Crime Prevention in Residential Areas A Systematic Review and Meta-Analysis of Environmental Design Approaches
摘要
Crime prevention in residential areas is essential for fostering safety and improving everyday well-being. Situational interventions offer practical tools for reducing crime opportunities, yet the evidence base for their effectiveness in residential settings remains dispersed. This study systematically reviews and synthesizes evaluations of environmental design interventions implemented in residential environments. Searches were conducted in three major databases and supplemented with backward citation searches, following PRISMA guidelines. Thirty-one studies met the inclusion criteria, and 29 provided sufficient data to calculate Relative Effect Sizes (RES). Random-effects meta-analyses were performed for violent and property crime, and heterogeneity and publication bias were assessed through standard diagnostic tools, including exploratory meta-regression. Intervention types were grouped into conceptual classes aligned with CPTED principles. Pooled analyses showed a statistically significant reduction in property crime overall. Target hardening and access control were associated with consistent decreases in property offenses, while demolition produced substantially larger effects across crime types, though based on limited evidence. Effects on violent crime were highly variable and not statistically significant on average. Across all analyses, heterogeneity was substantial, indicating wide contextual variation in intervention impact. The findings suggest that certain environmental design interventions may reduce crime in residential areas, particularly property offenses. However, high heterogeneity and methodological variation across studies warrant cautious interpretation. Crime-prevention effects appear to depend on intervention type and contextual conditions. Future research should prioritise more consistent outcome reporting, stronger research designs, and systematic examination of contextual and implementation factors. No trial registration was applicable for this study.