NEIGHBORHOODS, LAND-USE, AND ROBBERY RATES: A TEST OF ROUTINE ACTIVITY THEORY
Open Access
- Graduate Program:
- Crime, Law and Justice
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- August 23, 2002
- Committee Members:
- D Wayne Osgood, Committee Member
- Richard B Felson, Committee Chair
- Stephen Matthews, Committee Member
- Thomas J. Bernard, Committee Member
- Keywords:
- GIS
- routine activity theory
- robbery
- spatial regression
- neighborhoods
- Abstract:
- Routine activity theory suggests that neighborhood-level activity patterns influence crime rates, and that the convergence of three elements in space and time—a motivated offender, a suitable target, and the absence of a capable guardian—result in increased likelihood of criminal events. Opportunities for crime increase when neighborhood land-use patterns are conducive to crime. Criminogenic land-uses include intermixed patterns of residential, commercial, industrial, and vacant lands within neighborhoods, as well as the presence of particular establishments, such as shopping mall. Routine activity theorists suggest that criminogenic land-uses influence crime in two ways: (a) by inhibiting an area’s social control capacity, and (b) by attracting particular types of routine activities (e.g., consuming alcohol at a bar, selling/using drugs in abandoned structures). These land-use patterns may explain why disadvantaged neighborhoods have higher crime rates than more advantaged areas. This dissertation examined whether the effect of neighborhood disadvantage on crime may be a function of its association with criminogenic land-use patterns. This research also examined whether criminogenic land-uses have greater effects in disadvantaged neighborhoods. By understanding the effects of land-use, planners and managers could make changes in land-use patterns that decrease crime rates. This dissertation examined the possible relationships between criminogenic land-use and crime, as measured by calls-for-service to the police. To address these issues, this research uses census and tax parcel data from three cities (Lincoln, NE; Columbus, OH; and San Antonio, TX), which vary in terms of size and racial composition. Within each city, census data are being used to create measures of neighborhood social composition, including concentrated disadvantage, population density, and residential mobility. In addition, land-use diversity indices, created from tax parcel data, are used to indicate the degree to which single-family residential, multiple-family residential, commercial, industrial, and vacant or abandoned lands are intermixed within neighborhoods. Land-use diversity is measured using the Shannon Index, where the highest value occurs when equal areas of all types of land-use are present in a neighborhood. The three cities vary in both the size of their populations and their social composition. Thus, this dissertation also examined whether the effects of land-use are consistent across the three research sites. One problem with prior research on communities and crime is that it has largely ignored spatial autocorrelation—the fact that locations that are near to each other are also likely to have similar levels of poverty, residential instability, and crime. Because ordinary regression models cannot control for spatial autocorrelation, traditional estimates are biased and inferences based on these models are likely to be incorrect. To address this problem, this study uses state-of-the-art spatial regression analytic techniques.