Longitudinal Evaluation of Chicago's Community Policing Program, 1993-2001 (ICPSR 3335)

Data Collection Notes View help for Data Collection Notes

Study Purpose View help for Study Purpose

The purpose of this study was to evaluate the long-term organizational transition of the Chicago Police Department (CPD) to a community policing model. The Chicago Alternative Policing Strategy (CAPS) was an ambitious plan to reorganize the CPD, restructure its management, redefine its mission, and forge a new relationship between police and city residents. With more than 16,600 employees, the CPD is the second largest police department in the United States, serving nearly three million people and responding to calls over a 225-square-mile area. The size and complexity of the CAPS initiative generated significant changes in the department's structure and goals during a multi-year implementation effort, which began in April of 1993. After an experimental period in five police districts, the program was expanded to encompass the entire city. A key aspect of the CAPS program was the implementation of a problem-solving model for policing. Within the CPD, a problem was defined as a group of related incidents or an ongoing situation that concerned a significant portion of those who lived or worked in a particular area. Although dealing with crime remained at the heart of the police mission, it was envisioned from the beginning that the police mandate would coordinate responses to a broad range of community concerns, including social disorder, municipal service problems, and code enforcement matters previously handled by civil courts. According to this strategy, teams of officers were given long-term assignments to each of the city's 279 police beats. They were to spend most of their time responding to calls and working on prevention projects in their assigned areas. To enable them to do so, rapid response units were assigned to excess or low-priority calls. The problem-solving efforts of beat officers were supported by a coordinated system for delivering city services. Beat sergeants were responsible for coordinating their efforts across a 24-hour clock. One mechanism for doing so was beat team meetings that brought together all of the officers serving the area on all watches. Beat sergeants, in turn, reported to a lieutenant charged with coordinating their projects across a larger geographical area. The views of the community were represented through District Advisory Committees (DAC), in which policies and strategies were discussed with commanders, and through monthly community meetings held in every beat. Another important feature of Chicago's problem-solving infrastructure was training for both neighborhood residents and police. Residents were expected to take an active role as partners with the police and on their own.

