You are on page 1of 19

This article was downloaded by: [Facultad Latinoamericana de Ciencias Sociales]

On: 27 March 2015, At: 14:54


Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Sociological Spectrum: Mid-South


Sociological Association
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/usls20

Untangling Fear of Crime: A Multitheoretical Approach to Examining the


Causes of Crime-Specific Fear
a

Mariel Alper & Allison T. Chappell

Department of Criminology and Criminal Justice , University of


Maryland , College Park , Maryland , USA
b

Department of Sociology and Criminal Justice , Old Dominion


University , Norfolk , Virginia , USA
Published online: 19 Jun 2012.

To cite this article: Mariel Alper & Allison T. Chappell (2012) Untangling Fear of Crime: A Multitheoretical Approach to Examining the Causes of Crime-Specific Fear, Sociological Spectrum: MidSouth Sociological Association, 32:4, 346-363, DOI: 10.1080/02732173.2012.664048
To link to this article: http://dx.doi.org/10.1080/02732173.2012.664048

PLEASE SCROLL DOWN FOR ARTICLE


Taylor & Francis makes every effort to ensure the accuracy of all the information (the
Content) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions

Sociological Spectrum, 32: 346363, 2012


Copyright # Taylor & Francis Group, LLC
ISSN: 0273-2173 print=1521-0707 online
DOI: 10.1080/02732173.2012.664048

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

UNTANGLING FEAR OF CRIME:


A MULTI-THEORETICAL APPROACH TO EXAMINING
THE CAUSES OF CRIME-SPECIFIC FEAR
Mariel Alper1 and Allison T. Chappell2
1

Department of Criminology and Criminal Justice, University of Maryland,


College Park, Maryland, USA
2
Department of Sociology and Criminal Justice, Old Dominion University,
Norfolk, Virginia, USA
Fear of crime has been a major research topic over the past several decades. However,
multiple explanations have been proposed and no comprehensive theoretical model exists.
In this study, we assess the predictive power of three theoretical modelsvulnerability,
disorder, and social integrationon offense type-specic, emotionally-based fear of violent
and property crime. Results suggest that the predictors of fear of property and violent crime
vary, and some theoretical models are a better explanation of one type of fear than the other.
Overall, the models are a better t with fear of violent crime than fear of property crime. The
vulnerability model is more strongly related to fear of property crime, while the social integration model is more strongly related to fear of violent crime. Implications and suggestions
for future research are discussed.

Fear of crime has been a major research area, due in part to the growing awareness
that the consequences of fear reach beyond feelings of personal anxiety. It undermines
the quality of life, increases divisions between the rich and the poor, transforms public
places into places to avoid, leads to increased punitiveness, and can increase crime
(Hale 1996). In fact, fear of crime has been called a major social problem (Box
et al. 1988, p. 40). Although it is benecial to have some awareness and concern about
crime, and it is sensible for people to take precautions against victimization, fear can
be detrimental when taken to extremes. Fear of crime and its far-reaching ramications have been highlighted in the media and in policy slogans by politicians, and it
has long been an important topic for social scientists (Hale 1996).
The study of fear of crime has been complicated by several methodological
shortcomings. First, many of our generalizations about fear are based on research
which measures perceived risk, a cognitive assessment of the likelihood of future
victimization, rather than emotional fear or worry (Ferraro 1995). Second, the importance of measuring crime-specic fear with references to a specic crime or victimization rather than an abstract global measure has been raised (Ferraro 1995; Ferraro
and LaGrange 1987; Miethe and Lee 1984; Moore and Shepherd 2007; Rountree
Address correspondence to Mariel Alper, Department of Criminology and Criminal Justice,
University of Maryland, College Park, MD 20742-8235, USA. E-mail: malper@umd.edu
346

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

CRIME-SPECIFIC FEAR

347

and Land 1996; Warr 1984). Many studies do not operationalize fear in a crimespecic manner, and often the measure is combined into one general index. By failing
to measure fear of specic crimes separately, results can be misleading because important subtleties between fears of different types of crime can be masked (Miethe and
Lee 1984).
Researchers have proposed numerous explanations of fear of crime; all of
which contribute to the overall understanding of fear of crime. However, there is still
no comprehensive theoretical model of fear (see Hale 1996). The theoretical explanations can be grouped into three distinct models: the vulnerability model, the disorder
model, and the social integration model. The vulnerability model predicts that those
who feel unable to defend themselves, such as women, the elderly, and crime victims
will be more fearful, while the disorder model posits that neighborhood incivilities
increase feelings of fear. The social integration model focuses on the role of variables
such as social capital, cohesion, and collective efcacy in inhibiting fear. Although
numerous studies have examined the efcacy of these theories, a notable aw is that
they have not been compared within the same study to measure offense type-specic,
emotionally-based fear of crime.
The current state of research includes many inconsistencies and open questions, some of which the current study will address. Specically, we address two
research questions in this study. First, what is the explanatory power of each the vulnerability model, the disorder model, and the social integration model? Second, what
are the differences between these three models in explaining fear of property crime
and fear of violent crime? Our work will contribute to the extant research on fear
of crime by expanding on the small amount of research that has examined the three
theoretical models simultaneously. In addition, no previous studies have used this
three-model approach to measure crime type-specic fears. Finally, few studies have
examined differences in predictors of property versus violent crime.
Our research is valuable because fear of crime has extensive, far-reaching
effects on quality of life, and as such, it is an important topic for policy. Fear is
an emotion that determines peoples behaviors, including where they go and how
much they interact. When we know exactly what the origins of fear are, we can work
to reduce these causes and to increase feelings of safety. By reducing fear, social ties
can be built, residents will feel free to be more involved in their community, and
crime can be reduced (Hale 1996).

