The relation between residential property and its surroundings and day- and night-time residential burglary

TL;DR: In this article, the authors examined how residential property and its surroundings influence day-and night-time residential burglary and concluded that two separate burglary prevention frameworks are needed: one for day and another one for night time burglary.

Abstract: This article examines how residential property and its surroundings influence day- and night-time residential burglary. Crime Prevention Through Environmental Design (CPTED) principles of territoriality, surveillance, access control, target hardening, image maintenance, and activity support underpin the study. Data were collected by observing 851 houses in the city of Enschede, half of which were burgled and half representing a random selection of houses not burgled. Multilevel multinomial regression models were estimated for predicting day- and night-time burglaries. The findings show that territoriality and access control predict daytime burglary while access control and target hardening predict night-time burglary. The analysis controls for offender availability, target attractiveness, and residential stability. The conclusion is that two separate burglary prevention frameworks are needed: one for day-and another one for night-time burglary.

Summary

Introduction

• Rational choice theory assumes that offenders pursue normal goals like non-offenders do:.
• The advantage of focusing on the built environment is that many characteristics are modifiable.

Territoriality

• CPTED studies reported that houses with front gardens and with symbolic barriers would have a lower risk.
• There might be several explanations of why this is important for deterring daytime burglary.
• The authors findings thereby support previous research about the importance of territoriality, but qualify it as well: Territoriality is important but only for daytime burglary.
• It is possible that having a front garden reflects surveillance because people sit or work there and provide informal surveillance.
• In support of this explanation, Kuo and Sullivan (2001) found that vegetation around buildings is associated with lower levels of crime.

Surveillance

• Three types of strategies were investigated: informal or natural, formal or organized, and mechanical.
• It was tested whether streets with higher speed limits were associated with higher risk because drivers had reduced surveillance opportunities.
• This illustrates that although conceptually, the CPTED concepts are clear, it is challenging to operationalize them into specific variables, and thus to identify concrete potential preventive measures.
• There was a tendency for the evidence of a dog to be associated with decreased night-time burglary.

Access Control

• Two access control aspects were investigated: property and road structure.
• The bicycle’s higher mobility compared with that of a car could explain why the road layout is less important in a country such as the Netherlands.

Image/Maintenance

• The relation between the physical condition and “image” of the built environment and crime and the fear of crime has long been acknowledged (Lynch, 1960).
• There are conflicting views as Wright and Decker (1994) consider that when selecting targets, burglars relate the size and condition of a house and the type of cars in the driveway to the type and value of the house’s contents.
• On the one hand, it seems possible that maintenance matters less by night due to low visibility, while on the other hand, the findings of Coupe and Blake (2006) suggest that the effect of maintenance might be confounded with the differences in target selection strategies of day- versus night-time burglars.
• This therefore suggests that it is necessary to control for neighborhood income.

Activity Support

• The authors investigated whether (a) there is lower burglary risk for houses close to buildings used for recreational, educational, commercial, or religious purposes; (b) medium intensity of vehicle and bicycle traffic at the nearest junction is associated with higher risk; and (c) medium intensity of pedestrian and vehicle traffic along the street segment is associated with higher risk.
• The authors findings contrast with those of many (Jacobs, 1961; Petterson, 1997; Poyner, 1983; Zelinka & Brennan, 2001) as no activity support variables were found to be statistically significant in the regression model.
• There is a tendency for proximity to commercial establishments to be associated with higher daytime burglary risk, which is in contrast to the expected beneficial surveillance effects.
• Therefore, “activity support” leads to increased risk rather than decreased risk, suggesting that the opportunity effect (i.e., bringing offenders to possible targets) is more important than the possible positive effects of increased surveillance.
• A possible explanation of the difference with previous research findings is that their study looked at proximity to specific land uses, rather than to land-use density.

Sample

• The study was carried out in Enschede, a city along the Dutch-German border.
• A case-control study involves the identification of “individuals” with (“cases”) and without (“controls”) a particular condition.
• The controls were selected randomly by the municipality of Enschede, out of a list containing all properties in the city.
• Four houses were not observed as they did not meet this study’s definition of residential property (refer to section “Concepts”).
• It was deemed unsuitable to select a neighboring house or a house from the same street as a control case due to “near repeat” victimization since research (Bernasco, 2008; Bowers & Johnson, 2004; Townsley, Homel, & Chaseling, 2003) shows that houses located near a burgled house have a higher victimization risk.

