U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics Special Report Community Policing in Local Police Departments, 1997 and 1999 February 2001, NCJ 184794 ----------------------------------------------------------- This file is text only without graphics and many of the tables. A Zip archive of the tables in this report in spreadsheet format (.wk1) and the full report including tables and graphics in .pdf format are available from: http://www.ojp.usdoj.gov/bjs/abstract/cplp99.htm ----------------------------------------------------------- By Matthew J. Hickman and Brian A. Reaves, Ph.D. BJS Statisticians ----------------------------------------------------------- Highlights Full-time sworn personnel regularly engaged in community poling activities, by type of agency, 1997 and 1999 Total 112,611 Local police 91,072 Sheriff 15,443 Other 6,096 Total 21,216 Local police 15,978 Sheriff 3,649 Other 1,588 * State and local law enforcement agencies had nearly 113,000 community policing officers or their equivalents during 1999, compared to about 21,000 in 1997. This included 91,000 local police officers in 1999, up from 16,000 in 1997. * 64% of local police departments serving 86% of all residents had full-time officers engaged in community policing activities during 1999, compared to 34% of departments serving 62% of residents in 1997. * In 1999, 87% of local police officers were employed by a department that provided community policing training for some or all new recruits, and 85% by a department that provided it for at least some in-service officers. * In 1999, 63% of local police departments serving 85% of all residents used routine foot and/or bicycle patrol. About half of all officers worked for a department that actively encouraged them to engage in problem-solving projects on their patrol beats. * As of June 30, 1999, 69% of local police departments, serving 96% of all residents, had met with community groups within the past year; and 40% of departments, serving 71% of residents, had formed a problem-solving partnership within the past 3 years. * During 1999, 92% of residents were served by a local police department that provided them with routine access to crime statistics or crime maps, compared to 70% in 1997. ---------------------------------------------------------- The community policing approach to law enforcement seeks to address the causes of crime and reduce the fear of crime and social disorder through problem-solving strategies and police-community partnerships. A fundamental shift from traditional reactive policing, community policing strives to prevent crime before it occurs. As part of an ongoing partnership between the Bureau of Justice Statistics (BJS) and the Office of Community Oriented Policing Services (COPS), data were collected on the community policing practices of State and local law enforcement agencies in 1997 and 1999. The Law Enforcement Management and Administrative Statistics (LEMAS) survey was used to assess the impact of community policing on personnel, training, policies, programs, and the use of technology. The LEMAS survey Begun in 1987, the LEMAS survey periodically collects data from State and local law enforcement agencies. All agencies employing 100 or more sworn personnel are included, as well as a representative sample of smaller agencies. LEMAS surveys have been conducted in 1987, 1990, 1993, 1997, and 1999. COPS provided funding for the 1997 and 1999 surveys, and both surveys included a section devoted to community policing. The 1999 LEMAS survey questionnaire was mailed to 3,319 agencies, including primary State law enforcement agencies, sheriff's departments and offices, local police departments, and certain types of special jurisdiction police departments. (See Methodology on page 11.) A total of 3,246 State and local agencies provided data for the 1999 survey for a response rate of 97.8%. Included were 2,052 of the 2,085 local police departments that received the questionnaire, a response rate of 98.4%. In 1997, 2,012 of 2,097 local police departments provided data, resulting in a response rate of 95.9%. More than 17,500 agencies and nearly 700,000 full- time sworn personnel are represented by the data collected in the 1999 LEMAS survey. Full-time sworn Agencies personnel Local police 13,524 436,182 Sheriff 3,087 185,859 State police 49 55,892 Special* 880 16,121 *Agencies with special jurisdictional areas or special law enforcement responsibilities. Unless otherwise indicated, the reference dates for all data cited in this report are June 30, 1997, and June 30, 1999. Community policing personnel As of June 30, 1999, local police departments employed an estimated 436,182 full-time sworn personnel and served approximately 195 million residents nationwide. About 90% of these departments served a population of less than 25,000 residents, accounting for about 30% of all full-time local police officers. Nearly half