REVIEW OF THE LITERATURE


Measurement of Fear of Crime
Two measurement issues surround the study of fear of crime: the distinction
between the emotional and the cognitive components of fear and the measurement
of fear as crime-specic. Fear of victimization is an emotional feeling, and can be
described as how much someone worries about being victimized (Ferraro 1995;
Ferraro and LaGrange 1987). Perceived risk, on the other hand, is a cognitive judgment and is described as peoples estimations of crime rates and the resulting assessments of the likelihood of victimization. These two concepts are related, but are very
distinct. While a person may judge his or her risk of crime as being high, that person

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

348

M. ALPER AND A. T. CHAPPELL

may not necessarily be afraid (Ferraro 1995). Many early studies only measured
perceived risk, often by using some version of the NCVS question which asks respondents how safe they feel being out alone in their neighborhood at night. As a result,
many of our generalizations about fear are based on research which measures
perceived risk and neglects emotional fear (Ferraro 1995).
The second measurement issue surrounding the study of fear of crime is the
importance of examining crime-specic fears rather than general crime fears. In
his book, Ferraro (1995) discusses the necessity for references to the specic crime
or victimization rather than generic references to crime (also see Ferraro and
LaGrange 1987; LaGrange and Ferraro 1989; Moore and Shepherd 2007; Rountree
and Land 1996; Rountree 1998; Warr 1984). Indexes of specic crimes or victimizations are useful in measuring fear and risk of crime. Most importantly, they enable
a researcher to distinguish between types of crime when assessing a persons perceptions of crime (Ferraro 1995, p. 38).
There are many important reasons why it is essential to distinguish between fears
of different types of crimes. Miethe and Lee (1984) suggest that more than fear of violent crime, fear of property crime represents a rational response to ones previous
exposure to crime or assessment of ones environment, implying that the two fears have
different causes. The differing causes of fear of the two types of crimes would be lost if
only one measure of fear was used. In addition, combining both types of fear in the
same summation index can produce misleading results (Miethe and Lee 1984, p. 412).
Studies have found that the causes of fear differ depending on whether the
dependent variable is perceived risk, fear of property crime, or fear of violent crime.
Moore and Shepherd (2007) found that the NCVS question (perceived risk) was more
analogous to worry about physical harm than worry about property loss. Since
women feel more vulnerable to personal attacks, the use of this measure has
erroneously led to the belief that women are more fearful than men. The results
showed that women reported considerably higher fear of personal harm compared
to men. However, no gender differences were found for property loss. The authors
conclude that the assumption that women are more fearful of all types of crimes
should be rejected and that this claim likely originated from the use of perceived risk
as a dependent variable. Other research has also examined gender differences between
fears of different types of crimes (LaGrange and Ferraro 1989; Reid and Konrad
2004; Rountree and Land 1996; Schafer et al. 2006) and has found that women are
only more fearful than men in regards to violent crimes and sexual assault.
LaGrange and Ferraro (1989) conducted a study on the effects of age on different measures of fear. They found that the age differences were only present among
women when the NCVS question was used, and mens fear did not vary with age.
These differences were not present when any other measure of fear was used. A
possible explanation for this variation could be the fact that this question presents
a foreboding scenario that few people, particularly older, female respondents who
are unlikely to travel alone at night, are prone to experience (LaGrange and Ferraro
1989). Examining personal and property crimes separately, they also found that the
relationship between fear of personal crime and age was negative, meaning that
younger people were actually more afraid. For property crimes, there were no signicant relationships with age. They conclude that the NCVS measure is responsible for
the exaggeration of fear among older people.

CRIME-SPECIFIC FEAR

349

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

Theoretical Models
Past research has primarily focused on three theoretical models to explain fear
of crime. The rst two models, known as the vulnerability and disorder models,
focus on facilitators of fear, and the third model, known as the social integration
model, focuses on inhibitors of fear (Franklin et al. 2008). The vulnerability model
posits that those who feel unable to defend themselves, either physically or socially,
will experience more fear. According to this model, women, the elderly, those with
low SES, and victims of crime feel more vulnerable, leading to greater feelings of fear
(Clemente and Kleiman 1976; Hale 1996; Hindelang 1974; Hindelang et al. 1978;
Kennedy and Silverman 1985; Taylor and Hale 1986; Warr 1984).
The second model is the disorder model which originated from Shaw and
McKays (1972) work at the Chicago school. The disorder model stems from the
broken windows perspective, which posits that neighborhood incivilities inuence
perceptions of crime (Wilson and Kelling 1982). The thesis is that unresolved physical deterioration and social disorder grows progressively worse, escalating into more
serious crime. Social control is then reduced as increasing disorder causes residents
to become more fearful and withdraw from their neighborhood. In this way, incivilities and disorder lead to increased feelings of fear of crime (Doran and Lees 2005;
Hinkle and Weisburd 2008; Hunter 1978; Perkins and Taylor 1996).
The third model is the social integration model which holds that when residents
respond collectively to neighborhood problems, fear can be reduced (Hale 1996).
Increased social capital, social cohesion, and collective efcacy can mediate the
relationship between structural characteristics (such as SES, poverty, and ethnic heterogeneity) and fear of crime (Kanan and Pruitt 2002; Markowitz et al. 2001;
McGarrell et al. 1997; Renauer 2007; Taylor et al. 1984; Taylor and Hale 1986;
Box et al. 1988; DeLone 2008).

Three Model Studies


Because an exhaustive review of all studies examining each theoretical model is
beyond the scope of this review, here we discuss three studies that have examined all
three models and specically measured the strength of each model separately.
Although none of them explores crime-specic fear, they do lay the groundwork
for a three-model comparison.
Taylor and Hale (1986) were the rst to examine the three-model approach to
explaining fear of crime. Although it is somewhat conceptually different from the
models proposed in the current study, they included many of the same concepts as
the three models that will be used in this study. Additionally, they establish support
early in the literature for a multiple-model approach.
Taylor and Hale (1986) examined three perspectives, which they referred to as
disorder, community concern, and indirect victimization. The rst model they proposed was the indirect victimization model which combined vulnerability and social
ties with indirect victimization. This model suggested that more vulnerable people
(women, blacks, the elderly, and people with lower income) are more likely to be victimized or see crime. Victims or witnesses of crime will pass on this information
through their respective local support networks. Consequently, those with more social

