The Spatial Extent of the Effect of Foreclosures on Crime 
 
Seth B. Payton 
Assistant Professor 
School of Public and Environmental Affairs 
Indiana University Public Policy Institute 
Indiana University–Purdue University Indianapolis 
 
Thomas D. Stucky 
Associate Professor, 
School of Public and Environmental Affairs 
Indiana University Public Policy Institute 
Indiana University–Purdue University Indianapolis 
 
John R. Ottensmann 
Professor Emeritus 
School of Public and Environmental Affairs 
Indiana University Public Policy Institute 
Indiana University–Purdue University Indianapolis 
 
 
 
 
Contact Information: 
 
Seth B. Payton, Ph.D. 
Assistant Professor 
School of Public and Environmental Affairs 
BS/SPEA Building 4076 
801 W. Michigan St.  
Indianapolis, IN 46202  
sbpayton@iupui.edu 
(317)278-4898 
 
Keywords: externalities, foreclosures, crime, spatial analysis 
 
This is the author's manuscript of the article published in final edited form as:  
Payton, S. B., Stucky, T. D., & Ottensmann, J. R. (2015). The spatial extent of the effect of 
foreclosures on crime. Social science research, 49, 288-298.  
http://dx.doi.org/10.1016/j.ssresearch.2014.08.012 
 
1 
 
Abstract: 
Although neighborhood stability has long been considered a substantial determinant of crime, 
foreclosures have not been the subject of concerted research among criminologists until recently.  A 
number of recent studies have examined the linkage between home foreclosures and crime.  Though 
generally finding a significant relationship, studies have used different approaches and units of 
analysis.  This variation led us to examine the spatial extent to which foreclosures affect a relatively 
small surrounding area.  In this paper, we consider the spatial extent of the foreclosure effect on 
crime by estimating fixed effect negative binomial models using geocoded UCR data for 2003-2008 
and foreclosure data to predict crime counts using the number of foreclosures within various small 
area radii. Results show that, independently and jointly, foreclosures are a predictor of crime up to at 
least a distance of 2,250 feet. Importantly, that effect declines with distance. We conclude with a 
discussion of the implications of those findings. 
 
 
 
1.Introduction 
Scholarly interest in the external effects of foreclosures on neighborhood conditions has 
increased over the past decade. The general consensus is that foreclosures contribute to 
neighborhood instability.1 Until recently, scholars have primarily focused on the external effect 
of foreclosures on property values (Immergluck and Smith, 2006a; Schuetz, Been, and Ellen, 
2008; Harding, Rosenblatt, and Yao, 2009; Lin, Ronsenblatt, and Yao, 2009; Leonard and 
Murdoch, 2009). However, a rapidly expanding strand of scholarly research has examined the 
effects of foreclosures on crime (Acevedo, 2009; Arnio and Baumer, 2012; Bess, 2008; Baumer, 
Wolff, and Arnio, 2012; Ellen, Lacoe, and Sharygin, 2013;Goodstein and Lee, 2010; Harris, 
2011; Immergluck and Smith 2006b; Jones and Pridemore, 2012; Katz, Wallace, and 
Hedberg2013; Kirk and Hyra 2012; Stucky, Ottensmann, and Payton 2012; Wallace, Hedberg 
and Katz, 2012; Williams, Galster, and Verma, 2013; Wolff, Cochran, and Baumer, 2013). Most 
research finds that more foreclosures are related to higher levels of crime. However, many of the 
studies examining the relationship between foreclosures and crime have employed relatively 
1 One example of a program acknowledging the neighborhood instability of foreclosures on neighborhoods is the 
federal Neighborhood Stabilization Program (NSP). The intergovernmental transfers of the NSP were directly 
related to relieving the negative neighborhood externalities of foreclosures (Immergluck 2011).  
                                                 
large units of analysis, such as counties (Arnio, Baumer, and Wolff, 2012; Goodstein and Lee, 
2010; Wolff, Cochran, Baumer, 2013) or census tracts (Immergluck and Smith, 2006b; Acevedo, 
2009). 
Larger units of analysis could mask the spatial extent of the foreclosure-crime 
relationship. Assessing the spatial relationship between foreclosures and crime at larger, 
commonly used, units of analysis (e.g., MSAs, Counties, and census tracts) is limited by at least 
three assumptions: (1) the relationship between two events is evenly distributed across the space 
of relatively large areas; (2) the relationship between two events is contained within and 
consistent with administrative or statistical boundaries of varying sizes and shapes; and (3) the 
magnitude of the effect of event x(e.g., foreclosures) on event y(e.g., crime) is not a function of 
distance between events. 
Some studies have used smaller units of analysis (e.g., blockfaces, uniform spatial 
lattices). However, those studies have confronted the question of the spatial extent of foreclosure 
effects on crime by making a priori decisions about the distance at which that effect occurs (Cui, 
2011; Ellen, Lacoe, and Sharygin, 2013; Stucky, Ottensmann, and Payton, 2012). The goal of the 
current study is to empirically test the distance within which foreclosures affect crime and the 
degree to which that effect decays with increasing distance. Specifically, our study examines the 
spatial extent of the effect of foreclosures on crime in an urban area by using a lattice of 500 x 
500 feet square cells that are uniformly distributed across space and radii of increasing distance. 
 Consistent with most previous studies, we generally find a significant and positive spatial 
relationship between foreclosures and crime. The magnitude of that relationship is greatest 
within the immediate 500 x 500 feet area and declines as the distance to foreclosures increases 
from the unit of analysis. The magnitudes of point estimates are similar when total crime is 
dissected into property crime and violent crime. However, the independent point estimates are 
less precise for violent crimes. 
 Our approach directly contributes to the literature that examines the spatial extent of 
externalities. Similar to the housing literature that has addressed the external effect of 
foreclosures on property values; we suggest that criminal justice practitioners can benefit from 
understanding the spatial extent at which foreclosure may affect crime. In particular, our findings 
indicate that the foreclosure-crime relationship is greatest in magnitude within very close 
proximity (i.e., within a 500 X 500 feet area), but the substantive impact of foreclosures extends 
to at least 2,250 feet and potentially as far as 3,250 feet.  
There are several theoretical reasons foreclosures may be associated with crime. The 
following section examines the relevant theoretical and empirical literature linking crime and 
foreclosures, exploring why one might expect a small area effect of foreclosures that declines 
with distance. Then, we describe the data and modeling strategy for examining the spatial extent 
of that relationship. Next, we provide the results of our analysis. Finally, we discuss the 
implications of our findings. 
 