of all local police officers were employed by 1% of departments that served a population of 150,000 or more. One important measure of the level of local police involvement in community policing is the number of full-time sworn personnel who serve as community policing officers, community resource officers, community relations officers, or otherwise regularly engage in community policing activities. From 1997 to 1999, the percentage of local police departments with community policing officers nearly doubled -- from 34% to 64%. During this time the number of such officers increased from about 16,000 to about 91,000. (See Methodological note on page 11 regarding comparisons of data from the 2 years.) In 1999 all departments serving a population of 1 million or more and about 90% of those serving 25,000 to 999,999 residents were using community policing officers. Among departments using community policing officers, the average number so designated increased from 3 in 1997 to 11 in 1999. Large increases were reported among departments in all population categories. In the smallest communities the average number of community policing officers doubled, and it more than doubled in all other categories. In 1999 the average number of community officers ranged from 1,728 in the largest jurisdictions to 2 in the smallest. ----------------------------------- The Crime Act of 1994 and the Office of Community Oriented Policing Service (COPS) The Violent Crime Control and Law Enforcement Act (the Crime Act) was signed into law by President Clinton on September 13, 1994. Of the $30 billion in expenditures authorized by this legislation, nearly $9 billion was authorized for Title I, also known as the Public Safety Partnership and Community Policing Act of 1994. Title I was the basis for the creation of the Department of Justice's Office of Community Oriented Policing Services (COPS) program. Since 1994, to advance community policing, COPS has provided over $7.5 billion in Federal funds to over 12,000 State and local law enforcement agencies. The COPS office was charged with changing the practice of policing in the United States by -- * Deploying an additional 100,000 community policing officers * Facilitating problem-solving efforts and encouraging interactions with communities by officers * Promoting innovation in policing * Enhancing existing technologies to assist officers in preventing and responding to crime and its consequences. To achieve these goals, the COPS program used three primary approach-es. The first was to award 3-year grants to agencies for hiring officers. The second was to award grants for acquiring technology and hiring civilians so officers could be redeployed to perform community policing activities. The third was to award grants to agencies for innovative programs with special purposes such as reducing gun violence or domestic violence. All COPS grant programs were designed to supplement local expenditures, not to supplant or replace them. Title I required that COPS funds be distributed equally between jurisdictions above and below a population of 150,000. The LEMAS data are a rich source of information about community policing; however, they are not intended to be used to determine whether the goals of the COPS program have been met. For more information on the COPS program, see the National Evaluation of the COPS Program, National Institute of Justice (NCJ 183643), or visit the COPS website at www.usdoj.gov/cops --------------------------------------- In 1997 about 4% of all local police officers were serving as community policing officers. By 1999 the percentage of officers serving as community policing officers or otherwise regularly engaged in community policing activities had increased to 21%. In jurisdictions with 500,000 or more residents, the percentage of full-time sworn personnel serving as community policing officers increased from 1.4% to 24.1% (figure 1). Percent of full-time sworn personnel in local police departments serving as community policing officers, by size of population served, 1997 and 1999 1999 1997 500,000 or more 24.1% 1.4% 150,000 - 499,99 14.4% 4.7% 50,000 - 149,999 16.9% 4.8% 10,000 - 49,999 18.4% 4.6% Less than 10,000 28.3% 5.6% Large increases occurred in other population categories as well: In jurisdictions with 150,000 to 499,999 residents, from 4.7% to 14.4%; in those with 50,000 to 149,999 residents, from 4.8% to 16.9%; in those with 10,000 to 49,999 residents, from 4.6% to 18.4%, and in jurisdictions with fewer than 10,000 residents, 28.3% of sworn personnel were serving as community policing officers in 1999 compared to 5.6% in 1997. At the agency level, a small proportion of officers devoted to community policing may indicate experimentation, while a majority of officers so assigned may reflect the adoption of a organizational philosophy based in community policing. In 1997 just 7% of all local police departments, and none