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

350

M. ALPER AND A. T. CHAPPELL

ties and those who are more vulnerable will be more fearful, along with victims or witnesses of crime. The second model was the perceived disorder model. In this model, it
was proposed that lower social class and increased physical and social incivilities
heighten the perception of local problems which then elevates levels of fear. The nal
model was the community concern model. This model included the hypotheses of the
disorder model but added one extra link: concern about the future of the neighborhood. As objective characteristics such as crime, physical conditions, and SES lead
to perceived problems, concern about the future of the neighborhood and the quality
of present and future residents increases, leading to increased fear. Overall, Taylor and
Hale found that social networks and disorder were signicant predictors of fear, but
vulnerability was not. The authors concluded that fear is largely a function of the individuals position in the larger society, though SES predictors should not obscure the
constant roles played by perceived local conditions and neighborhood involvement.
McCrea and colleagues (2005) also examined the three-model approach, but
only as it related to perceived risk (though it is called fear throughout the study).
The dependent variable was measured by asking respondents their level of agreement
with the statement, I feel safe walking around this neighborhood after dark. To
measure the vulnerability hypothesis, the authors focused on age and gender. For
the disorder hypothesis, measures of physical disorder were used (social disorder
was not measured). The third model was based on social disorganization theory,
and it included measures of social capital, neighborhood involvement, neighborhood
friendliness, and sense of community. Findings suggested that the individual attribute variables from the vulnerability hypothesis were the most important for predicting risk, with females and older respondents reporting more risk. Neighborhood
disorder was the second strongest model, and the social processes and neighborhood
structural variables of the social disorganization model were the least explanatory.
Possibly, the results would have been different if social disorder was included as well
as other individual attributes that t with the vulnerability model.
The study that bears the closest resemblance to the current study is Franklin and
colleagues (2008) multilevel study of 21 cities across Washington State. They examined each of the three theoretical models (vulnerability, disorder, and social integration), using both perceived risk and worry about crime as the outcomes. To test
the vulnerability model, race, age, sex, education, and income were included in the
analysis.1 To test the disorder model, perceived disorder was measured by summing
the responses regarding the seriousness of eight physical and social neighborhood
problems. Lastly, the social integration model was a four-part measure that included
the questions Would you describe the area where you live as a place where people
help one another or mostly go their own way? Do you feel the area where you live
is more of a real home or more just like a place where you live? How often do you
talk with your neighbors? and When you do a favor for a neighbor, can you trust
the neighbor to return the favor? The results indicated that the disorder model
explained the greatest proportion of variation for both risk and worry. The social
integration model was the second strongest determinate, but it explained signicantly
less variation than the disorder model. The vulnerability model explained the least
1

The authors note that past victimization also ts the theme of the model; however, it was not
available in their data.

CRIME-SPECIFIC FEAR

351

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

amount of variation. Also, within the vulnerability model, the variables behaved
differently across the two dependent variables. When measuring perceived risk, the variables behaved as expected with minorities, females, the elderly, and those with lower
income and education levels reporting higher perceived risk. When examining worry,
however, race, age, income, and education all remained signicant but changed direction
as minorities, the elderly, and those with lower income and education levels displayed
lower levels of fear. Only gender remained consistent across measures with females displaying more risk and worry. In a later article, the authors extend their work by examining how the different models function across genders (Franklin and Franklin 2009).
Current Study
The current study lls a void in the extant literature by comparing the explanatory power of the three theoretical explanations of fear with offense type-specic,
emotionally-based fear of violent and property crime. We address two specic research
questions: (1) What is the explanatory power of the vulnerability model, the disorder
model, and the social integration model? (2) What are the differences between these
three models in explaining fear of property crime and fear of violent crime?
DATA AND METHODS
Data
The data for this project came from a 2008 telephone survey. The sample comes
from residents living in a medium-sized southeastern U.S. city with a population of
approximately 220,000. The purpose of the survey was to gain insight on neighborhood life, crime and safety, and police experiences. The sampling method began with
a pool of telephone numbers provided by a commercial sampling company. The
phone numbers were stratied by police precinct population and included a sample
of randomly generated telephone numbers supplemented by listed telephone numbers.
To conduct the survey, trained interviewers asked to speak with the person in
the household at least 18 years old who had the most recent birthday. Call attempts
were made in the evenings from 4:30 to 8:30 p.m. Monday through Friday and in the
mornings from 10:00 a.m. to 12:00 noon on Wednesday and Thursday. Numbers were
called a minimum of 10 times or until the number was determined to be invalid (such
as non-residential or out of service). There was an average of three call attempts for a
completed survey. If respondents initially gave a soft refusal by politely refusing,
the phone number was re-tried no sooner than three days later. After two soft refusals for a phone number, the third was coded as a hard refusal and removed from
the calling pool. The refusal rate was 40%. This resulted in a total of 628 successfully
completed interviews and a response rate of 22.5%.
Variables
Dependent variables. This study has two dependent variables: fear of
personal crime and fear of violent crime. These two variables were created from a
four-part question which asked, How often do you worry about the following things?

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

352

M. ALPER AND A. T. CHAPPELL

Do you worry about them often, sometimes, rarely, or never? (1) Having someone
break into your home; (2) Having your property vandalized; (3) Having property stolen from you; (4) Having yourself or family assaulted or harmed. To create the fear of
violent crime variable, one of the four parts was used. This question asked respondents
how much they worried about having yourself or family assaulted or harmed. The
responses were coded as 1 never, 2 rarely, 3 sometimes, and 4 often. The range
of the fear of violent crime variable was 1 to 4 and the mean was 1.69.
To create the fear of property crime question, an index was created with two
other parts of the question, which asked respondents how much they worried about
having your property vandalized and having property stolen from you. An
alpha reliability test showed that these questions had a high reliability of .824. These
questions coded so that a lower score indicated a lower level of worry. They were
then added together resulting in a range of 2 to 8 and a mean of 3.88.
Independent variables. Descriptive statistics for all variables are presented
in Table 1. Gender was coded as male 0 and female 1. The majority of the
Table 1 Descriptive characteristics
Variable
Gender
Male
Female
Age
Range 75
Mean 50.575
SD 15.317
Education
Less than high school
High school diploma or GED
Some college but no degree
Associates degree or technical training
Bachelors degree
Graduate or professional degree
Race
White
Other
Victimization
Yes
No
Disorder
Range 4.212
Mean .031
SD .920
Neighborhood trust
Not at all
A little bit
Somewhat
A lot
Organizational participation
Yes
No