2. Relevant Literature Linking Crime and Foreclosures  
2.1 Theoretical Framework 
Recent studies of foreclosures and crime have suggested several theoretical reasons to 
expect that foreclosures in an area will be associated with increased crime nearby (e.g. Wilson 
and Paulsen 2008; Kirk and Hyra 2012;Wallace et al 2012; Ellen et al 2013; Katz, et al 2013; 
Williams et al 2013; Wolff et al 2013). Universally, the foreclosure process involves several 
steps that occur over an extended period of time.  To begin, the pre-foreclosure process is 
generally initiated when homeowners neglect timely mortgage payments.  Homeowners are 
usually forced to vacate a property after an extended period of neglected payments. The vacancy 
of those properties is variable as lenders of foreclosed, vacant, properties seek buyers or renters. 
Theories of crime focus on various aspects of the foreclosure process.For example, social 
disorganization theory posits that crime results from the inability of neighborhood residents to 
develop or maintain social cohesion and ties necessary to effectively maintain informal social 
control (Sampson and Groves, 1989; Shaw and Mckay 1972). From that perspective, 
homeowners facing foreclosure are likely to be under severe psychological stress and therefore 
may have reduced capacity or willingness to “invest” in maintaining neighborhood informal 
social control (Ellen et al 2013).  The social organization of the neighborhood posited to be vital 
for maintaining social control is further disrupted when homeowners are forced out of their 
house (Harris, 2011; Pandit, 2011). 
Similarly, broken windows theory suggests that physical disorder (e.g., deteriorating 
houses, unkempt lawns, visible trash and debris) sends a signal to would-be criminals that 
neighborhood residents care less about, or lack resources to effectively maintain, their 
neighborhood (Wilson and Kelling, 1982).Visible neighborhood deterioration may occur as 
homeowners facing foreclosure begin to defer maintenance on their properties (Katz et al 
2013).Visible signs of decay are likely to increase as homes become vacant. The appearance of 
decay may cause contagion effects within an area as neighbors withdraw from social life due to 
increased fear of crime (Katz et al 2013) or reduced identification with the “neighborhood” 
(Wallace et al 2012).   
 Foreclosures may also increase opportunities for crime.  Routine activities theory, which 
assumes that criminals are rational, suggests that offenders are drawn to accessible and desirable 
targets that lack adequate guardianship, reducing the likelihood of criminal apprehension (Cohen 
and Felson, 1979). The neglect and ultimate deterioration of foreclosed, or vacant, properties 
may send a signal to potential criminals that residents are less invested in the neighborhood 
(decreased guardianship) as the appearance of decay increases.Vacant homes may also be seen as 
suitable targets for theft or burglary (e.g., stripping of copper pipes or wiring), which could lead 
to higher crime in the immediate area (Spelman, 1993). Vacant homes may also become 
locations for squatters, drug-related crimes and prostitution (Ellen et al 2013). Arnio, et al 
(2012:1600) also note that foreclosures can cause shifts in secondary markets which can increase 
opportunities for property and violent crime as “more people come into contact with risky 
situations and by increasing the need for violent resolutions of unregulated transactions.” 
From these perspectives, one might observe negative impacts on a neighborhood at 
various points in the foreclosure process.  Homeowners facing foreclosure might begin to 
“withdraw” from community life, impacting informal social control. As homeowners withdraw, 
they might allow their houses to deteriorate. Property deterioration may lead to the perception of 
neighborhood decay, which may be associated with increased criminal activity. 
There also is potential for the foreclosure process to lead to extended periods of vacancy. 
Vacancy over an extended period of time may lead to visible signs of deterioration. Thus, theory 
suggests that foreclosures may generally lead to crime, but some research suggests that the 
effects of foreclosures on crime are greater when homes are unoccupied (e.g. Cui 2011; Ellen et 
al 2013). 
 
2.2 Prior Empirical Research 
 
Methods used to study the relationship between foreclosures and crimes have varied. 
Immergluck and Smith (2006b) began the recent surge of related scholarly research by 
examining foreclosures and other determinants of census tract crime rates in Chicago, Illinois. 
Their results suggest that foreclosures affect census tract violent crime rates but not census tract 
property crime rates. From 2009 through 2013, at least sixteen published studies have examined 
the relationship between foreclosures and crime using various methods and levels of aggregation.  
For example, studies have used Metropolitan Statistical Areas (MSAs)(Pandit, 2011); counties 
(Arnio, Baumer, and Wolff, 2012; Goodstein and Lee, 2010;Wolff, Cochran, Baumer, 
2013);police beats (Harris, 2011); census tract clusters (Kirk and Hyra2012, Williams, Galster, 
and Verma, 2013); census tracts (Arnio, and Baumer, 2012; Baumer, Wolff, and Arnio, 2012); 
and block groups (Katz, Wallace, and Hedberg, 2011; Wallace, Hedberg, and Katz, 2012).  With 
a few exceptions (Kirk and Hyra 2012; Wolff, Cochran, and Baumer, 2013), research generally 
finds some relationship between foreclosures and crime, although that relationship may be of 
fairly short duration (Katz et al 2013), vary by neighborhood context (Arnio and Baumer 2012), 
and vary by crime type.  Some studies find foreclosures affect property crime but not violent 
crime (e.g. Williams, Galster, and Verma 2013), violent crime but not property crime (e.g. 
Immergluck and Smith 2006b), or both (e.g. Stucky, Ottensmann, and Payton 2012).2 
Although extant studies have generally examined fairly large units of analysis such as 
counties or cities, or sub-city units such as census tracts (which can be large and often vary 
substantially in size), it is possible that the spatial extent of the effect of foreclosures is small 
enough that those larger units of analysis may obscure the estimated effect of foreclosures on 
crime.  We believe that the spatial extent of the effect of foreclosed residential properties on 
crime may extend only for relatively short distances. Our assertion is consistent with the few 
2 Some scholars argue that only certain types of crime (e.g., larceny and aggravated assault) occur as a result of 
foreclosures (Goodstein and Lee, 2010). Other scholars argue that social disorganization perspectives relate the lack 
of informal social control to all types of crime, including rape and murder (Kirk and Hyra, 2012). 
                                                 