of those serving 50,000 or more residents, had at least half of their officers serving as community policing officers. By 1999, the overall percentage had risen to 29%, and the percentages among population categories of 50,000 or more ranged from 4% to 25%. Among departments serving 2,500 to 9,999 residents, the percentage with at least half their officers serving as community policing officers increased from 2% to 23%, and in departments serving fewer than 2,500 residents, from 15% to 41%. Training for community policing In support of their widespread use of community policing officers, just over half of local police departments offered 8 or more hours of training in community policing to some or all new officer recruits in both 1999 and 1997. About two-fifths of all departments trained all of their new recruits in community policing. Such training helps officers develop skills in areas integral to successful community policing such as problem-solving, SARA***Footnote 1: SARA is a problem- solving approach first proposed by Herman Goldstein. For a brief description see .*** (scanning, analysis, response, assessment), and community partnerships. In 1999, 96% of the departments serving 500,000 or more residents trained all new recruits in community policing, as did more than 80% of those serving 150,000 to 499,999 residents. Nearly half of the departments serving 2,500 to 9,999 residents, and about a fourth of those serving fewer than 2,500 residents trained all recruits in community policing. In 1999, 63% of local police departments provided community policing training to at least some of their in-service sworn personnel. Twenty-eight percent of departments trained all in-service officers. As with new recruits, the percentage of departments that trained in-service officers in community policing was about the same in 1999 as in 1997. The percentage of local police officers that were employed by a department that trained some or all new recruits in community policing increased slightly from 1997 to 1999 (figure 2). Percent of local police officers employed by a department providing community policing training to some or all employees, by category of trainee, 1997 and 1999 1999 1997 New officer recruits 87.0% 83.0% In-service officers 85.0% 84.0% Civilian personnel 50.0% 48.0% In 1999, 87% of officers were employed by a department that trained new recruits in community policing compared to 83% in 1997. For training of in-service officers the comparable figures were 85% in 1999 and 84% in 1997; for training of civilian personnel, 50% in 1999 and 48% in 1997. Community policies and programs and policies These nationwide increases -- in the number of departments using community policing officers and in the proportion of sworn personnel designated as community policing officers -- have been accompanied by the continued prevalence of community-oriented policies, programs, and activities. Agency community policing plans In both 1997 and 1999, about two-thirds of local police departments had some type of community policing plan. About a fourth of those with a plan, a sixth of departments overall, had a formally written plan, indicating that community policing approaches and goals were incorporated into the overall agency strategic plan. The remainder had informal, unwritten community policing plans. In both years a majority of the community policing plans of departments serving a population of 50,000 or more were formal, written plans, while departments in smaller jurisdictions were more likely to have an unwritten plan. Foot and bicycle patrol Routine foot or bicycle patrol is a commonly used means to increase police-community interaction. From 1997 to 1999, the percentage of local police departments using routine foot patrol increased from 28% to 34%, and the percentage using routine bicycle patrol went from 50% to 53%. In 1999, 63% of departments, including nearly all of those serving a population of 50,000 or more, used one or both of these patrol types on a routine basis. Of the approximately 195 million U.S. residents served by a local police department, 85% were served by one that used foot and/or bicycle patrol on a routine basis during 1999, compared to 78% in 1997 (figure 3). Percent of population served by a local police department using foot or bicycle patrol units on a routine basis, by size of population served, 1997 and 1999 1999 1997 Either type 85.0% 78.0% Bicycle 76.0% 66.0% Foot 68.0% 59.0% From 1997 to 1999, the percentage of residents served by a department using bicycle patrol went from 66% to 76%, and the percentage served by a department with foot patrol from 59% to 68%. Geographic assignment Assigning patrol officers in fixed geographic- and time-stable beats or areas, often referred to as "territorial stability," is thought to increase the level of officer attachment to an area and its residents. Further, stability of assignment may facilitate