225
403

35.8
64.2

30
144
145
80
120
96

4.9
23.4
23.6
13.0
19.5
15.6

413
199

67.5
32.5

61
567

9.7
90.3

31
58
192
335

5.0
9.4
31.2
54.4

216
410

34.5
65.5

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

CRIME-SPECIFIC FEAR

353

respondents were female (64.2%) while 35.8% were male. Age was continuous. Due
to the sampling method of landline telephones, the sample was generally older with
53.5% of respondents being age 50 or older. Education level was operationalized
with the question What is the highest grade of school or year of college you have
completed? The education variable was coded as less than high school diploma 1,
1, high school diploma or GED 2, some college but no degree 3, associates
degree or technical training 4, bachelors degree 5, and graduate of professional
degree 6. Race was recoded into two groups: nonwhite 0 (32.5%) and white 1
(67.5%).2
Victimization was operationalized by asking, During the past six months, have
any of the following things happened to you? The options were: Something belonging to YOU was stolen, such as: a wallet, clothing, jewelry, or other personal property;
Someone broke into your home; Someone stole or ATTEMPTED to steal one of your
motor vehicles (motorcycle or car); Someone tried to steal something from you by
force or by threatening you; Someone threatened you with a weapon (gun, knife, or
club); and, Someone physically attacked you. If the respondent answered that he
or she had been a victim of one or more of the crimes, victimization was coded as 1.
Due to the low number of victimizations (9.7%), all types of victimizations were combined. However, the majority of the victimizations were property victimizations.
The disorder variable was derived from a ten-part question which asked
respondents about ten types of disorder. The question asked, How often do you
encounter or see the following problems in your neighborhood? Abandoned=vacant
buildings, excessive noise, littering=garbage in streets, poor street lighting, public
drinking=intoxication, abandoned=inoperable vehicles, people hanging out in the
street (loitering), grafti, drug dealing or activities that look like drug dealing, and
prostitution. The options were almost never 1, sometimes 2, usually 3, and
almost always 4. A factor analysis showed that all types loaded into one factor
which explained 42.96% of the variance.3 Then, each item in the disorder scale
was normalized (using Bloms formula; see Blom 1958) to eliminate skewness and
kurtosis. A reliability analysis resulted in an alpha of .787 which indicates acceptable
reliability. Then, the ten items were combined into one disorder variable with a range
of 4.21, a mean of .031, and a standard deviation of .920.
Neighborhood trust was measured with the question How much do you trust
people in your neighborhood? The options were a lot, somewhat, a little bit, or not
at all. The variable was recoded so that not at all 1, a little bit 2, somewhat 3,
and a lot 4. Participation in neighborhood organizations was measured by asking
Do you participate in any neighborhood associations or activities? The options
were yes 1 or no 0.

According to the 2000 U.S. Census, the population of the city was 48.6% male and 51.4% female.
The median age was 34.7 and 28% of the residents had a high school (or equivalent) education while 25%
had a bachelors degree or higher. The racial make-up was 66.9% white, 28.5% black, and 2% Hispanic or
Latino. Compared to the general population, our sample is more likely to be older, female, and college
educated.
3
Although some past research has found that social and physical incivilities are related differently
to fear (LaGrange et al. 1992; Rohe and Burby 1988), the data did not support the creation of two separate
variables.

354

M. ALPER AND A. T. CHAPPELL

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

DATA ANALYSIS
We begin with correlations to examine the bivariate relationships between the
independent variables and the two dependent variables. Next, a series of three OLS
regression models are used for each of the two dependent variables, fear of property
crime and fear of violent crime. The rst model for each included the variables
related to the vulnerability model: race, age, gender, education, and victimization.
By denition, many of the variables in the vulnerability model are demographic variables and will be controlled for in the subsequent models. The second model included
the demographic variables (race, age, gender, and education) as well as disorder. The
third model included the demographic variables as well as the social integration
variables (trust and organizational participation). Next, to determine whether there
are three separate processes at work or if all three theoretical models are explaining
the same phenomenon, a nal set of regressions were run to control for the effects of
the other models. In all, there were eight OLS models.
RESULTS
Bivariate Analyses
To examine the relationships between the independent variables and the two
dependent variables, Pearson correlations were used (see Table 2). Gender is not signicantly correlated with either type of fear. While race is not signicantly correlated
with fear of violent crime, it does have a signicant relationship with fear of property
crime with whites being more fearful. Education is signicantly and positively correlated with both fear of property crime and fear of violent crime. The results of the
correlations also show that those who have been victimized, those who perceive more
disorder, and those who trust their neighbors less are more fearful of both types of
crime. Organizational participation, however, is not signicantly correlated with
either dependent variable.
Overall, the independent variables are more strongly correlated with fear of
property crime than fear of violent crime. The only exception is neighborhood trust
where the correlation with fear of violent crime was higher (r .291) than with fear
of property crime (r .245).
Multivariate Analyses
The rst set of three regressions was run with fear of violent crime as the dependent variable (Table 3). The rst model was a test of the vulnerability model and
included the variables of age, gender, race, education, and victimization. Recall that
the vulnerability model predicts that those who are older, female, have been previously
victimized, have less education, and are racial minorities will be more fearful of crime.
Some (2.8%) of the variation in fear of violent crime was explained (R2 .028).
Education was signicant at the .05 level with those having higher education
reporting more fear. Victimization was also positively associated with fear of violent
crime as those who had been victims of a crime were more fearful of violent crime.
The next model represented the disorder model and included the control
variables as well as the disorder variable. Some (11.1%) of the variation was

355

Education

Victimization

Disorder

Neighborhood trust

Org participation

Fear of roperty Crime

(5)

(6)

(7)

(8)

(9)

(10)

p < .01.

