Adolescent fear of crime: Testing Ferraro’s risk interpretation model

Abstract

The aim of this study is to broaden the scope of knowledge on fear of crime by examining if Ferraro’s risk interpretation model of fear of crime also holds true for the adolescent population. Using data on approximately 1500 Czech students in middle and grammar school, we test three different models. First, the classic model of fear of crime, applied originally to adults, is estimated. Second, the role of perceived school disorganization is examined, taking into account that adolescents spend a substantial part of the day at school. Finally, we propose an extension of Ferraro’s model by considering parental supervision as a factor influencing adolescent fear of crime through risk perception and avoidance behaviour. The results indicate that Ferraro’s original model of fear of crime can be appropriately applied to both adults and adolescents. Nevertheless, adolescent risk perception, which remains the most influential determinant of individual fear, seems to be influenced by stimuli stemming from the school rather than the neighbourhood environment. Furthermore, the relationship between parental supervision and fear of crime is mediated by avoidance behaviour, while no direct effect of parental supervision on risk perception and fear of crime was found in the data.

Keywords

Adolescent fear of crime Czech Republic disorganization parental supervision risk perception

Introduction

Research on fear of crime among adults has a long tradition. The first studies started to appear as early as the 1970s in order to explore fear of crime levels among the public, to find out whether there is a rational explanation for individual fear and to describe to what extent it influences people’s everyday lives (see, for example, Furstenberg, 1971; Garofalo, 1979; Lotz, 1979; President’s Commission on Law Enforcement and Administration of Justice: Task Force Report, 1967). It was subsequently revealed that fear of crime presents a complex social phenomenon that is influenced by a number of individual (gender, age, victimization experience) and social (neighbourhood disorganization) factors. In addition to theories about victimization, vulnerability, social incivilities and social disorganization that investigate why certain people are more prone to fear crime, the influential work of Ferraro (1995) introduced a risk interpretation model of fear of crime, emphasizing the role of risk perception as a strong correlate of fear of crime and constrained behaviour and as a mediator between the aforementioned individual and social factors and fear of crime. Later, the importance of the relationship between risk perception and fear of crime, and the mediating effect of risk perception, were also confirmed by other authors (Farrall et al., 2009; Gainey et al., 2011; Jackson, 2004).

Whereas research on adult fear of crime is well established, knowledge about adolescent fear is still largely limited (De Groof, 2008; May et al., 2002). Overall, existing research has shown that adolescent fear of crime is likely to be influenced by similar factors as in the case of adults. Some authors, however, argue that, unlike adults, the decisions adolescents make, the activities they undertake and the people they encounter can be greatly influenced by their parents. Attention is therefore paid to the level of control parents exercise over their offspring (Cops, 2013; De Groof, 2008; May et al., 2002, 2015; Wallace and May, 2005). On the other hand, we have found no study that would test Ferraro’s risk interpretation model of fear of crime (see Ferraro, 1995) among adolescents and that would allow us to study the complex relationships between adolescent fear of crime, risk perception, neighbourhood or school disorganization, constrained behaviour and, possibly, parental supervision.

The first aim of this study is therefore to determine if Ferraro’s (1995) risk interpretation model of fear of crime also holds true for the adolescent population. In addition to testing the original model, which postulates perceived neighbourhood disorganization as one of the factors influencing risk perception and consequently fear of crime, we take into account that adolescents spend a substantial part of the day at school. We thus also test the role of perceived school disorganization in informing adolescent fear of crime. Finally, and with respect to studies pointing to the significance of the relationship between parenting style and fear of crime, we propose an extension of the model by including parental supervision as a factor influencing adolescent fear of crime through risk perception and avoidance behaviour.