Data Collection

• The observation form that was used to collect information was based on a design by the University of Huddersfield (Armitage, 2006) and it was adapted to the Dutch context during a pilot phase.
• The most important adaptation related to the inclusion of proximity to non-residential activities and cyclingrelated characteristics.
• Differences identified in the coding were discussed and an appropriate guideline was developed to standardize coding.
• The data collection was carried out by observing the houses from the sidewalk or footpath.
• The observers were unaware whether the houses had been burgled or not.

Concepts

• The Dutch police definition of burglary was used: “the theft from a house and/or attached storeroom, garage, shed, etc.” Burglary was measured using a four category variable: (a) not burgled, (b) burgled during the day, (c) burgled during the night, (d) burgled during both day and night (i.e., dual day/ night).
• Due to space limitations, the descriptive statistics of the dichotomous independent variables were simplified by describing them as if these were continuous variables (see Table 2).
• Residential stability was measured using the percentage of home owners and the length of stay in the neighborhood of the household head expressed in Territoriality 0.877 −.
• Offender availability was measured using the percentage of suspected residential burglars in a neighborhood, as recorded by the police.

Analyses

• Cross-tabular statistics and Pearson chi-square values tested the relationship between a given dichotomous independent and the dichotomous dependent variable.
• For continuous variables, an independent-sample t test was used.
• Multilevel regression was used since the average correlation between the variables measured on houses from the same neighborhood is higher than that of houses from other neighborhoods (see Table 3).
• The variance inflation factor (i.e. VIF), a diagnostic measure of multicollinearity, indicated that two variables have values above 4 (i.e., home ownership = 4.31; social cohesion = 4.13).
• Finally, on the basis of the individual CPTED regressions, a multilevel multinomial logistic regression model of day- and night-time burglary was elaborated using only those variables with p values below .10 on either the day- or the night-time parts of the model.

Results

• Table 2 summarizes the results of the descriptive statistics of the 42 dichotomous and 5 continuous control variables.
• It also presents the bivariate analyses between daytime, night-time burglary, and dual day/night-time and each of the characteristics.
• That is, it is not only the opposite of what was expected, but it is not likely that this result happened by chance.
• For comparability, the model contained the same 17 variables in the dayand night-time parts of the model (2 of which were control variables).

Daytime Burglary

• The model shows that daytime burglary is related to territoriality and access control.
• Specifically, having a front garden associated with lower daytime burglary, while holding all other variables constant (RRR1 = 0.46, for example, houses that have a front garden have 0.46 times lower risk of being burgled than those that do not have such feature).
• Being an un-detached house is also associated with lower daytime burglary risk (RRR = 0.55).
• Neighborhood stability reduces the likelihood of daytime burglaries (RRR = 0.89).
• Proximity to commercial establishments (i.e., activity support) and window screening (i.e., target hardening) were on the verge of statistical significance (RRR = 1.45 and 0.49, respectively).

Night-Time Burglary

• The model shows that night-time burglary is related to access control and target hardening.
• The results show that being an un-detached house is associated with lower night-time burglary, while holding all other variables constant (RRR = 0.68).
• Window screening is also associated with lower night-time burglary risk (RRR = 0.44).
• Three variables were on the verge of statistical significance.
• Finally, offender availability was associated with a large increase in burglary risk (RRR = 11.58).

Discussion

• The aim of this study was to investigate whether (a) burgled houses and nonburgled houses differ with respect to the CPTED concepts and whether (b) a day and night effect exists.
• To account for several plausible alternative explanations, the analysis controlled for target attractiveness, social cohesion, and offender availability.
• In addition, the multilevel analysis took the ICC (i.e., nesting within neighborhoods) into account.
• First, this study found that 34.7% of the burglaries took place during the night while 65.3% took place during the daytime.
• Many of these relationships disappeared in the multivariate analysis.

Image Maintenance

• The authors examined whether risk is lower when there is good maintenance of (a) the house, (b) neighboring houses, (c) window and door frames, and when (d) the front garden is attractive.
• In contrast to the findings of other researchers (Hedayati Marzbali et al., 2012; Pruitt et al., 2000; Ross & Mirowsky, 1999; Taylor, 1991; Wilson & Kelling, 1982), this study found that good house or neighborhood maintenance is not associated with lower burglary.
• It is possible that the relatively homogeneous level of upkeep of Dutch neighborhoods explains why image maintenance does not predict burglary.