Race

(4)



Gender

(3)

p < .05;

Age

(2)

Fear of violent crime

(1)

Pearson
N
Pearson
N
Pearson
N
Pearson
N
Pearson
N
Pearson
N
Pearson
N
Pearson
N
Pearson
N
Pearson
N
r

1
625
.067
608
.058
625
.005
610
.080
614
.129
625
.300
583
.291
613
.016
624
.590
625

(1)

1
609
.025
609
.033
601
.133
602
.102
609
.148
567
.169
597
.031
608
.060
608

(2)

1
628
.087
612
.022
615
.002
628
.013
584
.041
616
.037
626
.022
625

(3)

1
612
.008
604
.014
612
.068
572
.254
600
.058
611
.159
610

(4)

Table 2 Correlations

1
615
.048
615
.066
574
.060
603
.136
614
.113
614

(5)

1
628
.157
584
.194
616
.023
626
.292
625

(6)

1
584
.378
573
.061
584
.364
583

(7)

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

1
616
.126
614
.003
624

(8)

1
626
.003
624

(9)

1
625

(10)

356

M. ALPER AND A. T. CHAPPELL


Table 3 OLS regression models predicting fear of violent crime
Model 1 (vulnerability)
n 594

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

B (SE)
Control variables
Age
Gender
Race
Education
Independent variables
Victimization
Disorder
Neighborhood trust
Organizational participation
R2


p < .05;



.002
.134
.02
.049

(.002)
(.077)
(.08)
(.029)

.376 (.126)

Beta

.035
.071
.01
.082

Model 2 (disorder)
n 555
B (SE)

.000
.138
.078
.063

(.002)
(.075)
(.078)
(.024)

Beta

.008
.074
.04
.108

Model 3 (social
integration) n 581
B (SE)

.000
.119
.165
.057

(.002)
(.075)
(.080)
(.024)

Beta

.003
.063
.084
.096

.122
.307 (.040) .318

.028

.360 (.045) .334


.065 (.017)
.034
.114

.111

p < .01.

explained in fear of violent crime (R2 .111). Again, education was positively
correlated with fear of violent crime. Disorder was also signicant, with those who
perceived more disorder being more fearful.
The third model was the social integration model which included the control
variables as well as neighborhood trust and organizational participation. Some
(11.4%) of the variation was explained in this model (R2 .114). Education was signicant with those having more education being more fearful of violent crime. Race was
also signicant with whites reporting higher levels of fear. Neighborhood trust was signicant with those who were less trusting displaying greater fear of violent crime.
The next set of regressions was run with fear of property crime as the dependent variable (Table 4). In the rst model, the vulnerability model, 12.7% of the
variation was explained (R2 .127). Race was signicant with whites reporting
Table 4 OLS regression models predicting fear of property crime
Model 1 (vulnerability)
n 594
B (SE)
Control variables
Age
.022
Gender
.172
Race
.674
Education
.153
Independent variables
Victimization
1.823
Disorder
Neighborhood trust
Organizational participation
R2
.127


p < .05;



p < .01.

Beta

(.005) .016
(.148)
.046
(.153)
.171
(.047)
.126
(.242)

Model 2 (disorder)
n 555
B (SE)

Beta

.003
.169
.827
.152

(.005)
(.148)
(.155)
(.047)

.027
.044
.208
.125

.757

(.078) .380

Model 3 (social
integration) n 581
B (SE)

.001
.180
.938
.138

(.005)
(.150)
(.162)
(.049)

Beta

.010
.047
.237
.114

.292

.183

.659 (.091) .303


.144 (.156)
.037
.120

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

CRIME-SPECIFIC FEAR

357

greater fear of property crime. Education was also signicant and positively correlated with fear of property crime. Those who had been victimized also reported
higher levels of fear of property crime.
The next model was the disorder model. Here, 18.3% of the variation was
explained (R2 .183). Again, race and education were signicant with whites and
those with more education displaying greater fear. Disorder was also signicant
and positively correlated with fear of property crime.
Lastly, the social integration model explained 12.1% of the variation of fear of
property crime (R2 .120). Whites and those with more education again reported
greater fear of property crime, and those reporting lower levels of trust of their
neighbors were also more fearful.
With regard to the rst research question, all three theoretical models were
signicant predictors of fear of crime, either in whole or in part. Measures of vulnerability, disorder, and social integration were signicant for both types of crime. Consistently, though, the models explained more of the variation in fear of property
crime than in fear of violent crime. Even the model with the highest R2 explained less
variance than the model with the lowest R2 explaining fear of property crime.
Additionally, there were differences in the strength of each model between the two
dependent variables. While the social integration model was the strongest predictor
of fear of violent crime, it was the weakest of the three models for fear of property
crime.
In regards to the second research question, some variables behaved differently
between the two types of crimes. For fear of violent crime, race was only signicant in
the social integration model and it was not signicant in the disorder or vulnerability
models. For fear of property crime, though, race was signicant in all three models.
Alternatively, age was not signicant in any model. This nding likely results
from the fact that fear was measured in terms of worry about crime rather than
perceived risk of crime. Prior research has shown that higher levels of fear among
the elderly are only present when the NCVS question (a measure of perceived risk)
is used as the dependent variable (LaGrange and Ferraro 1989). Gender was also not
signicant in any of the models.
Neighborhood trust was signicant for both types of crime while organizational participation was not signicant in either model. This is similar to Ross and
Jangs (2000) ndings that informal participation was a signicant predictor of fear,
but formal participation was not. This could be due to the fact that participation in
formal neighborhood organizations is less common than informal ties to neighbors,
and that participation in organizations can have trade-offs such as making residents
more aware of crime.
Although past literature has found mixed results on the relationship between
fear and victimization (Hale 1996), in this study, victimization was a signicant
predictor of both fear of property crime and fear of violent crime. This may be due
to the fact that victimization was operationalized by whether the respondent had
recently (within the past six months) been a victim of a specic criminal act. Because
this measure only includes recent and relatively serious offenses, it is more apt to capture victimization experiences that are likely to affect the respondent rather than
relatively minor or prior experiences that are unlikely to still have an effect. This

358

M. ALPER AND A. T. CHAPPELL

Table 5 OLS regression models predicting fear of violent crime and fear of property crime: Full models
Fear of violent crime n 544