certain goals of community policing, such as the development of citizen and business contacts and formation of problem-solving partnerships. In both 1997 and 1999, 42% of local police departments gave their patrol officers responsibility for specific geographic areas or beats. This included about 90% of those serving 50,000 or more residents. Overall, about 4 in 5 local police officers worked for a department that followed this practice. Departments using routine foot or bicycle patrol and giving officers responsibility for specific geographic areas or beats were about 5 times as likely to have full-time community policing officers as departments not implementing these strategies. In both 1997 and 1999, fewer than 1 in 10 local police departments gave their investigators responsibility for specific geographic areas or beats. However, a majority of departments serving 50,000 or residents used this practice and nearly half of all local police officers were employed by a department that did so. Problem solving A central component of community policing is the process of identifying and responding to community problems. This process is often put into operation at the street level using the SARA problem-solving model. The process may be facilitated by activities such as holding meetings with citizen groups, forming problem-solving partnerships, and training citizens in community policing. In 1999 nearly a third of local police departments, employing about half of all officers, actively encouraged patrol officers to engage in SARA-type problem-solving projects on their beats. From 1997 to 1999 the largest increase occurred among departments serving 50,000 to 149,999 residents. Including problem-solving projects in the criteria of performance evaluations of patrol officers is an important way departments can encourage and support community policing efforts. About a sixth of local police departments, employing about a third of all officers, did so in 1999. The largest increase from 1997 to 1999 in the use of such criteria was among departments serving 250,000 to 499,999 residents. During the 3-year period ending June 30, 1999, 40% of local police departments formed problem-solving partnerships with community groups, municipal agencies, or others through specialized contracts or written agreements. This included at least 70% of the departments in each population category of 25,000 or more, and covered 71% of all served by a local police department. Meetings with citizens are considered to be essential to the development of long-term solutions to community problems. About two-thirds of local police departments, serving 96% of all residents, reported meeting with citizens on a regular basis during the year ending June 30, 1999. Nearly all agencies serving populations of 10,000 or more residents regularly met with citizen groups during this period. Among departments holding regular meetings, a variety of citizen groups were typically involved (figure 4). Types of groups that local police departments regularly met with to discuss crime-related problems, 1999 School groups 76.0% Neighborhood associations 52.0% Business groups 50.0% Senior citizen groups 45.0% Domestic violence groups 41.0% Youth service organizations 33.0% Religious groups 30.0% Advocacy groups 26.0% Tenants' associations 19.0% The largest percentages of departments met with school groups (76%), followed by neighborhood associations (52%), business groups (50%), senior citizen groups (45%), and domestic violence groups (41%) (figure 4). Meetings with youth service organizations (33%), religious groups (30%), advocacy groups (26%), and tenants' associations (19%) also occurred. In addition, 40% of departments reported regular meetings with local public agencies. During the 3-year period ending June 30, 1999, about a third of departments offered community policing training to citizens on topics such as community mobilization and problem solving. This included a majority of those serving 10,000 or more residents and about 90% of those serving 150,000 or more residents. Local police departments with a formally written community policing plan were much more likely to have implemented strategies such as geographic patrol assignments (72%), support for officer problem-solving projects (57%), community policing training for citizens (55%), and problem-solving partnerships (49%) than those with only an informal community policing plan or no community policing plan at all (figure 5). Community-oriented activities and policies of local police departments, by type of community polcing plan, 1999 Written plan Unwritten plan No plan Geographic patrol assignments 72.0% 48.0% 19.0% Officer problem-solving projects 57.0% 37.0% 6.0% Community policing training for 55.0% 27.0% 5.0% Problem-solving partnership 49.0% 29.0% 8.0% Information sharing and measurement Citizen input and feedback Community policing requires law enforcement