recent studies using smaller units of analysis.  For example, Stucky, Ottensmann, and Payton 
(2012) employ negative binomial fixed effects (FENB) models to examine 1,000 x 1,000 feet 
grid cells in Indianapolis, Indiana. They find that foreclosures in contiguous grid cells have a 
positive effect on UCR reported violent and property crime (Stucky, Ottensmann, & Payton, 
2012). Using block faces as the unit of analysis and Ordinary Least Squares (OLS) and FENB 
models, Ellen, Lacoe, and Sharygin (2013) find that foreclosures on a blockface are associated 
with elevated crime in New York. Consistent with a potential distance decay effect, they note 
that the number of foreclosures in nearby blockfaces also increases crime to a lesser extent 
(Ellen, Lacoe, and Sharygin, 2013). Finally, Cui (2010) finds that foreclosures, after vacancy, are 
related to increased levels of both property and violent crime within a 250 feet radius when 
compared to an outer-radius (control effect) extending to 353 feet (250 feet * square root of 2) 
through a difference-in-difference modeling approach in Pittsburgh, PA.  The Cui (2010) 
approach is quite innovative and allows the researcher substantial confidence in isolating a 
causal effect of foreclosures on crime.  However, this approach assumes that the effect of 
foreclosures does not extend into the “control” area and the 250 feet radius appears to have been 
somewhat arbitrarily chosen. 
In summary, theorists linking foreclosures and crime have often focused on social 
disorganization theory, broken windows theory, and routine activities theory.  Both broken 
windows theory and routine activities theory suggest that the visible physical signs of decay 
associated with foreclosures (and ultimately vacant property in some instances) are expected to 
lead to increased perceptions of opportunity for crime and reduced surveillance.  However, if the 
visible appearance of disorder associated with foreclosed and/or vacant homes is a determinant 
of increased crime, it seems reasonable that this effect would be localized to within a small area 
around the foreclosed property. 
Similarly, social disorganization perspectives are often invoked as a mechanism that 
might link foreclosures and crime. If foreclosures lead to reductions in informal social control 
and collective efficacy due to the social and physical withdrawal of homeowners facing 
foreclosure, it seems reasonable that this effect is likely to be greatest in magnitude at relatively 
short distances from the affected home. However, the physical and social effects of foreclosures, 
especially if several are concentrated within a small area, may also have an effect on crime that 
extends beyond the immediate area. 
Examination of small area effects in crime is a robust area of research (see Weisburd, 
Groff, and Yang, 2012 for a review of recent research on the micro-ecology of place and crime).  
For example, Taylor (1997) argues that street blocks create important “behavioral settings” for 
crime.  Focusing on small areas around foreclosures would also make sense from the perspective 
of crime pattern theory, which argues that people operate within geographically finite 
“awareness spaces” (see Brantingham and Brantingham 1984, 1993), one of which is near their 
residence. Additionally, research has now well-documented that there are “hot spots” of crime 
which are typically very small areas (Sherman, Gartin, and Buerger, 1989). Tobler’s (1970) First 
Law of Geography, “Everything is related to everything else, but near things are more related to 
each other,” also suggests that the effect of foreclosures on crime can be expected to operate over 
finite distances. Yet, the extant research on foreclosures has not explored the spatial extent of the 
effect of foreclosures on crime.  Prominent theories of crime, though suggesting mechanisms for 
why they may be related, do not provide specific guidance on the distances at which one might 
expect to observe an impact of foreclosures on crime. Therefore, the goal of the current study is 
to examine the distance at which foreclosures affect crime and the degree to which that effect 
declines with increasing distance. 
 
3. Data and Methods 
 
3.1 Study Area and Unit of Analysis 
 
 We chose Indianapolis, Indiana to study the spatial extent of foreclosures on crime. The 
study area for this analysis is the former administrative boundaries of the Indianapolis Police 
Department (IPD). Several Marion County and City of Indianapolis government functions were 
consolidated in 1970. In 2008, IPD merged with the Marion County Sheriff’s Department. To 
maintain consistency over time, our data do not extend beyond 2008. The former IPD 
administrative boundaries essentially correspond with the old central city boundaries, which is 
the district in which most reliable longitudinal data are available for this analysis. Therefore, we 
use only data that were collected by IPD prior to consolidating two data management systems. 
Additionally, limiting the study area to the former IPD boundaries increases comparability to 
other Midwestern and Northeastern cities. 
We use a 500 x 500 feet grid cell unit of analysis to more effectively address the distance 
effect of foreclosures on crime. Map 1 illustrates the IPD service area and the grid cells used in 
the current study. Specifically, our dataset includes annual foreclosure listings and crime counts 
within 6,984 grid cells entirely contained within the former IPD service district. Our data span 
from 2003 to 2008. It should be noted that the study period of our analysis includes only the 
beginning of the recognition of the national housing crisis. In a national study, Immergluck 
(2011) found that the Indianapolis Metropolitan Statistical Area was among similar Midwestern 
metropolitan areas that experienced high initial foreclosures well before national attention on the 
2007 housing crisis. These data also allow us to employ a modeling strategy to mitigate annual 
unobserved heterogeneous effects. 
 
Map 1: Units of Analysis and Study Area 
Use of grid cells at the chosen resolution addresses the three limiting assumptions 
associated with larger, inconsistently sized units of analysis. First, smaller units of analysis (500 
x 500 feet grid cells) mitigate the issue associated with aggregated data being evenly distributed 
across space. Smaller units of analysis more closely represent the spatial pattern of the 
underlying data. Second, usage of uniform grid cells addresses the limitation of using 
administrative or statistical boundaries that may be inconsistent with the spatial relationship of 
the phenomena being tested.  Finally, the examination of the relationship of foreclosures and 
crimes within relatively small units of analysis (500 x 500 feet grid cells) allows for the test of 
distance effects within relatively short distances between two events (i.e., foreclosures [x] and 
crimes [y]).  
 
3.2 Crime Data 
Given the differences in past findings on violent versus property crimes of previous 
studies, we examine whether or not the distances at which foreclosures affect crime differs by 
crime type. Geocoded Uniform Crime Report (UCR) point data were obtained from the 
Indianapolis Police Department. We examine total, violent, and property UCR crime counts 
within each 500 x 500 feet cell.  Violent crimes include homicide, rape, robbery, and aggravated 
assault.3 Property crimes include burglary, larceny, and motor vehicle theft.  Total crime 
3 As one reviewer noted, rape counts in the UCR data suffer from under-reporting. Rapes represent a very small 
percentage of reported UCR violent crimes, however.  Additionally, we do not have any reason to believe that 
                                                 
includes all property and violent offenses.  We calculated the number of crime and foreclosure 
incidents per cell, by year. Because property crimes occur much more frequently than violent 
crimes, total crime counts are relatively highly correlated with property crime counts. 
 