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

B (SE)
Control variables
Age
Gender
Race
Education
Independent variables
Victimization
Disorder
Neighborhood trust
Organizational participation
R2
y

p < .10;  p < .05;



Beta

.03
.122
.171
.066

(.002)
(.074)
(.080)
(.024)

.043
.066y
.088
.111

.156
.236
.259
.051
.166

(.127)
(.042)
(.049)
(.076)

.050
.243
.237
.027

Fear of property crime n 544


B (SE)

Beta

.009
.140
.935
.164

(.005)
(.143)
(.155)
(.046)

.074y
.037
.234
.135

1.382
.614
.367
.104
.264

(.245)
(.080)
(.095)
(.146)

.216
.309
.164
.027

p < .01.

relationship between victimization and fear supports the vulnerability theory as those
who have been previously victimized were more likely to fear crime in the future.
As the independent variables for all three theoretical models were signicant,
the possibility still remained that all three models were explaining the same phenomenon. Victimization, disorder, and social integration could be related and the only
way to disentangle the three explanations would be to run a regression model
containing all of the variables suggested by the three theories. With this step, the
variables related to each theoretical model would be able to be tested while simultaneously controlling for the variables related to the other theoretical models. When
the variables pertaining to the alternate theories are controlled for, it can be seen
whether each individual theoretical explanation of fear can stand on its own, or
whether one theoretical model emerges as a better predictor. The results of the full
regression models are presented in Table 5.
In the rst model, fear of violent crime, 16.6% of the variation was explained.
The R2 for fear of property crime roughly doubled from the previous models, and
26.4% of the variation was explained. An examination of the R2 values showed that
the combined model explained more of the variation than any of the separate models. Additionally, consistent with the separate models, more of the variation in fear
of property crime was explained than in fear of violent crime.
While age was not signicant in any of the previous models, it approached signicance in the full model of fear of property crime with those who are older reporting higher levels of fear.4 However, age was not signicant for fear of violent crime.
Gender approached signicance for fear of violent crime and remained insignicant
for fear of property crime. Race also displayed a similar pattern as the previous models. Race was a strong predictor of property crime with whites displaying a higher
level of fear. Race was also a signicant predictor of fear of violent crime in the full
model, though it was only signicant previously in the social integration model.
4

This may be due to a suppression effect using listwise deletion or due to a small difference in the
Ns between the dependent variables..

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

CRIME-SPECIFIC FEAR

359

While victimization was a signicant predictor of both fear of property crime and
fear of violent crime in the previous models, it lost signicance in the nal model
predicting fear of violent crime. However, it remained signicant for fear of property
crime. This result is consistent with Miethe and Lees (1984) ndings that
crime-related variables such as prior victimization are a stronger predictor of fear
of property crime than fear of violent crime.
Disorder and neighborhood trust remained signicant predictors of both
dependent variables. Those who trusted their neighbors less and those who perceived
more disorder expressed higher levels of fear of property crime and fear of violent
crime. Like all previous models, organizational participation was not signicant.
As the independent variables generally remained signicant even after the variables related to the alternative theoretical models were controlled for, it is evident that
all three theoretical explanations are not explaining the same phenomenon, lending
further support to the theories. Disorder explains fear of property and violent crime
beyond its relationship with neighborhood trust and vice versa. This supports a
multi-theoretical approach as there are multiple processes that are contributing to fear
of crime and one theory cannot account for all of the explanations. Multiple explanations do not take away from each other; rather, they strengthen the overall explanation.
DISCUSSION
Although an abundance of research has been conducted on fear of crime, there
is still no clear comprehensive theoretical model of fear (Hale 1996). While it has
been established that there are substantial differences in the causes of fear of different types of crime (Miethe and Lee 1984; Ross and Jang 2000; Rountree 1998; Warr
1987), the efcacy of existing theoretical explanations has not been compared within
the same study to measure offense type-specic, emotionally-based fear. The current
study addressed several methodological issues in an attempt to uncover the differences in the predictors of fear of property crime and fear of violent crime. Using
offense type-specic dependent variables, the strength of three theories (the vulnerability model, the disorder model, and the social integration model) was compared.
The primary objective of this study was to compare the explanatory power of the
three proposed theoretical models and to uncover the differences between the three
models in explaining fear of property crime and fear of violent crime.
Results indicate that all three theoretical models help to explain fear of crime.
Aspects of vulnerability, disorder, and social integration were signicant predictors
of both fear of property crime and fear of violent crime. However, these explanations
appear to be better predictors of fear of property crime as they consistently explained
more of the variation in fear of property crime than in fear of violent crime. This
suggests that alternative explanations may be needed to explain fear of violence.
For example, psychological or cognitive explanations may be better suited to explain
fear or violent crime, and factors such as exposure to violence and attitudes towards
violence may help to explain such fear.
The best predictor of fear varies between the two crime types. The strongest
model predicting fear of violent crime was the social integration model.
Alternatively, the strongest model predicting fear of property crime was the disorder
model, and the social integration model was the weakest model in predicting such

