Study 1
Schneider (1986) analyzed three sets of data using a preprogram–postprogram comparison. To estimate the proportion of burglaries reported to the police, two victimization surveys were analyzed. The first survey was designed specifically for the evaluation and included questions on knowledge of and participation in crime prevention activities, attitudes toward crime, actions taken to avoid being victimized, and other various questions related to burglary prevention. The survey was administered in 1974 and had a recall period of May 1973 to April 1974. The interviews were conducted in person; approximately 3,950 interviews were conducted. Out of the total interviews, 1,909 were conducted within Portland, OR’s city limits; the remainder were conducted in suburban Portland. Two high-crime areas were selected as high-priority areas: the Street Lighting Area of Portland and another Crime Prevention Bureau (CPB)–designated high-priority area.
The second victimization survey contained previously collected data, which was obtained by the Law Enforcement Assistance Administration (LEAA) in 1972. However, Census Bureau rules prohibited the data from being divided into subareas within Portland. The rules also prohibited the Bureau from providing individual-level data, which could have been used for a historical control group. While there were some limitations in using this data in the analysis, the data was used for a pre–post examination of change in citywide burglary rates.
The third set of data used in the evaluation was police statistics on offense rates for burglaries and other Crime Index offenses for the entire city.
The evaluation sought to compare the private benefits of the program and the collective benefits of the program. Further, the displacement effect was also taken into account when conducting the evaluation. If certain residents take precautions to protect themselves, burglars may simply move to easier targets—whether they are in the same neighborhood or in other nearby areas. This would not actually reduce the volume of crime, it would simply be moved to other areas. To analyze possible displacement effects, subareas of Portland were included in the evaluation.
The private benefits were assessed by analyzing the effects of participating households compared with households that did not participate in the program. This was done using the victimization survey designed for the evaluation. Households were denoted as participants if they displayed antiburglary decals outside of their homes. Data on participants and nonparticipants who lived in the same section were compared, and—to isolate the effect of the program—statistical controls were introduced for other variables related to participation and burglary rate. The information was obtained by calculating the percentage of homes with stickers that had been burglarized one or more times after the stickers were displayed. The number of months of opportunity for burglaries to occur was calculated (based on dates when the stickers were displayed), and the rate was then adjusted to a yearly equivalent. To supplement the evaluation of private effects of the program, survey items that specially benefited the individual (as opposed to the neighborhood overall) were analyzed. Multiple regression analysis was then used to assess the independent effect of attending block meetings on these types of activities.
Collective benefits were assessed by estimating the change in burglary rates for the entire city of Portland. Since no baseline victimization data was available for specific areas within the city, preprogram citywide burglary rates were calculated using data from the LEAA victimization surveys. Postprogram victimization rates were assessed using the second victimization survey designed specifically for the evaluation. Using this data, survey victimization rates before implementation of the program (preprogram rates) were compared with survey victimization rates after the implementation of the program (postprogram rates). In addition, official police burglary statistics for preprogram and postprogram rates were analyzed for the entire city and adjusted for differences in reporting percentages as reflected in the victimization surveys.