Study Design View help for Study Design

This evaluation of the CAPS program included surveys of police officers, residents, and program activists. In addition, observational data were collected from beat meetings, and aggregate business establishment and land-use data were added to describe the police beats and districts. Parts 1-13 contain data from surveys of police officers conducted between 1993 and 1999. For Part 1 (1993 Prototype District Training Officer Survey Data) surveys were distributed in the spring of 1993 to personnel serving in five CAPS prototype districts as they assembled in groups of 25 to 40 for an initial orientation session at the training academy. Twenty-four sessions were held over a few weeks. The questionnaires were briefly described by evaluation staff members who attended each briefing, and the same staff members collected the questionnaires for keying. The same questionnaire was also distributed to members of the exempt staff at a morning meeting in the summer of 1993. For Part 2 (1993 Comparison Districts Roll Call Officer Survey Data) surveys were distributed during roll call in five districts selected to match the CAPS prototype areas. They were administered on all shifts to officers passing through roll call. This survey was much shorter than that for the prototype officers to facilitate roll-call administration. Part 3 (1994 Supervisor Training Officer Survey Data) surveys were distributed during training for sergeants in all districts. They included 81 questions that were largely a subset of the 1993 prototype district survey (Part 1). Part 4 (1995 Patrol Division Training Officer Survey Data) surveys were conducted at the beginning of two-day training sessions on CAPS and problem-solving. Part 5 (1996 Exempt Staff and Senior Managers Training Officer Survey Data) surveys were conducted during one of the last in a weekly series of training sessions for exempt staff and civilian managers. Part 6 (1997 Training for Assorted Managers Officer Survey Data) surveys were administered at training for captains, lieutenants, sergeants, and civilian employees at the 911 Center. The training covered leadership and management under CAPS. The questionnaire closely paralleled the one distributed during the 1996 exempt staff training (Part 5). Part 7 (1997 Beat Team Leader Training Officer Survey Data) surveys were conducted at training for beat team leaders (sergeants) in beat plan development and supervision of patrol. Officers were trained in about 15 different groups. At the first half of the sessions, the questionnaires were administered before the training began, and at the other half, at the end of training. Part 8 (1998 Beat Meeting Officer Survey Data) surveys were administered to officers attending beat community meetings. Due to the fluid nature of beat meetings with officers coming and going, and because some officers attended out of uniform, it was difficult for the observers to assess the completion rate for the survey. More complications were introduced by the fact that some officers attended multiple beat meetings at which questionnaires were distributed. They were to be completed each time, but some officers either misunderstood the request or did not want to complete a questionnaire that often. Part 9 (1998 Twenty-Five District Roll Call Officer Survey Data) surveys were administered during Monday through Thursday afternoon roll call. The 3:00 p.m. and 4:00 p.m. roll calls were surveyed because these officers had the greatest involvement in CAPS, and officers who worked Mondays through Thursdays were most likely to be involved in community meetings and beat meetings. Generally the watch commander or sergeant in charge introduced the evaluation staff, who then explained the purpose of the survey and described its confidentiality. Officers were asked to put their completed questionnaires in a large envelope labeled "Northwestern University" before leaving roll call. For the most part, officers were receptive to the survey, but a few in each group refused to fill it out. Part 10 (1998 Beat Facilitation Training Officer Survey Data) surveys were administered to officers at beat facilitators training. Part 11 (1998 Beat Facilitation Training Citizen Survey Data) surveys were administered to civilian beat facilitators with questions similar to those in Part 10. Part 12 (1999 Beat Team Leader (Sergeant) Training Officer Survey Data) surveyed sergeants during beat team leader training. Part 13 (1999 Sector Team Leader (Lieutenant) Training Officer Survey Data) surveyed lieutenants during beat team leader training. Parts 14-20 and 46 contain data from eight city-wide sample surveys conducted for the CAPS evaluation. The surveys were conducted by telephone using random-digit dialing samples. Part 14 (1993 Citizen Survey Data) was conducted in English only and the results for Latinos should be interpreted with care. Parts 15-20 and 46 (1994 through 1999 and 2001 Citizen Survey Data) allowed respondents to choose to be interviewed in Spanish. These files contain a variable that identifies the language in which the interview was conducted. Parts 14-20 and 46 contain weight variables that can be used to correct the observed sample for household size and number of telephone numbers. Surveys for Parts 14-17 (1993 through 1996 Citizen Survey Data) were conducted by the Northwestern University Survey Research Laboratory. Surveys for Parts 18-20 and 46 (1997 through 1999 and 2001 Citizen Survey Data) were conducted by the Survey Research Laboratory of the University of Illinois-Chicago. Detailed methodology reports for Parts 14-20 and 46 from the principal investigators are provided in the PDF documentation for this data collection. Parts 21-26 contain data from two observational studies of beat community meetings in Chicago. Two kinds of data were collected: (1) systematic observations of the meetings, and (2) questionnaire responses from citizens who attended the meetings. Parts 21-23 contain data from beat community meetings conducted during 1995 and 1996. There was one observation in each of 165 beats, which were selected to provide an approximately 50-percent sample of all beats in the city (Part 21). There were 2,145 respondents to the participant survey (Part 22), and a sample of 291 respondents were later recontacted and reinterviewed after four months to assess their subsequent involvement in problem-solving activities (Part 23). Parts 24-26 contain data from beat meetings observed in 1998. There were a total of 459 observations of 253 beats. Part 24 contains the data from the main observation form, and Part 25 contains the coding of problems and issues discussed. These two parts can be linked by beat and date. For city-wide analysis, a weight variable can be used to weight cases by the inverse of the number of times they were observed so that the apparent