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

360

M. ALPER AND A. T. CHAPPELL

fear. So, trusting ones neighbors has more of an effect on fear of violent crime than
perceptions of disorder or prior victimization. But, trust has less of an effect on fear
of property crime than perceptions of disorder or prior victimization.
While the disorder and social integration models received consistent support
throughout the analyses, the vulnerability model was less consistent. When the variables
from the other theoretical models were added, the effects of victimization on fear of violent crime disappeared. While those who have been victimized (primarily by property
crimes in this data set) are more fearful of violent crime, that variation is explained away
by the level of perceived disorder and the level of perceived trust. Previous research has
shown that property victimization only increases fear of property crime while violent victimization increases fear of both types of crime (Rountree 1998). Since most of the victimizations reported in this study were property crime victimizations, it would make
sense that victimization would not have much of an effect on fear of violent crime. A
stronger relationship may have been found if measures of crime-specic victimizations
had been used, something which future research should investigate.
Additionally, other correlates of the vulnerability model did not follow the
hypothesized relationships. Neither race nor education behaved in the expected direction as minorities and those with less education were both less fearful. Age and gender
were not consistent predictors of fear. It does not appear that the idea of vulnerability
as conceptualized in this study is strongly related to fear, and it may be another
phenomenon that is at work. Other factors, such as physical vulnerability including
physical condition, handicaps, and whether the person would be seen as an easy target
from the point of view of the attacker, may provide a better explanation (c.f. Killias
and Clerici 2000; Stiles et al. 2003). Conceptually, it would seem that physical vulnerability would be particularly useful in explaining fear of violent crime.
The question of the role of age and gender has been a long-standing paradox in fear
of crime research (Hale 1996). By separating the crime types, the role of these two characteristics was uncovered where it might otherwise have been unclear or misleading.
Although neither was very strongly related to fear, it is evident that if females are more
fearful, it is only true for fear of violent crime. Additionally, if older people are more fearful, it is only applicable to fear of property crime (LaGrange and Ferraro 1989).
The theoretical implications of these results are numerous. First, the separate
and cumulative models show that there are multiple processes at work and that
no one theoretical model can account for the all of the variation. This highlights
the importance of accounting for the several channels through which fear develops,
rather than accounting for only one individual model. Second, while many of these
variables make some contribution to the explanation of fear of crime, some are better suited to explain one type of crime more than the other. For example, race and
education were better predictors of fear of property crime than fear of violent crime.
Third, support is established for the separation of crime types. This aligns with prior
research which has found signicant differences in the predictors of fear for different
specic crimes (Miethe and Lee 1984; Ross and Jang 2000; Rountree 1998;
Warr 1987).
In addition to theoretical implications, this study also has practical contributions in regards to the design and implementation of programs designed to reduce
crime by reducing fear. For example, a more effective program would target multiple
sources of fear rather than just one. Additionally, a program that reduced disorder

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

CRIME-SPECIFIC FEAR

361

and victimizations or builds neighborhood trust would be more effective than a program that encouraged participation in neighborhood organizations.
This study has limitations, like all research. As the study is cross-sectional,
causality and the possibility of reciprocal relationships cannot be addressed. Few
fear of crime studies are longitudinal (for exceptions, see Robinson et al. 2003;
Markowitz et al. 2001). As the original purpose of this data was not to measure
police satisfaction, some of the measures are less than ideal. For example, fear of
crime is measured using levels of worry, rather than using the word afraid as used
by Ferraro (1995). However, worry about crime is an often-used measure of fear,
and the concept of worry has been found to be a preferable way to measure the
vague term of fear of crime (Williams et al. 2000) Further, it reects the way people
dene fear of crime in their daily conversations (Lane and Meeker 2000). Due to the
low number of victimizations, the victimization variable was not crime type-specic.
Prior research has shown that violent and property victimizations affect fear of specic crimes differently (Miethe and Lee 1984; Rountree 1998). This was supported in
the present study by the fact that, as most of the victimizations were from property
crimes, the relationship with fear of property crime was stronger than with fear of
violent crime. Second, measures of social integration lack operational uniformity
with prior research. While this study uses two possible measures of integration, there
are multiple other ways that this concept could be operationalized and that may provide a more thorough test of the model. Lastly, because of the sampling method of
landline telephones, the response rate was low and the sample was older, more educated, and contained a higher percentage of females than the general population,
raising questions about the generalizability of the sample.
The ndings from this study conrm the importance of using crime-specic
measures of fear as non-specic measures may obscure the results. This study also
found that support for several different theories of fear of crime and established
that multiple processes are at work that cannot all be accounted for by any one
of the models. Additionally, the models tested in this study were better able to
explain fear of property crime than fear of violent crime. Future studies may also
explore other processes which may be at work such as cognitive processes or
physical characteristics that may be better explanations of fear of violent crime.
We recommend that future researchers replicate this study with larger samples
and more comprehensive measures of social integration and crime-type specic
victimization.
AUTHOR NOTES
Mariel Alper is a PhD student in criminology and criminal justice at the University
of Maryland. She received her masters degree in applied sociology from Old Dominion
University in 2010. Her research focuses on fear of crime and perceptions of crime.
Allison T. Chappell is Assistant Professor in the Department of Sociology and
Criminal Justice at Old Dominion University in Norfolk, Virginia. She earned her
PhD in sociology from the University of Florida in 2005. She conducts research in
the areas of policing and juvenile delinquency. Her work has appeared in journals
such as Journal of Contemporary Ethnography, Crime and Delinquency, and Justice
Quarterly.