agencies to take a new approach to both the identification of problems and the measurement of police performance. Success in addressing issues such as social disorder and the fear of crime cannot be determined solely through data about arrests or calls-for-service. Some level of citizen input and feedback is often thought to be essential to accurately assess police performance, and systematic citizen input and feedback is preferred. To this end, 27% of local police departments surveyed citizens to gather information in 1999, and 30% did so in 1997. In both years, a majority of the departments serving 25,000 or more residents conducted at least 1 citizen survey. Among departments conducting surveys, 89% collected information on public satisfaction with police services, 63% collected information on public perceptions of crime and disorder problems, and 38% collected information on the personal crime experiences of citizens. Local police departments reported using their survey data for a wide variety of purposes (figure 6). Purposes for which police departments used information collected in citizen surveys, 1999 Providing information to officer 72.0% Evaluating program effectiveness 57.0% Prioritizing crime/disorder problem 52.0% Officer training 42.0% Formulating policy/procedures 42.0% Allocating resources to neighbor 39.0% Redistricting beat/reporting are 16.0% Among departments conducting surveys in 1999, the most common use of the data was providing information to patrol officers (72%). A majority also used survey data to evaluate program effectiveness (57%) and prioritize crime and disorder problems (52%). Other uses of survey information included formulating agency policy and procedures (42%), officer training (42%), allocating resources to targeted neighborhoods (39%), and redistricting beats and reporting areas (17%). Police input and feedback To identify and prioritize local problems, the police and citizens need to share information. At a minimum, citizens need to have some access to local crime information. In 1999, 73% of local police departments provided citizens with routine access to crime statistics, compared to 36% in 1997. This doubling largely occurred from the increased availability among departments serving fewer than 50,000 residents. The percentage of residents served by a local police departments that provided such access rose from 70% in 1997 to 92% in 1999. Among departments providing access to crime statistics in 1999, nearly all provided it to those making their request in person (90%) (figure 7). Methods for accessing crime statistics provided to citizens by local police departments, 1999 In-person 90.0% Telephone 57.0% Written requests 56.0% Newspaper 40.0% Agency reports 38.0% Fax 16.0% Radio 14.0% Television 11.0% Newsletter 11.0% Internet 10.0% More than half provided access by telephone (57%) or written request (56%), and about two-fifths provided it through newspapers (40%) or agency reports (38%). Less common methods of routine access provided to citizens were fax (16%), radio (14%), newsletter (14%), television (11%), and Internet (10%). Although still relatively low in 1999, the percentage of departments providing access to crime statistics that allowed citizens to use the Internet for this purpose was double that of 1997. During the year ending June 30, 1999, 3% of the departments providing routine citizen access to crime statistics, covering 11% of the total population served by such departments, conduct-ed training classes for citizens on how to use or analyze crime statistics or crime maps. In 1999, 83% of the departments providing routine citizen access to crime statistics did so with data compiled at the county or city level. Statistics at a more detailed level such as neighborhood (29%), street (22%), patrol beat (19%), address (17%), apartment complex (14%), district or precinct (12%), or census tract (2%) were much less likely to be routinely available (figure 8). Detailed levels of crime statistics provided by local police departments, 1999 Neighborhood 29.0% Street 22.0% Patrol beat 19.0% Address 17.0% Apartment complex 14.0% District/precinct 12.0% Census tract 2.0% Technology and community policing Crime mapping Technology plays a central role in the advancement of community policing by helping law enforcement agencies to become more efficient and effective problem solvers. Technological tools such as computerized crime mapping assist agencies in identifying and analyzing problems, developing long-term solutions, and evaluating their problem- solving efforts. In 1999, nearly twice as many local police departments were using computers for crime mapping purposes as in 1997 (32% versus 17%). This included more than 90% of those serving 150,000 or more residents, and 76% of those serving 50,000 to 149,999 residents. Overall, departments using computerized crime mapping in 1999 covered 71% of the total population served by