NCJ Number
101865
Date Published
1985
Length
112 pages
Annotation
This study examined the effects of community context on recidivism and the impact of criminal offenders on their communities.
Abstract
The principal hypothesis of this study is that an analysis of the interaction between offender and community characteristics will increase the accuracy of recidivism prediction beyond that attained by separate analyses of the two variable types. To obtain information on community contexts, the physical attributes of 90 Baltimore neighborhoods were determined through onsite assessment. For 66 of the neighborhoods, additional information was obtained from the 1980 census and from telephone interviews with a sampling of residents. Data were obtained on all offenders released to the sample of neighborhoods between October 1978 and October 1980, and followup studies through January 1982 determined any arrests of offenders to that point. Because offenders were not released to all of the communities studied, analyses based on census and survey information were limited to 487 offenders in 57 neighborhoods. Analyses based on the physical assessment of neighborhoods involved 619 offenders in 67 neighborhoods. The analysis used covariance by regression methods (Cohen and Cohen, 1975). Offender characteristics predominated in determining recidivism, with community context having a relatively small effect on outcome. The impact of offenders on the community, however, was significant, as their presence affected residents' perceptions of the neighborhood climate and their fear of crime. Tabular data and approximately 250 references.