537 18 Discretion 468 19 Technology 304 20 Riots 288 21 Human rights 249 22 Criminal investigation 80
CRIMINAL JUSTICE RESEARCH ON POLICE 45 TABLE 2A-5 Topics Rank-Ordered in 12 Criminal Justice Journals Topics Citations 1 Crime 1,463 (with drugs, domestic (2,056) violence, and crime prevention) 2 Organization 690 (with accountability/discipline) (771) 2a COP with police-community relations (485) 3 Police community relations 397 4 Women 377 5 Discrimination 342 6 Drugs 319 7 Evaluation 306 8 Ethics 189 9 Crime prevention 182 (with strategy and COP) (156) 10 International/comparative 121 11 Patrol 108 12 Discretion 96 13 Domestic violence 82 14 COP 88 15 Accountability/discipline 81 16 Strategy 68 17 Criminal investigation 61 18 Firearms 48 19 Riots 38 20 Traffic 33 21 Human rights 33 22 Technology 25
46 FAIRNESS AND EFFECTIVENESS IN POLICING TABLE 2A-6 Topics Rank-Ordered in Three Professional Police Journals Topics Citations 1 Crime 1,808 (With drugs, domestic (3,239) violence, and crime prevention 2 Organization 1,709 (With accountability/discipline) (1,839) 3 Drugs 1,379 (3a Community policing with police-community relations 973) 4 Police community relations 936 5 Crime prevention 677 6 Traffic 614 7 International/comparative 566 8 Patrol 550 9 Ethics 490 10 Criminal investigation 375 11 Discrimination 370 12 Technology 327 13 Firearms 302 14 Evaluation 239 15 Women 150 16 Accountability/discipline 130 17 Human rights 115 18 Strategy 112 19 Riots 105 20 Domestic violence 85 21 Community policing 83 22 Discretion 42