3.3 Foreclosure Data 
The foreclosure data are geocoded point locations of real estate owned (REO) residential 
properties listed for sale in the multiple listing service (MLS) database maintained by the 
Metropolitan Indianapolis Board of REALTORS (MIBOR). REO properties are properties for 
which lending institutions retain ownership after not selling at a foreclosure auction. REO 
properties also are the type of foreclosures (i.e., high probability of vacancy) that have had the 
greatest effect on crime in a previous study (Ellen, et.al.2013). REO properties are identified in 
the MIBOR MLS data as HUD-owned or bank-owned. MIBOR estimates that approximately 80 
percent of single-family home listings are recorded in the MLS.4 
 
3.4 Modeling Strategy 
Our overarching hypothesis is that higher levels of nearby foreclosures lead to higher 
levels of crime. Based on that assumption, the annual number of crimes that occur in each grid 
cell is our dependent variable. A consequence of using count data as a dependent variable 
(especially at smaller units of analysis) is that those data are often skewed toward zero. Such a 
distribution violates the normality assumption necessary for linear regression. Additionally, there 
are multiple years during the study period in which zero crimes are recorded. Those observations 
reported rates of rape changed during the period under consideration.  Generally, we believe that inclusion or 
exclusion of rape counts in the violent and total crime indices is appropriate and unlikely to affect reported results. 
To maintain consistency, we include all data reflected in each UCR crime category. 
4 The percentage of REO properties listed in the MLS is likely higher than  non-REO properties. 
                                                 
lead to over-dispersion in a Poisson model. Due to our research question and structure of our 
data, we chose negative binomial fixed effects regression (see Greene, 2000; Osgood, 2000, 
Osgood & Chambers, 2000).5 
The purpose of the fixed effects estimator is to mitigate problems associated with omitted 
variable bias. The fixed effects model allows us to hold time-invariant neighborhood 
characteristics constant. By holding time-invariant characteristics constant, the potential bias 
associated with heterogeneous unobserved determinants across observations that may be 
correlated with both crime and foreclosures is reduced. That is, we have removed the influence 
of many unobserved factors that may simultaneously affect crime and foreclosures.  As a result, 
we may make stronger inferences about the relationship between foreclosures and crime. 
Based on a hypothesis consistent with Tobler’s (1970) First Law of Geography, we test 
the potential distance decay of the effect of foreclosures on crime. To do this, we examine the 
number of crimes within a 500 x 500 square feet grid cell (i.e., unit of analysis) and number of 
foreclosures within four mutually exclusive distance measures. The distance measures include 
the number of foreclosures within the grid cell; between the grid cell and 1,250 feet; from1,250 
to 2,250 feet; and from 2,250 feet to 3,250 feet. All radii are calculated from the centroid of 500 x 
500 feet grid cell.  
The use of uniform grid cells in this analysis is equivalent to including an area-based 
exposure variable. The small size of the grid cells makes explicitly including census-based 
residential population estimates or a similar measure as an exposure variable problematic. Rather 
than standardizing the risk of crime by residential population, our standardization of crime risk is 
5The xtnbreg command in Stata, based on procedures developed by Hausmann, Hall, and Griliches (1984), was used 
to predict crimes in this study (StataCorp, 2009). 
 
                                                 
based on physical area.6 That standardization is inherent in our units of analysis that do not vary 
in size or shape. 
 
4. Results 
4.1 Descriptive analysis 
 
 Table 1 shows the means, standard deviations, and ranges of variables used in our 
analyses. Standardized foreclosure measures are also shown to compare the variation of 
foreclosures by area (i.e., density of foreclosures). On average, there were approximately four 
crimes per grid cell during the study period. The number of crimes per grid cell range from zero 
to 260. Foreclosure listings per cell range from zero to 18, with a mean of 0.31. The mean 
number of foreclosures within various radii increases as distance increases. However, that 
increase is partly a function of the area within those ranges. Foreclosure density (converted to 
number of foreclosures/square mile) is fairly uniform across the various radii. However, standard 
deviations of foreclosure densities show variation increases as radius increases. 
 
Table 1: Descriptive Statistics 
 
6 Exposure variables are often included in count models of crime to address potential variation in underlying risk of 
victimization. In models predicting counts of crimes for larger units of analysis (e.g., census tracts or counties), one 
may expect more crime where more people live and spend much of their time. Accordingly, residential population 
serves as a proxy for relative victimization risk in larger units of analysis. Residential population is not a suitable 
proxy for risk in our small area units of analysis, however. There are many grid cells in our analysis that include 
exclusively nonresidential areas where people may spend time, but do not live. Crime occurs in such areas even 
though the (residential) population is zero. In addition, there are many mixed use areas where few people live 
relative to the number of people that may come and go from these areas at any given point. Therefore, residential 
population or population-based measures such as the number of residences are problematic as exposure variables in 
this context. The use of the grid cell approach here accounts for risk of victimization by equalizing the physical area 
of exposure. 
 
                                                 
Before examining the spatial extent of the effect of foreclosures on crime, it is important 
to address the directionality of the foreclosure-crime relationship. That is, we must address 
whether crimes lead to foreclosures or foreclosures lead to crimes.Although criminological 
theory suggests that increased levels of foreclosures lead to increased levels of crime (e.g., social 
disorganization and routine activities), it is not unreasonable to hypothesize that increased levels 
of crime may lead to increased levels of foreclosures. For instance, residents may exit a high 
crime area, choosing to stop payment and default on their mortgages. Another possibility is that 
mortgage companies execute the foreclosure process more rapidly in high crime areas due to 
perceived increased property risks. More expedient foreclosures in targeted high crime areas may 
reverse the direction of the relationship.  
Table 2 shows the results testing for causal order. First, the lag of crimes(t-1year) is 
examined as a predictor of foreclosure listings within each 500 x 500 feet grid cell. Next, the 
temporal lag of foreclosures(t-1year) is examined as a predictor of crime by cell. Both models are 
negative binomial fixed-effects with year binary variables. A comparison of the two models 
supports the hypothesis that higher levels of foreclosures in the immediate area lead to higher 
levels of crime, as opposed to crimes leading to higher levels of foreclosures. Specifically, the 
coefficient for lagged foreclosures is a significant predictor of crime and lagged crime is not a 
significant predictor of current foreclosures.7 These results are consistent with the findings of a 
longitudinal study examining the foreclosure-crime relationship by Katz, Wallace, and Hedberg 
(2013), which also showed that temporally-lagged foreclosures were predictors of crime.8 Given 
7 Our methodology for determining directionality is consistent with Hipp, Tita, and Greenbaum (2009).  
8 We also tested current foreclosures listings, lagged foreclosures sales, and current foreclosure sales in place of 
lagged foreclosures listings in preliminary models. Listings were consistently better predictors than foreclosure 
sales. Lagged foreclosure listings were better predictors of crime than current foreclosure listings. In fact, when 
current and lagged foreclosure listings were included in a preliminary model, lagged foreclosures were significant 
and current foreclosures were not significant.   
                                                 
