Physical Environment and Crime

Connecting physical environment, crime, and crime prevention is a huge topic. Broader literature reviews appear in the Annotated Further Readings of this volume. This entry narrows the focus by concentrating on the question of causal impacts at the neighborhood or streetblock level, as well as contextual impacts on individuals or households. Further, this entry notes problems with inferring causal impacts of physical environment features. It also highlights difficulties in clearly demonstrating ecological connections between features of physical environment—or other features of neighborhood fabric, for that matter—and hypothesized mediating social, behavioral, or psychological processes taking place at the small group or community level.

The focus of the current entry is a response to questions raised by Robert Sampson, Per-Olof Wikström, and others about whether, despite the many recent statistical and methodological advances in communities and crime research, criminologists are actually any closer to demonstrating neighborhood effects on crime which are causal in nature. A case can be made that conclusive demonstrations of causal neighborhood effects involving physical environment features affecting crime, delinquency, or victimization have not yet appeared and are extremely unlikely to ever appear.

At least since the mid-1800s, hopes have run high that changes in the physical environment of a neighborhood or a street would reduce social problems, including delinquency, offending, and victimization rates. Tearing down a neighborhood of slum housing in the early 20th century, or destroying blocks of alley housing as part of mid-20th-century urban renewal, or dynamiting high-rise public housing communities at the end of the 20th and beginning of the 21st centuries, have clearly had effects on the social problems, offending, delinquency, and victimization rates in those locations. The location-based rates have shifted in part because the original households were no longer there, and the replacement households often had different household structures, held different positions in the broader structure of society, and were more spread out spatially. Although the drastic changes in the physical environment were clearly a precondition for the other structural changes, parsing out how much of the social problem rate reductions in those places arose from physical environment changes per se, as compared to the other changes including population shifts, is extraordinarily difficult if not impossible. Estimating physical environment impacts when physical environment changes are less dramatic, or are ongoing, are similarly difficult. The sections below explain in more detail why.

These controversies were highlighted in a trio of articles authored by Jens Ludwig, Susan Clampet-Lundquist and Doug Massey, and Robert Sampson in the July 2008 issue of the American Journal of Sociology. The authors debated whether a recent randomized experiment involving the relocation of households eligible for subsidized housing had or had not demonstrated an impact on delinquency, victimization, and related outcomes. No consensus emerged.