Risk interpretation model of fear of crime

Fear of crime, risk perception and constrained behaviour

Although research on fear of crime is well established, the conceptualization and operationalization of fear of crime within crime and victimization surveys has been a subject of debate (Hale, 1996; Warr, 2000). Research on fear of crime initially included a ‘general’ indicator in questionnaires, asking respondents about their feeling of safety in their neighbourhood. Later on, however, batteries of questions focusing on explicit worries about different types of offences were preferred over the simple indicator of feeling of safety (Ferraro and LaGrange, 1987; Warr and Stafford, 1983). For example, Gainey et al. (2011) measure the fear of: having someone break into their home, having their property vandalized, having property stolen from them and having themselves or their family assaulted or harmed. Farrall et al. (2009) consider two indicators of fear of crime, one of which measures everyday worry about crime, capturing its intensity, and the other measures anxiety about crime, capturing its frequency. Offences included in the analysis are burglary, robbery and car theft. In the 1980s, Ferraro and LaGrange (1987: 72) defined fear of crime as ‘a negative emotional reaction to crime or the symbols associated with crime’ and pointed to the difference between emotionally based fear and risk perception, which is rather based on cognition. Similarly, the difference between individual fear and behavioural responses to risk, often referred to as constrained behaviour, should be made clear (Doran and Burgess, 2012; Liska et al., 1988).

Risk perception can be defined as the subjectively assessed likelihood of victimization. Similarly to fear of crime, risk perception is usually measured with a battery of questions using the same range of offences as in the case of fear of crime indicators. Nevertheless, instead of examining the respondent’s fear, the aim is to assess his or her perceived risk of being victimized by these offences (see Farrall et al., 2009). According to Ferraro (1995), people evaluate their risk based on who they are and how they live. Drawing on symbolic interactionism, social incivilities and criminal opportunity theories, the author argues that risk perception is predominantly influenced by individual characteristics or the environment in which the individual lives and, more importantly, the way he or she interprets it.

Finally, constrained behaviour refers to activities people do to reduce their risk of victimization. In other words, they tend to avoid situations or places that could possibly lead to victimization or protect themselves by different means such as alarms or weapons. Whereas Liska et al. (1988: 829–30) consider only two indicators of constrained behaviour, asking respondents how often they go out in the evening for entertainment and whether they have limited or changed their activities in the past year due to crime, Ferraro (1995: 56) takes into account both avoidance and defensive dimensions. The first is therefore measured by indicators such as ‘Do you avoid unsafe areas during the night because of crime?’, the latter by indicators such as ‘[Have you] installed extra locks on windows or doors?’ or ‘[Have you] started carrying something to defend yourself?’

The relationship between fear of crime and risk perception is well documented (Farrall et al., 2009; Ferraro, 1995; Gainey et al., 2011; Jackson, 2004; LaGrange et al., 1992; Rountree and Land, 1996; Warr, 1984; Warr and Stafford, 1983). Those who find themselves at risk of becoming a victim of crime fear crime to a greater extent than those who perceive their victimization risk to be lower. As Ferraro (1995) and LaGrange et al. (1992) argue, risk perception increases fear of crime and (largely) mediates the effect of individual (for example, gender, age, victimization etc.) and social (for example, social incivilities and disorganization) factors. The same was later confirmed by other studies (Farrall et al., 2009; Gainey et al., 2011; Jackson, 2004; Wyant, 2008). On the other hand, findings about the relationship between fear of crime and constrained behaviour are not as unequivocal as in the case of fear of crime and risk perception. Some authors argue that constrained behaviour is a cause of fear of crime (Ferraro, 1995; Rountree and Land, 1996), whereas others see it as a result (Garofalo, 1981; Williams et al., 1994). Moreover, Liska et al. (1988) found that there is a reciprocal relationship between both phenomena: fear of crime can lead to constrained behaviour as much as constrained behaviour can further heighten individual fear of crime. Unlike the other authors, Ferraro (1995) considers all of the phenomena – fear of crime, risk perception and constrained behaviour – in one model. He argues that risk perception plays an important role in eliciting both fear of crime and constrained behaviour. His analysis also confirms that constrained behaviour heightens fear of crime instead of reducing it. More precisely, people who find themselves at risk of victimization constrain their behaviour and fear crime to a greater extent than their counterparts. Constrained behaviour, however, does not lower individual fear of crime, but rather increases it.

