Community Policing in Madison, Wisconsin: Evaluation of Implementation and Impact, 1987-1990 (ICPSR 6480)

Time Period(s) View help for Time Period(s)

Study Purpose View help for Study Purpose

This study sought to evaluate efforts by the Madison, Wisconsin, Police Department to create a new organizational design (both structural and managerial) to support community-oriented and problem-oriented policing that would result in better, more responsive service to the community. The department's plan was a sequential one: internal organizational changes were necessary before the external goal of improved service could be accomplished. One-sixth of the organization serving approximately one-sixth of the community was used as a test site for the new approach. This Experimental Police District (EPD) was charged with implementing "quality policing," which emphasized quality of service delivery, quality of life in the community, and quality of life in the workplace. The first objective of the Madison Police Department was the implementation of three conditions: (1) quality leadership, emphasizing the role of managers as facilitators whose job was to improve systems, involve employees in decision-making, employ data-based problem-solving approaches, promote teamwork, encourage risk-taking and creativity, and give and receive feedback from employees, (2) a healthy work environment, treating employees as "internal customers" whose problems should be identified and resolved, and (3) physical decentralization, creating a small work group to improve conditions in the workplace and, at the same time, obtain closer physical proximity to citizens to get to know them and become aware of their problems. The researchers' task was to: (1) document the process of developing the Experimental Police District, (2) measure any attitude changes of the police personnel during the experimental period, and (3) measure the effects of change on the community.

Study Design View help for Study Design

As the first part of the program evaluation, attitude changes among officers working in the EPD were compared with those of officers working in the rest of the police department. Written surveys were administered by the Project Director to small groups of personnel during normal working hours. Although the analysis was to be based on a panel design, efforts were made to survey all commissioned personnel during each survey administration period. Since the ultimate goal of the department was change across the entire organization, changes for the organization as a whole were also monitored. Part 1, Commissioned Personnel Data, Wave 1, contains responses from 269 commissioned personnel surveyed in December 1987 before the creation of the EPD. Part 2, Commissioned Personnel Data, Wave 2, consists of responses from 264 police officers who completed a Wave 2 survey in December 1988, and Part 3, Commissioned Personnel Data, Wave 3, supplies responses from 230 police officers who completed a Wave 3 survey in December 1989. As the second part of the program evaluation, attitude changes among residents served by the EPD were compared with those of residents in the rest of the city using a quasi-experimental design. These data are presented in Part 4, Residents Data, Waves 1 and 2. A few days prior to contact, letters from the Office of the Mayor were sent to the selected addresses. Interviewers who had been recruited, trained, and supervised by Police Foundation personnel carried a copy of that letter and presented photo identification cards at each residence. Selection of the respondents was made by the interviewers at the selected household, using a Kish selection table included in each questionnaire. Individuals under the age of 18 were not included in the household listing. Interviewers made a total of six attempts to interview the selected respondent in each household. All refusals in which the respondent was not hostile were reassigned to different interviewers. Twenty-five percent of all completed interviews were validated by recontacting the respondent to verify that the interview took place, that it had required the appropriate amount of the respondent's time, and that a few key questions were answered the same way during the validation call as in the original contact. Data for Wave 1 consist of personal interviews with a random sample of 1,166 Madison residents in February and March 1988, prior to the opening of the EPD station. Since 97.6 percent of the respondents provided their telephone numbers, the decision was made to conduct the Wave 2 survey by telephone. The telephone interviews were conducted by the Wisconsin Survey Research Lab at the University of Wisconsin in February and March 1990. In-person interviews were also attempted with about 70 percent of the Wave 1 respondents who did not provide telephone numbers. Of the 772 completed interviews for Wave 2, 45 interviews contained substantial mismatches between information provided in 1988 and 1990. These 45 respondents were removed from the panel, leaving an analysis panel of 727 respondents.

Sample View help for Sample

The EPD program site was not randomly selected, but was selected by the department, based on several indicators of need. Police officers were also not randomly assigned to work in the EPD, but were allowed to bid for assignments in the EPD. Households for the resident survey were randomly selected from the 1980 Census block statistics, excluding city blocks that consisted primarily of business areas or student housing.

Universe View help for Universe

Police officers in the Madison Police Department and residents of the city of Madison, Wisconsin.

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

Police personnel provided their assessments on how successfully quality leadership had been implemented, the extent to which the officers felt they worked closely with and received feedback from other officers, the amount of their interaction with detectives, the amount of time available for problem-solving, ease of arranging schedules, safety of working conditions, satisfaction with working conditions, type of work they performed, their supervisors, commitment to the department, attitudes related to community policing and problem-solving, perception of their relationship with the community, police views of human nature, attitudes toward change, attitudes toward decentralization, and demographic information. Residents provided their perceptions of police presence, frequency of police-citizen contacts, quality of police-citizen contacts, estimates of the magnitude of various problems in their neighborhoods, evaluation of the problem-solving efforts of the police, perception of neighborhood conditions, levels of fear of crime, personal experience of victimization, knowledge of victimization of other residents, and demographic information.

Response Rates View help for Response Rates

For the police personnel surveys, 97 percent of the total commissioned personnel in the Madison Police Department participated in the employee survey in 1987, 97 percent participated in 1988, and 86 percent participated in 1989. Of the respondents to the Wave I survey, 14 had left the department by the time of the third survey. Two hundred and two persons participated in all three survey waves, resulting in a participation rate of 79 percent for the panel. Some personnel changes occurred after the first year of the evaluation period, resulting in an analysis panel equivalent to 61 percent of the total sworn personnel at any one of the three survey times. Participation rates are provided rather than response rates because at each survey period, a very small number of individuals came to the survey site and completed a survey identification form, but did not actually complete the survey. For the resident survey, the response rate in the EPD area was 77.8 percent and 75.1 percent for the rest of the city. The 772 interviews completed for Wave 2 resulted in a panel completion rate of 66.2 percent.

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

1996-06-10

Version History View help for Version History