local police departments (figure 9). Percent of population served by a local police department with computerized geocoding and mapping, 1999 Any type 71.0% Crime incidents 60.0% Calls for service 50.0% Arrests 46.0% Business locations 32.0% Census data 14.0% A variety of data types were geocoded and mapped, with the most common type being crime incidents, mapped by departments covering 60% of the total population served by local police departments. Other data types and the corresponding population coverage percentages were as follows: calls for service (50%) arrests (46%), business locations (32%), and census data (14%). Community outreach Technology can also assist in community outreach. In 1999, 18% of local police departments, covering 62% of all residents served by a local police department, maintained a agency home page on the Internet. This represented an increase over 1997 when 11% of departments, serving 48% of all residents, had an Internet home page. In 1999, about 9 in 10 departments serving a population of 250,000 or more residents had a home page, as did about 7 in 10 departments serving 50,000 to 249,999 residents. Such websites are typically used to provide citizens with crime and police-related information. They can also increase police-citizen contact through e-mail and information sharing regarding crime problems and related issues. In-field computers In-field computers such as laptops, car-mounted computers/terminals, and hand-held computers/terminals assist problem-solving efforts by providing patrol officers with direct access to vital information. As of June 30, 1999, nearly a third of local police departments, employing nearly three-fourths of all local police officers, were using in-field computers. These proportions were similar to those reported in 1997. Nearly 9 in 10 departments serving 150,000 or more residents were using in-field computers in 1999. In departments where patrol officers used in-field computers, a variety of information resources were available to them. In 1999, a majority of local police officers worked for a department where information on wanted suspects (62%) and driving records (54%) were accessible to least some patrol officers through the use of in-field computers(figure 10). Percent of local police officers employed by a department in which patrol officers have direct access to information through in-field computers, 1999 Wanted suspects 62.0% Driving records 54.0% Stolen property 48.0% Prior calls at dispatch location 42.0% Criminal history records 39.0% Crime analysis software 15.0% Mapping programs 13.0% Other information accessible to patrol officers by in-field computer in local police departments and the corresponding percentage of officers employed by such departments included stolen property records (48%), calls for service history at dispatched location (42%), criminal history records (39%), crime analysis software (15%), and mapping programs (13%). --------------------- The COPS MORE Grant Program COPS offers support to State and local agencies through a variety of different grant programs. One such program, COPS MORE (Making Officer Redeployment Effective), was begun in 1995 to redeploy officers to the street by funding technology, equipment, and support personnel. When agencies can purchase computers and hire civilians for support positions, officers who would otherwise be engaged in support functions can be redeployed to the street in community policing roles. With the assistance of COPS MORE funding, the growth in the use of computers for administrative purposes that began in the early 1990's continued through the decade. From 1990 to 1999, the percentage of departments using personal/desktop computers (PC's), nearly doubled, from 40% to 78%. The growth in new use of PC's is primarily concentrated among departments serving smaller populations. Among departments serving populations of less than 10,000 residents, the percentage using PC's went from 29% in 1990 to 72% in 1999. Nearly all departments serving populations of 50,000 or more have reported using PC's since 1990. COPS MORE grants have funded these agencies to update or enhance their computer capabilities. Percentage of local police departments using personal computers (PC's), by size of population served, 1990-99 250,000 50,000 10,000 Less than or more 249,000 49,000 10,000 1990 99% 88% 69% 29% 1993 99% 92% 84% 45% 1997 99% 98% 88% 60% 1999 100% 100% 95% 72% ---------------- Methodology The Law Enforcement Management and Administrative Statistics (LEMAS) survey collects data from a nationally representative sample of the publicly funded State and local law enforcement agencies in the United States. See the BJS report Local Police Departments, 1997 (NCJ 173429) for information on the 1997 LEMAS sampling methodology. The 1999 LEMAS survey questionnaire was mailed to 3,319 State and local law enforcement agencies. The initial mailing of the