number of cases is 253, the number of beats observed. There were also 5,293 participants surveyed in either English or Spanish (Part 26). A methodology report for the 1998 beat meeting observation study from the principal investigator is provided in the PDF documentation for this data collection. The approach for the data collection and difficulties in field operations described in this report were typical for the 1995 data collection as well. Parts 27 and 28 contain panel data from residential surveys conducted in the five original CAPS prototype districts and in a set of matched comparison areas. Part 27 contains the responses from Wave 1 of the resident surveys conducted in 1993. Part 28 contains data from the same residents who were reinterviewed a year later. To prepare for the evaluation, 1990 Census tract data were used to select large, mainly contiguous sections of the city that closely matched the demography of the five newly-announced prototype areas. These comparison areas were used to represent what would have happened in the prototype districts if there had been no CAPS program, for the program was not put in motion there until the end of the prototyping period. The matching factors for selecting the comparison areas were race and ethnicity, home ownership, and the percentage of residents living in buildings of ten units or more. Because none of the CAPS prototypes included large blocks of public housing, those areas of the city were avoided. Parts 29-32 contain data from four annual surveys of CAPS activists. The data from 1996 to 1998 were collected by the Institute of Policy Research, while the 1999 survey was conducted by the Metropolitan Chicago Information Center. The samples were convenience samples of activists identified in a variety of ways, and each year complete respondents from the previous year were retained again in the sample. Each year the scope of the sampling base and the size of the survey grew. Respondents were selected from members of the District Advisory Committees (DAC), chairs of DAC subcommittees, activists nominated by the Chicago Alliance for Neighborhood Safety (CANS), trainees at CANS biannual training sessions, registrants at city-wide workshops, beat facilitators, executive directors of community organizations, and nominees of other respondents. For Part 33, data were collected to evaluate a program to train citizens in problem-solving, in a joint effort between the Chicago Police Department and the Chicago Alliance for Neighborhood Safety. Questionnaires passed out at training sessions provided baseline data on trainees. Those questionnaires were also used to select respondents for follow-up telephone questionnaires, which were administered three months later (Part 34) to assess their later involvement in problem-solving activities. Parts 35-37 contain data from three neighborhood surveys. The survey in Part 35 was conducted by the Institute for Policy Research in 1997 as part of a problem-solving study in 15 Chicago neighborhoods. The survey in Part 36 was fielded in 1998 in 29 beats in anticipation of an effort by the city's CAPS Implementation Office to organize communities, increase attendance at beat meetings, and involve residents in problem-solving. Part 37 contains data from a 1999 survey of six beats with predominantly African-American residents. Parts 38 and 39 contain data coded from police reports of beat community meetings. Police attending each beat meeting completed a two-sided report form titled "Beat Community Meeting Log," which was then approved by their sergeant. Part 38 contains attendance information for every beat community meeting held between January 1995 and December 2000. For Part 39, three coders spent two months intensively coding 2,580 beat community meeting logs for 1998. They recorded basic information about the meeting, who attended, and what was talked about. There are data for 277 of 279 beats, because no one attended the few meetings held in two completely nonresidential beats in the central business district. A few other nonresidential beats did manage to attract business owners and location association members, and they are also included in this data collection. Beats that were subdivided and held regular sub-beat meetings appear in the data file as separate records with the same beat number, and they are flagged by a sub-beat variable. Some beats also held joint meetings on occasion, and the data for these meetings are duplicated under each beat number in the data file. The data file contains a weight variable to adjust the data for multi-beat meetings. In Part 40, survey data were gathered from hundreds of organizational informants on the roles that their groups played in Chicago's community policing program in 1994. It documented how they mobilized to influence the shape of community policing in the city's five prototype areas. The goal of the survey was to determine to what degree community organizations in the five prototype districts were involved in and were promoting CAPS during its first year of implementation. In particular, the survey was designed to capture information on differences in CAPS involvement among various community organizations, and among the five prototype districts in which it was being tested. The operational definition of a community organization was a turf-based group with a name. The survey was budgeted to yield data on 50 organizations in each of the five prototype districts, or a total of 250 organizations. Each organization was to be represented by two respondents who were knowledgeable about their group. When they were available, the names of individual respondents associated with each organization were included on a call sheet for that group. These respondents were often the heads of the organizations, or other high-level personnel. When there was no valid name on the call sheet, interviewers were instructed to ask for the person who was most knowledgeable about the organization, perhaps the executive director or president of the organization. Interviewing was conducted by graduate students, professional interviewers, an Illinois Criminal Justice Information Authority (ICJIA) staff member, a temporary employee, and the project manager. Parts 41 and 42 contain counts of business establishment types for Chicago aggregated to the beat level and district level, respectively. These data were compiled from a proprietary database of 94,058 business address listings purchased from Claritas Corporation, dated circa 1998. Parts 43 and 44 contain land-use data aggregated to the beat level and district level, respectively. These data were aggregated from a parcel-level land-use file belonging to the city of Chicago. Part 45 provides a crosswalk between Census and Chicago Police Department area designations. The file is organized so that each record is a census block group.

