Three different conceptual models were tested (experience with police, quality of life, and neighborhood context) for directional accuracy and ability to explain satisfaction with the police, and the investigation also looked at whether these models helped explain the common finding that blacks were more dissatisfied with the police than whites.
Hierarchical linear modeling was used to analyze data obtained from the Project on Policing Neighborhoods. The data file consisted of survey responses from 5,361 citizens residing in 58 neighborhoods located in Indianapolis and St. Petersburg, Florida. Interviews with respondents showed, at the citizen level, the psychologically based quality of life model accounted for the greatest proportion of explained variance and provided the greatest directional accuracy. Also, residents of neighborhoods characterized by concentrated disadvantage expressed significantly less satisfaction with the police. In addition, neighborhood context reduced the negative effect of black status on satisfaction when the police when a sparse citizen-level specification was used. Racial variation in satisfaction with the police persisted, however, when citizen-level hierarchical models were specified more fully. Policy implications of the findings are considered, particularly with respect to the use of attitudinal measures as indicators of police performance. A zero-order correlation matrix of citizen-level independent variables is appended. 33 references and 5 tables