Application and Synthesis

• There are a number of practical design recommendations that stem from these findings.
• For proper decision making, the stakeholder(s) that would be involved in the modification of a particular property or property surrounding characteristic should be first identified.
• In addition, further analyses are required before a decision can be made about the selection of measures.
• Second, the time and costs associated with physical adaptation measures should also be evaluated.
• Without being exhaustive, next some practical recommendations are presented.

Home Owners

• The visibility into the back gardens should be blocked by means of fences or tall and dense bushes.
• Window screening is often seen as an expensive home improvement to reduce crime risk and increase thermal insulation.
• This research has confirmed that particularly for those who work night shifts, it is a worthwhile option.

Local Governments/Property Developers

• It is worth noting that the operationalization of some concepts can highlight ambiguities.
• Some aspects of territoriality (e.g., “front garden”) may be better conceived as “informal surveillance.”.
• The authors conclusion is that CPTED concepts provide an important framework for studying burglary although further research is needed to fill-in some gaps.
• From the management viewpoint, a number of design recommendations were discussed.

Potential Research Design Improvements

• There are four areas where the research design could be further fine-tuned.
• Second, the existence of street lights along the sidewalk could be limited by, for example, actual illumination and vegetation.
• Third, the research design (i.e., observations) could cause underestimation in the case of target hardening as there is sometimes a deliberate effort to conceal objects.
• Similarly, measures implemented as the result of a burglary cannot be identified through this method.
• A survey of home owners could therefore complement the collected data.

Acknowledgments

• This article was possible thanks to the burglary data set provided by Krista Smid, Marga Nieborg, and Herman Meuleman (Municipality of Enschede) and Axel Zengerink (Police Department).
• The authors also wish to thank Feike Aantjes, Mark Broekhuis, Kim Brunninkhuis, Elise Spanjer, and Aniek Oude Alink for the field data collection, as well as Marcus Felson and Ken Pease for providing valuable comments on a previous version of this article.

Author Biographies

• Lorena Montoya is senior researcher at the Department of Services, Cyber-Security and Safety at the University of Twente, The Netherlands.
• Her research interests include crime science in general and in particular information security and spatiotemporal analysis of crime.
• Marianne Junger is professor in cyber-security and business at the Department of Industrial Engineering and Business Information Systems at the University of Twente, The Netherlands.
• Yfke Ongena is assistant professor at the Department of Communication and Information Sciences at the University of Groningen, The Netherlands.
• Her research interests include research methods such as questionnaire design and behavior coding.

Figures

Table 2. (continued)

Table 3. Multilevel Multinomial Regression Model (N = 788).

Table 1. (continued)

Table 1. (continued)

Table 3. Multilevel Multinomial Regression Model (N = 788).

Table 2. (continued)

Full text

University of Groningen
The relation between residential property and its surroundings and day- and night-time
residential burglary
Montoya, Lorena; Junger, Marianne; Ongena, Yfke
Published in:
Environment and Behavior
DOI:
10.1177/0013916514551047
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2016
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Montoya, L., Junger, M., & Ongena, Y. (2016). The relation between residential property and its
surroundings and day- and night-time residential burglary.
Environment and Behavior
,
09
(2014), 516-549.
https://doi.org/10.1177/0013916514551047
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license.
More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-
amendment.
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum.