362

M. ALPER AND A. T. CHAPPELL

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

REFERENCES
Blom, Gunnar. 1958. Statistical Estimates and Transformed Beta Variables. New York: Wiley.
Box, Steven, Chris Hale, and Glen Andrews. 1988. Explaining Fear of Crime. British Journal
of Criminology 28(3):340356.
Clemente, Frank and Michael B. Kleiman. 1976. Fear of Crime Among the Aged. Gerontologist 16:207210.
DeLone, Gregory J. 2008. Public Housing and the Fear of Crime. Journal of Criminal
Justice 36:115125.
Doran, Bruce J. and Brian G. Lees. 2005. Investigating The Spatiotemporal Links between
Disorder, Crime, and the Fear of Crime. The Professional Geographer 57:112.
Ferraro, Kenneth F. 1995. Fear of Crime: Interpreting Victimization Risk. Albany: State
University of New York Press.
Ferraro, Kenneth F. and Randy LaGrange. 1987. The Measurement of Fear of Crime.
Sociological Inquiry 57:70101.
Franklin, Cortney A. and Travis A. Franklin. 2009. Predicting Fear of Crime: Predicting
Differences Across Gender. Feminist Criminology 4:83106.
Franklin, Travis W., Cortney A. Franklin, and Noelle E. Fearn. 2008. A Multilevel Analysis
of the Vulnerability, Disorder, and Social Integration Models of Fear of Crime. Social
Justice Research 21:204227.
Hale, Chris. 1996. Fear of Crime: A Review of the Literature. International Review of
Victimology 4(2):79150.
Hindelang, Michael J. 1974. Public opinion regarding crime, criminal justice, and related
topics. Journal of Research in Crime and Delinquency 11:101106.
Hindelang, Michael J., Michael Gottfredson, and James Garofalo. 1978. Victims of Personal Crime:
An Empirical Foundation for a Theory of Personal Victimization. Cambridge: Ballinger.
Hinkle, Joshua C. and David Weisburd. 2008. The Irony of Broken Windows Policing: A
Micro-Place Study of the Relationship between Disorder, Focused Police Crackdowns
and Fear of Crime. Journal of Criminal Justice 36:503512.
Hunter, Albert. 1978. Symbols of Incivility: Social Disorder and Fear of Crime in Urban
Neighborhoods. Presented at the annual meeting of the American Society of Criminology, Dallas, TX.
Kanan, James W. and Matthew V. Pruitt. 2002. Modeling Fear of Crime and Perceived
Victimization Risk: The (In)signicance of Neighborhood Integration. Sociological
Inquiry 74:527548.
Kennedy, Leslie W. and Robert A. Silverman. 1985. Perceptions of Social Diversity and Fear
of Crime. Environment and Behavior 17:275295.
Killias, Martin and Christian Clerici. 2000. Different Measures of Vulnerability in Their
Relation to Different Deminsions of Fear of Crime. British Journal of Criminology
40:437450.
Knoke, David, George W. Bohrnstedt, and Alisa Potter Mee. 2002. Statistics for Social Data
Analysis, 4th edition. Belmont, CA: Wadsworth.
LaGrange, Randy L. and Kenneth F. Ferraro. 1989. Assessing AGE and Gender Differences
in Perceived Risk and Fear of Crime. Criminology 27:697720.
LaGrange, Randy L., Kenneth Ferraro, and Michael Supancic. 1992. Perceived Risk and
Fear of Crime. Journal of Research in Crime and Delinquency 29:311334.
Lane, Jodi and James W. Meeker. 2000. Subcultural Diversity and the Fear of Crime and
Gangs. Crime & Delinquency 46:497521.
Markowitz, Fred E., Paul E. Bellair, Allen E. Liska, and Jianhong Liu. 2001. Extending
Social Disorganization Theory: Modeling the Relationships Between Cohesion, Disorder,
and Fear. Criminology 39:293320.

Downloaded by [Facultad Latinoamericana de Ciencias Sociales] at 14:54 27 March 2015

CRIME-SPECIFIC FEAR

363

McCrea, Rod, Tung-Kai Shyy, John Western, and Robert J. Stimson. 2005. Fear of Crime in
Brisbane: Individual, Social and Neighborhood Factors in Perspective. Journal of
Sociology 41:727.
McGarrell, Edmund F., Andrew L. Giacomazzi, and Quint C. Thurman. 1997. Neighborhood Disorder, Integration, and the Fear of Crime. Justice Quarterly 14:479500.
Miethe, Terance D. and Gary R. Lee. 1984. Fear of Crime Among Older People: A
Reassessment of the Predictive Power of Crime-Related Factors. Sociological Quarterly
25:397415.
Moore, Simon C. and Jonathan Shepherd. 2007. Gender Specic Emotional Responses to
Anticipated Crime. International Review of Victimology 14(3):337351.
Perkins, Douglas D. and Ralph B. Taylor. 1996. Ecological Assessments of Community
Disorder: Their Relationship to Fear of Crime and Theoretical Implications. American
Journal of Community Psychology 24:63107.
Reid, Lesley W. and Miriam Konrad. 2004. The Gender Gap in Fear: Assessing the Interactive Effects of Gender and Perceived Risk on Fear of Crime. Sociological Spectrum
24:399425.
Renauer, Brian C. 2007. Reducing fear of crime: Citizen, police, or government responsibility? Police Quarterly 10:4162.
Robinson, Jennifer B., Brain A. Lawton, Ralph B. Taylor, and Douglas D. Perkins. 2003.
Multilevel Longitudinal Impacts of Incivilities: Fear of Crime, Expected Safety, and
Block Satisfaction. Journal of Quantitative Criminology 19:237274.
Rohe, William M. and Raymond J. Burby. 1988. Fear of Crime in Public Housing. Environment and Behavior 20:700720.
Ross, Catherine E. and Sung J. Jang. 2000. Neighborhood Disorder, Fear, and Mistrust:
The Buffering Role of Social Ties with Neighbors. American Journal of Community
Psychology 28:401420.
Rountree, Pamela W. 1998. A Reexamination of the Crime-Fear Linkage. Journal of
Research in Crime and Delinquency 35:341372.
Rountree, Pamela W. and Kenneth C. Land. 1996. Perceived Risk Versus Fear of Crime:
Empirical Evidence of Conceptually Distinct Reactions in Survey Data. Social Forces
74:13531376.
Schafer, Joseph A., Huebner, Beth M., and Bynum, Timothy S. 2006. Fear of Crime and
Criminal Victimization: Gender-Based Contrasts. Journal of Criminal Justice 34:285301.
Shaw, Clifford R. and Henry D. McKay. 1972. Juvenile Delinquency and Urban Areas, Revised
edition. Chicago: The University of Chicago Press.
Stiles, Beverly L., Shaheem Halimand, and Howard B. Kaplan. 2003. Fear of Crime among
Individuals with Physical Limitations. Criminal Justice Review 28:232253.
Taylor, Ralph B., Stephen D. Gottfredson, and Brower Sidney. 1984. Block Crime and Fear:
Defensible Space, Local Social Ties, and Territorial Functioning. Journal of Research in
Crime and Delinquency 21:303331.
Taylor, Ralph B. and Margaret Hale. 1986. Testing Alternative Models of Fear of Crime.
Journal of Criminal Law and Criminology 77:151189.
Warr, Mark. 1984. Fear of Victimization: Why Are Women and the Elderly More Afraid?
Social Science Quarterly 65:681702.
. 1987. Fear of Victimization and Sensitivity to Risk. Journal of Quantitative
Criminology 3:2946.
Williams, Frank P., Marilyn D. McShane, and Ronald L. Akers. 2000. Worry about Victimization: An Alternative and Reliable Measure for Fear of Crime. Western Criminology
Review 2(2):126.
Wilson, James Q. and George Kelling. 1982. The Police and Neighborhood Safety: Broken
Windows. Atlantic Monthly 127:2938.

You might also like