Sample View help for Sample

For Parts 1-13, respondents consisted of all police officers attending respective meetings or training sessions. For Parts 14-20 and 46, all households were reached via random-digit dialing, a standard telephone survey sampling technique that allows interviewing to be done with households regardless of whether or not their telephone number is listed. Use of this technique is important to gather representative data, as over one-third of the households had unlisted telephone numbers. In each household where an interview was conducted, one adult, 18 years of age or older, was interviewed. These individuals were randomly selected using the standard "last birthday" technique. The PDF documentation for this data collection contains methodology reports from the principal investigator that explain the sampling procedures in greater detail. In Parts 21-23, observational and participant survey data were collected for meetings in 165 randomly selected beats, approximately 50 percent of all beats in the city. For Parts 24-26, observational and participant survey data were collected for meetings in 253 of 270 residential beats in the city. For Parts 27 and 28, the researchers were targeting relatively small geographical areas that did not match telephone company switching areas, so it was necessary to develop a sampling plan that reached the target areas in a cost-effective fashion. To accomplish this, half of the sample numbers for each area were selected at random from an up-to-date reverse telephone directory. Because they were listed numbers, the investigators could ensure that those households lay within the targeted areas. The remaining sample of telephone numbers was selected by randomly scrambling the last digits of the listed numbers. Households responding to those numbers were then asked a brief series of geographical screening questions to ensure that they were within the targeted areas. In the end, there were 1,294 completed randomly-sampled interviews and 1,278 listed-number interviews, for a total of 2,572 interviews. For Parts 29-32, respondents were selected by convenience and snowball sampling. For Part 33 beat training sessions were randomly selected and all participants in the selected sessions were surveyed. The sample also included all District Advisory Council members for the district. Part 34 consists of a random subset of the respondents surveyed in Part 33. For Parts 35-37, the telephone number samples for each area were randomly drawn from a recent reverse telephone directory to ensure that the numbers were within beat boundaries. Part 38 contains data for all beat community meetings held between 1995 and 2000. Part 39 contains data for beat meetings held in 277 of 279 beats in 1998. For Part 40, a list of turf-based named groups, along with the names and telephone numbers of persons associated with them, was developed from several sources. A start-up list was contributed by a research team at DePaul University, which shared an inventory they had developed for their own community organization study. Northwestern and Illinois Criminal Justice Information Authority (ICJIA) staff members supplemented this list with organizations and contacts they came across while attending various CAPS-related meetings, or doing interviews. Each time an interview was conducted with an organizational informant they were asked about other organizations that they had run across, to contribute a snowball component to the sample. The investigators also culled local news articles and newsletters to locate additional organizations and named contacts to add to the list. Several key community contacts in prototype districts were asked to go through the lists for their areas and make additions and corrections to them. As part of the evaluation of CAPS, team members conducted interviews with district commanders, neighborhood relations officers, beat team members, and other officers. Among other topics, lists of group names and contacts were gleaned from these key informants. There was no sampling for Parts 41-45.

Universe View help for Universe

Residents and police officers of the city of Chicago from 1993 to 2001.

Unit(s) of Observation View help for Unit(s) of Observation

Data Source View help for Data Source

Data Type(s) View help for Data Type(s)