the structure of our data (i.e., annual counts), the temporally lagged foreclosure variable (t-1year) 
also ensures that the foreclosures were listed prior to crimes occurring in a cell. 
 
Table 2: Direction of Relationship between Crimes and Foreclosures 
 
Multicollinearity is another potential issue when examining the relationship of spatial 
externalities within multiple, small radii of an outcome. We intentionally chose several “short 
distance” radii to examine the spatial extent of foreclosures on crime.  As a result, it should be 
noted that there is an inherent expectation of high collinearity between radii. Table 3 shows the 
correlation matrix of the number of foreclosures at each radius chosen for this study. As 
expected, there are relatively high correlations between many of the radii, suggesting that 
foreclosures are highly concentrated in some areas. These higher correlations suggest that we 
should consider the significance of independent and joint effects at each foreclosure distance 
measure.  
 
 Table 3: Correlation Matrix for number of foreclosures by Radius 
 
 
 
Table 4 shows models for total crimes, property crimes, and violent crimes. The 
separation by types of crimes shows a similar relationship to the findings associated with 
foreclosures and total crimes. In all models, the distance measures are jointly significant (χ2at 
p<0.05). The magnitude of the relationship is the greatest within the 500 x 500 feet grid cell 
across models. The distance of foreclosures to property crimes is independently significant 
(p<0.05) to at least 1,250 feet (marginally significant anywhere between 1,250 feet and 2,250 
feet). Independent estimates for distance effects of foreclosures on violent crimes (approximately 
37% of total crimes during study period) were much less precise. Although jointly significant to 
2,250-3,250 feet, the magnitude and precision of estimates drops sharply past 1,250 feet. The 
weakness of estimates associated with violent crimes may partly be a function of the relative 
rarity of such crimes. Nevertheless, the empirical pattern of a distance decay effect is apparent 
across all models. 
 
Table 4: Total Crimes, Property Crimes, Violent Crimes and Foreclosures at Multiple 
Radii 
 
Figure 1 illustrates the distance decay of the relationship between foreclosures and 
violent and property crimes. Specifically, the percentage change in the incident rate of crimes for 
a one unit change in foreclosures at each radius, holding all else constant, is shown. One 
foreclosure within the 500 x 500 feet grid cell increases the incident rate of violent crimes and 
property crimes by 1.15 percent and 1.11 percent, respectively.  That rate drops to about a 0.05 
percentage increase in violent crimes and a 0.04 percentage increase in property crimes for each 
additional foreclosure at 2,250 to 3,250 feet. 
 
5. Discussion 
 A large body of literature examines the extent to which foreclosures are a negative 
neighborhood externality.  Several studies have examined the spatial extent of foreclosures from 
an implicit value perspective (i.e., hedonic framework). Recently, scholars have focused more 
attention on questions related to other potential neighborhood outcomes associated with 
foreclosures. We join the recent interest in empirically examining the relationship between 
foreclosures and crime. 
 In the current study, we extend the existing literature that takes advantage of longitudinal 
micro-level foreclosure and crime data (Stucky, Ottensmann, and Payton, 2012; Katz, Wallace, 
and Hedberg, 2013). Using 500 x 500 feet grid cells, we explicitly measure uniform distance 
effects between foreclosures and crimes that may be masked by commonly used, larger, 
“neighborhood level” units of analysis that vary in shape and size. We estimate the relationship 
between the number of foreclosures listed during the entire previous year at various radii and 
crime. The magnitude of our results may be considered conservative as our estimates assume 
linear temporal effects over the year.9 
Our study period extends from 2003 to 2008 in a Midwestern city that experienced 
substantial increases in foreclosures prior to the 2007 housing crisis. Foreclosure listings actually 
peaked in 2007 in Indianapolis. The fact that our findings are generally consistent with prior 
small area studies of foreclosures and crime, suggests that the time period did not affect the 
results. It is our contention that the distance effect of foreclosures on crime is expected to remain 
reasonably the same before, during, or after the housing crisis. Additionally, our fixed effects 
modeling strategy absorbs many of the time-invariant unobserved determinants that may have 
been correlated with both the number of foreclosures and crime. However, the extent to which 
the “housing crisis” may have affected the distance decay effects of foreclosures on crime that 
occurred during the study period is not clear and a question that can only be answered by 
examination of other places and during different study periods.  
Generally, we find that the magnitude of the relationship between foreclosures listed in 
the previous year and total crime within the immediate area (1.15 percent increase in incident 
rate for each one unit change in foreclosure) is consistent with findings of previous studies that 
9Katz et.al. (2013) find a relatively shorter-term effect of each additional foreclosure starts (less than three months). 
                                                 