questionnaire was conducted in June 1999, and June 30, 1999, was generally used as the reference date for survey questions. Allowing for the exclusion of certain types of special jurisdiction police, 895 law enforcement agencies in the U.S. with 100 or more sworn officers as of June 30, 1996, were included in the sample as self- representing (SR) agencies. This included 529 local police departments, 306 sheriffs' offices, the 49 primary State law enforcement agencies, and 11 special jurisdiction police agencies. The self-representing (SR) agencies were supplemented by a nationally representative sample of agencies with fewer than 100 sworn officers. These nonself- representing (NSR) agencies were chosen using a stratified random sample with cells based on the type of agency (local police, sheriff, or special police), size of population served, and number of sworn officers. The 2,424 NSR agencies selected included 1,556 local police departments, 694 sheriffs' departments and offices, and 174 special jurisdiction police agencies. A total of 3,246 agencies responded to the LEMAS survey for a response rate of 97.8%. This included 883 SR agencies and 2,363 NSR agencies. The final database includes 2,052 local police departments, 967 sheriffs' departments, 178 special jurisdiction police departments, and the 49 primary State police departments. The base weight for all SR agencies is 1. For NSR local police departments, the base weights are determined by number of sworn officers employed as of June 30, 1996. For agencies with 0 to 6 sworn officers, the base weight is 15.55; for 7 to 13 officers, it is 9.65; for 14 to 23 officers, it is 7.09; for 24 to 39 officers, it is 4.95; for 40 to 62 officers, it is 3.38; and for 63 to 99 officers, the base weight is 2.14. The final weight associated with every agency, both SR and NSR, is the product of the base weight, a factor that adjusted for changes in the universe since 1996, and a factor that adjusted for any nonresponding agencies in each cell. Some responding agencies did not fully complete the LEMAS questionnaire. When an agency did not supply a response to an item, median value imputation or ratio imputation was used. Median value imputation uses the median value of an item reported by agencies in the same sample cell. Ratio imputation uses the median value of a ratio reported by agencies in the same sample cell. Median value imputation was used to impute counts of facilities, mainframes, minicomputers, reserve sworn officers, community service officers, and nonsworn volunteers. Ratio imputation was used to impute counts of employees and all other computer equipment. Every imputed item on the database has an indicator that the value has been imputed. Complete documentation regarding sampling procedures and nonresponse adjustments is available upon request. Because the data from agencies with fewer than 100 sworn personnel were collected from a sample, the results are subject to sampling error. Statements of comparison in this report have been tested to ensure that observed differences between values are significant at 2 standard errors (the 95-percent confidence level) or higher. The survey data are also subject to response and processing errors. Methodological note In 1997 agencies were asked for the number of full-time sworn personnel whose regular assigned duties included serving as a community policing officer. In 1999 this definition was reworded to ask for the number of full-time sworn personnel serving as community policing officers, community resource officers, community relations officers, or others regularly engaged in community policing activities. The Bureau of Justice Statistics is the statistical agency of the U.S. Department of Justice. Lawrence A. Greenfeld is deputy director. Matthew Hickman and Brian Reaves, BJS statisticians, prepared this report. Debra Cohen, Ph.D. and Matthew Scheider of the Office of Community Oriented Policing Services (COPS) provided assistance. Dr. Cohen, Karin Schmerler, and Meg Townsend developed the community policing content of the 1999 LEMAS survey instrument. Pam Cammarata, Greg Cooper, Gil Kerlikowske, and Bob Phillips of COPS also assisted. COPS provided full funding for the 1999 LEMAS survey and partial funding for the 1997 LEMAS survey. The data were collected and processed by the U.S. Census Bureau under the supervision of Stephanie Brown of the Governments Division. Census Bureau project staff included Carolyn Gates, Theresa Reitz, Martha Haselbush, Martha Greene, Bill Bryner, Patricia Torreyson, and Henrietta Herrin. Sample design, sample selection and imputation procedures were conducted by Suzanne Dorinski and Yukiko Ellis of the Economic Statistical Methods and Programming Division of the Census Bureau under the supervision of Carma Hogue. Tom Hester edited the report, and Jayne Robinson administered final report production. February 2001. NCJ 184794 End of file 2/22/01 ih