Environment and Behavior
2016, Vol. 48(4) 515 –549
© 2014 SAGE Publications
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0013916514551047
eab.sagepub.com
Article
The Relation Between
Residential Property and
Its Surroundings and
Day- and Night-Time
Residential Burglary
Lorena Montoya
1
, Marianne Junger
1
, and
Yfke Ongena
2
Abstract
This article examines how residential property and its surroundings influence
day- and night-time residential burglary. Crime Prevention Through
Environmental Design (CPTED) principles of territoriality, surveillance,
access control, target hardening, image maintenance, and activity support
underpin the study. Data were collected by observing 851 houses in the
city of Enschede, half of which were burgled and half representing a random
selection of houses not burgled. Multilevel multinomial regression models
were estimated for predicting day- and night-time burglaries. The findings
show that territoriality and access control predict daytime burglary while
access control and target hardening predict night-time burglary. The analysis
controls for offender availability, target attractiveness, and residential
stability. The conclusion is that two separate burglary prevention frameworks
are needed: one for day-and another one for night-time burglary.
Keywords
Crime Prevention Through Environmental Design (CPTED), environmental
criminology, urban and neighborhood design
1
University of Twente, Enschede, The Netherlands
2
University of Groningen, Groningen, The Netherlands
Corresponding Author:
Lorena Montoya, Faculty EWI, Services, Cyber-Security and Safety Group, University of
Twente, P.O. Box 217, 7500 AE, Enschede, The Netherlands.
Email: a.l.montoya@utwente.nl
551047EABXXX10.1177/0013916514551047Environment and BehaviorMontoya et al.
research-article2014

516 Environment and Behavior 48(4)
Introduction
The rational choice model (Clarke & Cornish, 2010), the routine activities
approach (L. E. Cohen & Felson, 1979), and the awareness space concept of
the crime pattern theory (Brantingham & Brantingham, 2008) argue that
crime can be explained by environmental rather than by offender characteris-
tics (Tilley & Laycock, 2007; Wortley & Mazzerole, 2008). It is plausible
that neither of these factors is solely responsible for crime but that both fac-
tors are responsible to some degree.
Research on property crime has focused on how offenders choose targets
and carry out offenses. Rational choice theory assumes that offenders pursue
normal goals like non-offenders do: They evaluate possible actions and
choose the action that maximizes gain and minimizes costs and risks (Clarke
& Cornish, 2010). The concept of limited rationality, however, proposes that
for behavior to be rational, it does not have to be carefully preconceived and
planned or require hierarchical, sequential decision making. It is enough that
decisions are perceived to be optimal (Cromwell & Olson, 2006). According
to the routine activities approach, “rational” choice is carried out during
everyday-life routine activities such as on the way to shopping, entertain-
ment, work, and school. Because some activities take place mainly during the
day while others mainly during the night, the benefits and risks of burglary
might vary by time and location. Offenders attempt to make rational deci-
sions which could shape the urban burglary patterns differently by day or by
night. The present study focuses on the physical environment and its relation
with day- and night-time burglary. More specifically, we investigate whether
burgled houses and non-burgled houses differ with respect to the CPTED
concepts and whether day and night effects exist.
Burglary is mostly opportunistic and often takes place at an opportune
moment when occupants are clearly absent and the house is perceived as
vulnerable (Cromwell & Olson, 2006). A motivated offender must first iden-
tify a vulnerable target and then enter the property without being detected. A
burglar’s decision to “hit” a specific target is based on environmental cues
that are perceived to have immediate consequences.
There is a considerable amount of research on residential burglary
(Comeau & Klofas, 2014; Cozens, 2002; Cozens, Hillier, & Prescott, 2001;
Cozens, Saville, & Hillier, 2005). The driving factors of burglary have been
described in many studies (Bernasco & Luykx, 2003; Bernasco &
Nieuwbeerta, 2005; Comeau & Klofas, 2014; Cozens et al., 2001; Cromwell
& Olson, 2004, 2006; Hakim, Rengert, & Shachmurove, 2001; Nee &
Meenaghan, 2006; A. Piquero & Rengert, 1999; N. L. Piquero, Cohen, &
Piquero, 2010; Ratcliffe, 2003; Snook, 2004; Wright & Decker, 1994).