examine the same relationship at relatively small units of analysis. The magnitude of the 
relationship is considerably higher in the immediate area (500 x 500 feet grid cell). However, our 
analysis shows that foreclosures exert a negative externality through increased crime that extends 
beyond the immediate area. Though the general distance effect may vary from one urban area to 
another, our findings suggest that detectable effects of foreclosures on crime may extend beyond 
a few city blocks (e.g., 2,250 feet (.43 miles) to 3,250 feet (0.62 miles).  
 The recent literature examining the relationship between foreclosures and crime reveals 
mixed results when total crimes are separated into types of crimes (e.g. violent and property). 
Some studies find no relationship between foreclosures and property crimes, but a significant 
positive relationship between foreclosures and violent crimes (Immergluck and Smith 2006b). 
Other studies find a significant positive relationship between foreclosures and property crimes 
but not violent crimes (Williams, Galster, and Verma 2013). Our study finds that foreclosures 
have positive significant relationships with both property and violent crimes (consistent with 
Stucky, Ottensmann, and Payton 2012). As with total crime, separate models indicate that the 
effect of both types of crime declines substantially for each unit change in foreclosures as 
distance increases. The relationship between foreclosures and violent crimes is much less 
precisely estimated. In other words, the empirical magnitude of the effects of foreclosures on 
violent and property crimes declines at a similar rate as distance increases, but the statistical 
significance of foreclosures on violent crime is lower due to larger standard errors. 
In this study we focused on nonlinear foreclosure effects that declined with distance. 
Another useful line of inquiry would be to consider nonlinear, cumulative effects of foreclosures 
on crime within an area.  It seems reasonable to assume that a single foreclosure in an area may 
not have nearly the impact that multiple foreclosures in a given block might have.  Cumulative, 
nonlinear effects of foreclosures would also make sense from social disorganization, broken 
windows, and routine activities theory, which focus on reduced informal social control and 
surveillance, and increased opportunities for crime associated with foreclosures.  A few recent 
studies of foreclosures suggest that crimes do not increase linearly with foreclosures in a given 
area.10 For instance, Ellen et.al (2013) find that the effect of foreclosures on crime increases 
demonstrably as the number of foreclosures on a street segment increases to three or more. 
Additionally, Arnio, Baumer and Wolf (2012) found that an additional foreclosure in a high 
foreclosure county had a greater effect on crime than an additional foreclosure in a low 
foreclosure county.  
Though direct examination of nonlinear effects was beyond the scope of the current 
study, estimates from the current study suggest that concentrations of foreclosures have 
significant impacts on rates of crime. For example, 2.9 percent of the observations in our study 
experienced three or more foreclosure listings in the immediate area (500 x 500 feet square area) 
during the previous year. The predicted number of total crimes for those observations is 3.5 
percent higher (100 * [1.01143-1]) than immediate areas with zero foreclosures. The predicted 
number of crimes is 5.5 percent higher in the immediate area (500 x 500 feet grid cells) when the 
area between 1,250 feet and the grid cell experienced 20 or more foreclosures (4.1 percent of the 
sample). In that example, three foreclosures in the immediate cell and an additional 20 
foreclosures within 1,250 feet of the cell are predicted to cumulatively increase the incidence of 
crime by 9.0 percent compared to similar cells with zero foreclosures within 1,250 feet. Thus, 
our results suggest that high concentrations of foreclosures are expected to have a demonstrable 
10 We are grateful to an anonymous reviewer for suggesting additional attention to this issue. 
                                                 
impact on crimes locally and we believe that additional exploration of nonlinear, concentration 
effects of foreclosures would be a fruitful avenue for future studies. 
Our findings are also consistent with recent research on the micro-ecology of place as it 
relates to crime (e.g. Weisburd et al 2012). Crimes tend to be higher in areas near foreclosures. 
Perhaps not surprisingly, the immediate area (e.g., within a 500 x 500feet area) is likely to 
experience the greatest impact on levels of crime. However, a detectable impact of foreclosures 
on crime extends as far as 3,250 feet. One limitation of our study is that we could not directly 
test the mechanisms by which social disorganization, broken windows, and routine activities 
theory connect foreclosures and crime (i.e. direct measures of surveillance, disorder, informal 
social control). Future studies could explore these relationships with more direct measures. 
We also used an artificial unit of analysis—500X500 feet grid cells, which facilitated the 
exploration of distance effects.  Ideally one would like to have a unit of analysis that matches the 
social reality of the world people inhabit.  This has been an ongoing struggle in macro-level 
research (see Hipp and Boessen 2013 for a discussion), especially as it relates to what constitutes 
a neighborhood.  The traditional approach in social disorganization theory is that a neighborhood 
is a discrete area that is somewhere along a continuum of organization or disorganization.  Yet, 
Hipp and Boessen (2013) note that residents tend to view themselves as the center of their 
neighborhood (in what they call “egohoods”) and they argue that neighborhoods are not discrete 
entities but rather overlapping waves that ripple throughout a city.  Hipp and Boessen (2013) also 
note that research shows there are distance decay functions in social ties among area residents 
and travel to crime patterns for criminals.  Our study found distance decay effects of one factor 
likely to be associated with physical and social disorder (i.e., foreclosures). The incorporation of 
the concept of egohoods into social disorganization and other macro-level theories of crime 
seems a fruitful avenue for future theorizing.  Our study provides empirical evidence that there 
are good reasons to investigate this further and perhaps begin to explore ways to explicitly 
incorporate physical distance within macro-level theories of crime and redefine what constitutes 
one’s “neighborhood”.  Though beyond the scope of this study, explicit consideration of how 
people conceptualize the extent of their neighborhood and awareness space would benefit extant 
macro-level theories of crime and provide direct guidance on how far one would expect to see 
effects of physical and social disorder extend. 
Practically, our results suggest yet another justification for recent efforts, such as the 
federal Neighborhood Stabilization Program, to revitalize neighborhoods that experience high 
levels of foreclosure activity. In addition, when stakeholders consider proposals for public 
interventions to address community impacts of foreclosures, they should consider that properties 
located outside areas with high concentrations of foreclosurescan experience negative external 
effects. Specifically, our findings indicate that properties on the edge of high foreclosure 
concentrations could experience increased crime. This suggests that police, community 
organizations, and residents should monitor not only the areas disproportionately affected by 
foreclosures, but also neighborhoods near greater concentrations of foreclosures.  
 
 
 
 
 
 
 
References 
 
Acevedo, Luis G. 2009. “An Empirical Analysis of the Impact of Home Foreclosure on the 
Crime Rate: Evidence in Atlanta, GA.” Research paper, Economics. Smithfield, RI: Bryant 
University. 
 
Arnio, Ashley, and Eric P. Baumer. 2012. “Demography, Foreclosure, and Crime: Assessing 
Spatial Heterogeneity in Contemporary Models of Neighborhood Crime Rates.” Demographic 
Research 26:449-486. 
 
Arnio, Ashley N., Eric P. Baumer, Kevin T. Wolff. 2012. “The Contemporary Foreclosure Crisis 
and US Crime Rates. Social Science Research. 41(6): 1598-1614. 
 
Baumer, Eric P., Kevin T. Wolff, and Ashley N. Arnio. 2012. “A Multicity Neighborhood 
Analysis of Foreclosure and Crime.  Social Science Quarterly. 93(3): 559-865. 
 