Description of Variables View help for Description of Variables

Parts 1-6 include variables about general job satisfaction and features of the officers' work, features they wanted to have in a job, the effectiveness of various patrol activities, the desired allocation of resources to traditional and new modes of service, how they spent their time, the activism of neighborhood residents, sources of information about their beat, their perceptions of police work, their personal qualifications for CAPS, and what they anticipated would be the impact of CAPS on assorted department activities. Additional variables specify sex, race, marital status, age, years on assignment, years with department, age when joining the department, rank, shift, education, and district assignment. Part 4 also includes a nine-item quiz of officers' knowledge of CAPS procedures and concepts, their prototype experience, and their perceptions of summer 1994 roll-call training. Part 7 provides information from officers on public and department support for CAPS, how realistic assumptions were about supervisors, beat team officers, and the community, beat plans, Information Collection for Automated Mapping (ICAM) and city services, beat problems, beat team meetings, and a grade for their personal CAPS activities. Additional variables specify sex, race, age, rank, and district and beat assignments. Parts 8-10 and Part 13 describe officers' attitudes toward beat meetings, their roles at the meetings, their perceptions of the community and problem-solving, the implementation of various elements of CAPS, their own problem-solving activities, the roles of their supervisors and their contacts with organizers and trainers. Also provided are officers' sex, race, and age. Part 11 includes the same variables as Part 10, as well as items about citizens' contacts with various police, beat, and district advisory committees, and contact with organizers and trainers. Other variables cover sex, age, race, length of residence, housing tenure, and education. Part 12 includes items similar to those in Part 10, as well as variables on public and department support for CAPS, beat team meetings, the roles of supervisors, views of the realism of assumptions about the capabilities of beat team officers and the community, beat and district plans, the use of ICAM and city service request forms, sources of information about problems, a self-rating of officers' own CAPS activities, when officers became beat team leaders, and what their regular job assignments and district assignments were. Variables in Parts 14-20 and 46 include how long the resident had lived in Chicago, types of problems in the neighborhood, satisfaction with quality of life in the neighborhood, evaluations of police performance, knowledge of community policing, and sources of news and other information. Demographic variables include age, sex, race and ethnicity, marital status, employment status. Political orientation is provided (only in Parts 14-20). Variables in Parts 21, 24, and 25 include number of beat meeting participants by race, number of police officers by rank, number of representatives from block clubs, organizations, or political offices, duties for which residents and police officers were responsible, characteristics of police and citizens relations, types of problems discussed, solutions proposed, and actions taken, who led the meeting, whether there was an agenda, who dominated the discussion, whether crime reports or maps were available, and the types of information requested for the next meeting. Variables in Parts 22, 23, and 26 include evaluations of the abilities of police and citizens to address community problems, citizens' expectations of police, involvement in community groups and other organizations, how citizens learned about beat meetings, and types of problems in neighborhood. Demographic variables include sex, age, race, years of residence, household income, type of education completed, housing type, and language of interview. Variables in Parts 27 and 28 include whether residents felt they were part of a neighborhood, their satisfaction with the neighborhood, problems in the neighborhood, whether residents participated in crime prevention meetings, knowledge about crime prevention programs, how safe they felt in their neighborhood, opinions about police, contact with the police, and crime victimization within the last year. Demographic variables include age, sex, race, employment status, marital status, household composition, educational attainment, household income, and language of the interview. For Parts 29-32 CAPS activists were asked how long they were involved with CAPS, how they rated the regularity and effectiveness of beat meetings, how they rated District Advisory Committees and subcommittees, and whether the police had improved their performance in terms of responsiveness to community concerns, working with residents, crime prevention, and their general level of concern regarding the community. Demographic variables include age, sex, race, education, home ownership, and police district. Variables in Part 33 include opinions about the abilities of police officers and citizens to solve problems, police expectations about citizens and citizens' expectations about police, citizens' willingness to take responsibility for problems, willingness to encourage other citizens to attend beat meetings, and participation in block clubs or other organizations. Demographic variables include age, sex, race/ethnicity, years of residence, type of housing, household income, and education completed. Part 34 variables repeat those in Part 33, and also include the number of beat meetings attended, how respondents heard about beat meetings, satisfaction with training sessions, whether respondents kept in touch with other participants from the training sessions, whether they tried to teach others about community policing, contacts with neighbors, police, city services, and politicians, problem-solving activities, community relations, and level of involvement in the community. Demographic variables include gender, marital status, number of children, labor force status, and language of questionnaire. In Parts 35-37, information is supplied on residents' views of neighborhood problems, their awareness of CAPS, their involvement in beat meetings, and reports of the quality of police service in their community. Other variables measured features identified as important components of the neighborhoods' capacity to deal with problems: residents' participation in community-based organizations and their perceptions of willingness of their neighbors to intervene to reestablish order. Part 40 presents information from organization leaders about when the organization had been established, by whom it was founded, its goals and activities, types of meetings held, contact with the media, fund-raising, CAPS-related activities, opinions about police, sources of funding, office space, number of staff, demographic composition of members, and the mission of the organization. Parts 41 and 42 contain counts of different types of business establishments, such as groceries, banks, restaurants, insurance sales, auto, sewage, liquor, and construction, as well as ranks of businesses based on size, sales, and employees. Parts 43 and 44 contain information on the percentage of land used for residential, commercial, or manufacturing purposes, buildings, parks and recreation, education and culture, religion, and parking, as well as the condition and age of the buildings. Part 45 contains the census block group number, census tract number, district beat number, police district number, and Chicago community area number. This is a crosswalk file linking census and police areas.

Response Rates View help for Response Rates

Parts 1-13, 21, 22, 24-26, 33, 34, 38, 39, 41-45: Not applicable. The response rates for the citizen surveys were: 60 percent for Part 14, 58.8 percent for Part 15, 74 percent for Part 16, 65 percent for Part 17, 44.2 percent for Part 18, 41 percent for Part 19, 42.7 percent for Part 20, and 42 percent for Part 46. The response rate for Part 23 (1995 Beat Meeting Participant Recontact Data) was 73 percent. The response rate for the panel survey data was 59.6 percent for Wave I (Part 27) and 59 percent for Wave II (Part 28). The response rates for the 1996 to 1998 activist surveys (Parts 29 to 31) are unknown. The response rate for the 1999 CAPS Activist Survey (Part 32) was 69.2 percent. The response rate for the 1995 Citizen Training Follow-up Survey Data (Part 34) was 61 percent. The response rates for Parts 35 and 37, the 1997 and 1999 Neighborhood Survey Data, are not known. However, the response rate for 1998 (Part 36) was 56 percent. The response rate for the 1994 Community Organizations Data (Part 40) was 81.6 percent. Parts 41 to 45: Not applicable.

Presence of Common Scales View help for Presence of Common Scales

Several Likert-type scales were used.

Original Release Date View help for Original Release Date

2002-10-11

Version History View help for Version History