Montoya et al. 517
Burglars are usually driven by money because it affords them a luxury life-
style, the so-called “Life as a party” described by Wright and Decker (1994).
Besides economic motives, there are also social ones such as peer approval,
status, and idiosyncratic motives, such as revenge, kicks, thrills, pathological
behavior, and rebellion (Cromwell & Olson, 2006).
In addition to burglars’ motives, research has also investigated target
selection and the process of estimating target suitability. Four main catego-
ries have been identified:
a. Familiarity: Most offenders do not travel very far to offend since fa-
miliarity with surroundings reduces stress (Block & Bernasco, 2009;
Rengert, Piquero, & Jones, 1999).
b. Occupancy: Most burglars prefer unoccupied targets. Occupancy cues
include the presence of visible residents or indications that someone
is at home (e.g., noises, lights, vehicles). Visibility and occupancy are
somewhat passive concepts and therefore some prefer the concept of
guardianship (Reynald & Elffers, 2009). Occupancy proxies such as a
dog or a car in the driveway can deter a burglar (Weisel, 2002; Wright
& Decker, 1994).
c. Potential rewards: The main driving factor of burglary is money. Signs
of potential financial rewards in the dwelling play a role in target selec-
tion (Macintyre, 2001).
d. The built environment: Within the environment, the built environment
has a prominent place. Crime Prevention Through Environmental De-
sign (CPTED) concepts represent an operationalization of environ-
mental models of crime. The CPTED approach states that the proper
design and effective use of the built environment can lead to a reduc-
tion in the fear and incidence of crime, and an improvement in the
quality of life (Cozens et al., 2005; Gibson, 2013).
The origins of CPTED can be traced mainly to the 60s and 70s (Angel,
1968; Jacobs, 1961; Jeffery, 1971; Newman, 1972; Robinson, 2013).
Research findings on the built environment may be applicable to many types
of crime but have until now focused on burglary, although they have also
been applied to robberies (Casteel & Peek-Asa, 2000) and to injury control
(Peek-Asa & Zwerling, 2003). The CPTED concepts are widely used by gov-
ernments as guidelines for “designing out crime.” In the Netherlands, for
example, these concepts can be identified in the Police Marque Secured
Housing, a certification scheme for new and old houses that can be requested
for individual houses, apartment complexes and the neighborhood as a whole.
The certification includes issues such as locks and door standards, type of

518 Environment and Behavior 48(4)
building materials to prevent fire and deter vandalism, as well as lighting
requirements to provide good visibility. The advantage of focusing on the
built environment is that many characteristics are modifiable. Consequently,
the built environment is a prime target for policymakers who wish to prevent
burglary.
Despite the wide application of CPTED principles, the empirical base on
which the CPTED concepts are founded is limited. Below we discuss several
limitations that exist and explain the approach used in this study to address
them.
First, with some exceptions (Armitage, 2007; Brown & Altman, 1983;
Reynald, 2009, 2011), observations have not been used to systematically
measure the CPTED concepts and study their relation with residential bur-
glary. The present research uses observations to measure the CPTED
characteristics.
Second, no systematic evaluation has been conducted to differentiate
between the six concepts and further differentiated them. For example, some
aspects of CPTED relate to characteristics of the property, but some to the
immediate surroundings of the property. Aspects of the property are the
responsibility of the owner, whereas the immediate environment is most
often the responsibility of the (local) government. The comparison of the
effectiveness of property-related versus property surrounding characteristics
is therefore relevant from an urban management viewpoint. The present study
will analyze the findings based on these two levels.
Third, studies investigating the CPTED principles seldom control for a
number of important possible interfering factors, such as target attractive-
ness, offender distribution and social cohesion. An alternative hypothesis to
the CPTED approach is that motivated individuals may not be deterred easily
by environmental characteristics. Despite research evidence showing the rel-
evance of protection measures (Bernasco & Luykx, 2003; Ratcliffe, 2003;
Rossmo, 2000), it is possible that motivated burglars find their way into a
property, regardless of whether it is adequately protected or not (Compton,
Conway, Stinson, Colliver, & Grant, 2005). Contradictory findings have been
reported with respect to the relationship between the socioeconomic charac-
teristics of houses and neighborhoods and burglary. For example, household
income was linked with higher rates of burglary in the United Kingdom and
lower rates in the United States (Tseloni, Wittebrood, Farrell, & Pease, 2004).
Malczewsk and Poetz (2005) reported that both low and high socioeconomic
areas can have relatively high burglary rates. It is possible that within low-
income neighborhoods, burglars prefer the “richer looking” houses (Shaw &
Gifford, 1994). Tilley and Webb (1994) showed that the relationship between
income, burglary, and security is complex and that it varies over time (e.g.,

Frequently asked questions

Q1. What contributions have the authors mentioned in the paper "The relation between residential property and its surroundings and day- and night-time residential burglary" ?

This article examines how residential property and its surroundings influence dayand night-time residential burglary. Crime Prevention Through Environmental Design ( CPTED ) principles of territoriality, surveillance, access control, target hardening, image maintenance, and activity support underpin the study.