Bess, Michael. 2008. “Assessing the Impact of Home Foreclosures in Charlotte Neighborhoods.”  
Geography and Public Safety 1(3): 2-5.  
 
Brantingham, Paul J., and Patricia L. Brantingham.1984.Patterns in Crime. New York: 
Macmillan. 
 
Brantingham, Patricia L., andBrantingham, Paul. J. 1993. Nodes, Paths and Edges: 
Considerations on the Complexity of Crime and the Physical Environment. Journal of 
Environmental Psychology 13: 3-28. 
 
Cohen, Lawrence and Marcus Felson. 1979. Social Change and Crime Rate Trends: A Routine 
Activity Approach. American Sociological Review. 44(4): 588-608. 
 
Cui, Lin. 2011. “Foreclosure, Vacancy, and Crime.”Unpublishedmanuscript.Available at 
(http://www.pitt.edu/~lic29/Foreclosure, Vacancy and Crime.pdf). 
 
Ellen, Ingrid Gould, Johanna Lacoe, and Claudia Ayanna Sharygin.2013.“Do Foreclosures 
Cause Crime?”Journal of Urban Economics. 74: 59-70 
 
Goodstein, Ryan M., and Yan Y. Lee. 2010. Do Foreclosures Increase Crime?Federal Deposit 
Insurance Corporation Center for Financial Research Working Paper, No. 2010-05.Available at 
(http://www.fdic.gov/bank/analytical/cfr/2010/wp2010/CFR_WP_2010_05goodsteinlee.pdf). 
 
Greene, William H. 2000. Econometric Analysis, 4th edition. Upper Saddle River, NJ: Prentice 
Hall. 
 
Harding, John P., Eric Rosenblatt, and Vincent W. Yao. 2009. “The Contagion Effect of 
Foreclosed Properties.” Journal of Urban Economics 66(3): 164-178. 
 
Harris, Paul E. (Lish). 2011. “The Criminal Consequences of Changes in Neighborhood 
Structure Due to Home Foreclosure: A Theoretical Discussion.” Chapter in In Economic Crisis 
and Crime, edited by Mathieu Deflem. Bingley, UK: Emerald. 
 
Hausman, Jerry, Bronwyn H. Hall, and ZviGriliches. 1984. “Econometric Models for Count Data 
with an Application to the Patents-R&D Relationship.” Econometrica 52(4): 908-938. 
 
Hipp, John. R., and AdamBoessen. 2013. “Egohoods as Waves Washing across the City: a New 
Measure of “Neighborhoods”. Criminology51(2): 287-327. 
 
Hipp, John R., George E. Tita, and Robert T. Greenbaum. 2009. “Drive-bys and Trade-ups: 
Examining the Directionality of the Crime and Residential Instability Relationship.” Social 
Forces 87(4): 1777-1812. 
 
Immergluck, Dan and GeoffSmith. 2006a. “The External Costs of Foreclosure: The Impact of 
Single-Family Mortgage Foreclosures on Property Values.” Housing Policy Debate 17(1): 57-79. 
 
Immergluck, Dan and Geoff Smith.2006b.“The Impact of Single Family Mortgage Foreclosures 
on Neighborhood Crime.”Housing Studies, 21(6): 851-866. 
 
Immergluck, Dan. 2010. Neighborhoods in the Wake of the Debacle, Intrametropolitan Patterns 
of Foreclosed Properties, Urban Affairs Review 46(1): 3-36. 
 
Jones, Roderick W., and William Alex Pridemore. 2012. The Foreclosure Crisis and Crime: Is 
Housing-Mortgage Stress Associated with Violent and Property Crime in U.S. Metropolitan 
Areas? Social Science Quarterly. 93(3): 671-691. 
 
Katz, Charles M., Danielle Wallace, and E.C. Hedberg. 2013. “A Longitudinal Assessment of 
the Impact of Foreclosure on Neighborhood Crime.” Journal of Research in Crime and 
Delinquency 50:359-398.  
 
Kirk, David, and Derek Hyra. 2012. “Home Foreclosures and Community Crime: Causal or 
Spurious Association?” Social Science Quarterly. 93(3): 648-670. 
 
Leonard, Tammy and James C. Murdoch. 2009. “The Neighborhood Effects of Foreclosures.” 
Journal of Geographic Systems 11(4): 317-332. 
 
Lin, Zhenguo, Eric Rossenblatt, and Vincent W. Yao. 2009. “Spillover Effects of Foreclosures 
on Neighborhood Property Values.” Journal of Real Estate Finance and Economics 38(4): 387-
407. 
 
Osgood, D. Wayne. 2000. “Poisson-based Regression Analysis of Aggregate Crime Rates.” 
Journal of Quantitative Criminology 16(1): 21-43. 
 
Osgood, D. Wayne, and Jeff M. Chambers. 2000. “Social Disorganization Outside the 
Metropolis: An Analysis of Rural Youth Violence. Criminology 38(1): 81-115. 
 
Pandit, Caaminee. 2011. “The Impact of Foreclosures on Crime in American Cities.” 
Sociological Insight 3: 82-109.   
 
Sampson, Robert J., and W. Byron Groves. 1989. “Community Structure and Crime: Testing 
Social Disorganization Theory.” American Journal of Sociology 94(4): 774-802. 
 
Schuetz, Jenny, Vicki Been, and Ingrid Gould Ellen. 2008. “Neighborhood Effects of 
Concentrated Mortgage Foreclosures.” Journal of Housing Economics 17(4): 306-319.  
 
Shaw, Clifford R., and Henry D. McKay [1942] 1972. Juvenile Delinquency and Urban Areas, 
rev. ed. Chicago, IL: University of Chicago Press. 
 
Sherman, Lawrence W., Patrick R. Gartin, and Michael E. Buerger. 1989. "Hot Spots of 
Predatory Crime: Routine Activities and the Criminology of Place*." Criminology 27: 27-56. 
 
Spelman, William. 1993. “Abandoned Buildings: Magnets for Crime?” Journal of Criminal 
Justice 21(5): 481-495. 
 
StataCorp. 2009.Stata Longitudinal-Data/Panel-Data Reference Manual: Release 11. College 
Station TX: Stata Press. 
 
Stucky, Thomas D., John R. Ottensmann, and Seth B. Payton. 2012. The Effect of Foreclosures 
on Crime in Indianapolis, 2003-2008. Social Science Quarterly. 93(3): 602-624 
 
Taylor, Ralph. B. 1997. Social Order and Disorder of Street Blocks and Neighborhoods: 
Ecology, Microecology, and the Systemic Model of Social Disorganization. Journal of Research 
in Crime and Delinquency34: 113-155. 
 
Tobler, Waldo R. 1970.  A Computer Movie Simulating Urban Growth in the Detroit Region. 
Economic Geography, 46, 234-40. 
 
Wallace, Danielle, E. C. Hedberg, and Charles M. Katz. 2012. The Impact of Foreclosures on 
Neighborhood Disorder and During the Housing Crisis: Testing the Spiral of Decay. Social 
Science Quarterly. 93(3): 625-647. 
 
Weisburd, David L., Elizabeth R. Groff, and Sue-Ming Yang. 2012.TheCriminology of Place: 
Street Segments and our Understanding of the Crime Problem. Oxford University Press. 
 
Williams, Sonya, George Galster, and NanditaVerma. 2013. “Home Foreclosures and 
Neighborhood Crime Dynamics.” Housing Studies.Published online 18 June 2013. 
DOI:10.1080/02673037.2013.803041 
 
Wilson, James Q., and George L. Kelling. 1982. “The Police and Neighborhood Safety: Broken 
Windows.”  Atlantic Monthly 127: 29-38. 
 
Wilson, Ronald E., and Derek J. Paulsen.2008. "Foreclosures and crime: A geographical 
perspective." Geography & Public Safety 1.3: 1-2. 
 
Wolff, Kevin T., Joshua C. Cochran, and Eric P. Baumer. 2013. “Reevaluating Foreclosure 
Effects on Crime during the “Great Recession”.” Journal of Contemporary Criminal Justice 
Published online 15 November 2013. DOI: 10.1177/1043986213509025.
Tables 
 
Table 1: Descriptive Statistics 
Variable Mean Std. Dev. Min Max 
Analyzed Variables 
Total Crime 4.11 8.27 0 260 
Violent Crime 1.5 3.75 0 184 
Property Crime 2.61 5.67 0 240 
Number foreclosures Cell 0.31 0.79 0 18 
Number of Foreclosures between Cell& 1250 5.69 6.87 0 51 
Number of Foreclosures between 1250 & 2250 13.37 12.80 0 94 
Number of Foreclosures between 2250 & 3250 20.80 17.16 0 134 
Standardized Foreclosure Variables (foreclosures/mile) 
Foreclosure Density Cell 34.2 87.71 0 2007.25 
Foreclosures Density < 1250 34.07 41.14 0 323.72 
Foreclosures Density < 2250 33.95 32.87 0 220.86 
Foreclosures Density < 3250 33.75 28.38 0 188.19 
 
 
   
Table 2: Direction of Relationship between Crimes and Foreclosures 
 Dependent Variables 
VARIABLES Foreclosure Crime 
   
Total Crime(t-1) -0.0025 -- 
 (0.0027)  
Total Foreclosure Listings (t-1) -- 0.0139*** 
  (0.0043) 
Year 2005 0.1687*** 0.0392*** 
 (0.0297) (0.0100) 
Year 2006 0.3147*** 0.0802*** 
 (0.0288) (0.0099) 
Year 2007 0.4782*** 0.0966*** 
 (0.0280) (0.0099) 
Year 2008 0.4071*** 0.0313*** 
 (0.0284) (0.0101) 
Constant 2.2404*** 2.4008*** 
 (0.1697) (0.0297) 
   
Observations 18,220 34,920 
Number of zone 3,644 6,984 
Wald χ2 370.92 143.20 
Log Likelihood -13,320.30 -51,775.84 
Standard errors in parentheses 
*** p<0.01, ** p<0.05, * p<0.10 
 
  
 
 
 
 
1 
 
Table 3: Correlation of Number of Foreclosures at Each Radius 
  Between 
 Cell (500x500ft) Cell &1,250ft 1,250ft &2,250ft 
Between Cell& 1,250ft 0.43   
Between 1,250ft &2,250ft 0.34 0.73  
Between 2,250ft &3,250ft 0.28 0.60 0.79 
 
 
 
 
Table 4: Violent Crimes, Property Crimes, and Number of Foreclosures at Multiple Radii 
VARIABLES Total Crimes Violent Crimes Property Crimes 
Cell (500x500ft) 0.0114*** 0.0114* 0.0110** 
 (0.0013) (0.0064) (0.0052) 
Between Cell & 1,250ft 0.0027*** 0.0022 0.0023** 
 (0.0010) (0.0014) (0.0012) 
Between 1,250ft &2,250ft 0.0017*** 0.0010 0.0018** 
 (0.0001) (0.0010) (0.0008) 
Between 2,250ft &3,250ft 0.0004 0.0005 0.0004 
 (0.0005) (0.0008) (0.0006) 
Year 2005  0.0251*** 0.0652*** 0.0120 
 (0.0105) (0.0170) (0.0124) 
Year 2006 0.0585*** 0.0393** 0.0797*** 
 (0.0112) (0.0177) (0.0131) 
Year 2007 0.0668*** 0.0801*** 0.0668*** 
 0.0122) (0.0194) (0.0142) 
Year 2008 -0.0113 0.0054 -0.0161 
 (0.0144) (0.0230) (0.0166) 
Constant 2.3538*** 1.901*** 2.1913*** 
 (0.0317) (0.0436) 0.0376 
    
Observations 34,920 29,320 34, 250 
Number of zone 6,984 5,864 6,850 
Wald χ2 166.10*** 67.08*** 152.44*** 
χ2 joint significance distance 33.38*** 10.16*** 23.03*** 
Standard errors in parentheses 
*** p<0.01, ** p<0.05, * p<0.10 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figures 
 
Figure 1: Percentage Change in Violent and Property Crimes for One Unit Change in Foreclosures 
 
 
 
*Jointly significant at p<0.05 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0
0.2
0.4
0.6
0.8
1
1.2
In Cell (500 x 500 feet) between cell & 1,250 ft between 1,250 ft & 2,250 ft between 2.250 ft & 3,250 ft
P
er
ce
n
ta
ge
 in
cr
es
e 
in
 C
ri
m
e
Violent Crimes
Property Crimes
Maps 
Map 1: Units of Analysis and Study Area 
 
Streets
500-foot Grid Cells
IPD Service Area
0 1 2 3 4 5 6 Miles
N