Social incivilities, vulnerability and victimization experience

As has already been mentioned, fear of crime and especially risk perception are shaped by factors related to the environment in which people live, individual vulnerability and victimization experience (Doran and Burgess, 2012; Farrall et al., 2009; Hale, 1996). One of the first studies to address the relationship between social incivilities and fear of crime was conducted by Lewis and Salem (1986), who argued that fear of crime is a result of the disruption of social control (disorganization) in the area rather than individual experience with crime. According to the authors, people tend to declare higher fear of crime in areas that are associated with a higher incidence of physical (vandalism, broken windows, dilapidated buildings and abandoned cars, trash and litter) or social (beggars, rowdy youth, inconsiderate neighbours) incivilities. Later, LaGrange et al. (1992), Ferraro (1995) and Farrall et al. (2009) examined the relationship between risk perception, fear of crime and social disorganization and found that the perceived level of social incivilities in a neighbourhood significantly influences individual risk perception, which in turn heightens fear of crime.

The vulnerability theory postulates that certain people are more prone to fear crime and perceive their victimization risk to be higher than that of others (Killias, 1990). A number of studies have confirmed that women, seniors and those with lower socioeconomic status (SES) are more fearful than men, younger people and those with higher education or more financial resources (Jackson and Stafford, 2009; LaGrange and Ferraro, 1989; Russo et al., 2013; Tseloni and Zarafonitou, 2008; Weinrath and Gartrell, 1996). The most consistent relationship is, however, found in the case of gender. Whereas women are more sensitive to risk and tend to perceive the consequences of potential victimization to be more serious (Warr, 1984, 1985), men either do not like to admit to fear or underestimate their potential victimization risk (Agnew, 1985; Gilchrist et al., 1998; Sutton and Farrall, 2005). In his risk interpretation model of fear of crime, Ferrarro (1995) confirms a moderate relationship between gender, risk perception and fear of crime and a strong relationship between gender and constrained behaviour. Women perceive their victimization risk to be higher and fear crime or change their behaviour to a greater extent than do men. With respect to education, Ferraro (1995) found no correlation with risk perception. On the other hand, lower education is to some extent related to higher fear of crime.

Finally, there is a relationship between victimization experience and fear of crime, risk perception and perceived social disorganization. According to a number of studies (Akers et al., 1987; Andreescu, 2010; Farrall et al., 2009; Hanslmaier, 2013; Tseloni and Zarafonitou, 2008), people who have become victims of crime fear crime to a greater extent than those who have no such experience; in the same vein, they perceive their victimization risk to be higher. Nevertheless, there is evidence that there are people who fear crime even though they have never become victims of crime themselves (Doran and Burgess, 2012; Hale, 1996). Such a result is usually explained by individual vulnerability (see above) or the fact that fear of crime can also be enhanced by indirect victimization, that is, vicarious experience of crime through people the individual is in frequent contact with, such as family members, friends, neighbours (Skogan and Maxfield, 1981; Taylor and Hale, 1986), or through the media (Gerbner and Gross, 1976; Hale, 1996; Reiner, 2002; Sacco, 1995). According to Ferraro (1995), being or knowing a victim of crime is associated with a higher perception of neighbourhood disorganization and risk perception, which in turn influences fear of crime. A stronger relationship is, however, found between victimization and perceived neighbourhood disorganization than risk perception. The same result was also reached by Gainey et al. (2011), who report a moderate relationship between these phenomena, while the correlation between victimization and risk perception was found to be weak.

Fear of crime among adolescents

There are several reasons why it is important to pay attention to fear of crime among the adolescent population. As Hashima and Finkelhor (1999) and Finkelhor et al. (2009) point out, adolescents become victims of crime more often than adults. Furthermore, adolescents are assumed to be more vulnerable than adults since it is more difficult for them to avoid or resist potential attacks. For example, they cannot easily avoid school attendance and, compared with adults, their physical strength is lower. Given the higher vulnerability of adolescents and their increased likelihood of victimization, we can expect their fear, as well as the consequences of such fear, to be greater or more serious than in adults. It is therefore surprising that research focusing on adolescent fear of crime is relatively scarce and, to our knowledge, there is no study aimed at testing Ferraro’s risk interpretation model of fear of crime with respect to the young population, even though findings related to adolescent fear of crime and its correlates can be crucial for our understanding of the causes of fear of crime among adults (Goodey, 1994).

The existing studies dealing with adolescent fear of crime usually assume that juvenile fear is likely to be influenced by similar factors as in adults. Such studies therefore frequently take into account factors such as gender, age, family SES, victimization experience and perceived neighbourhood (or school) disorganization. Similarly, as in the case of adults, gender remains the most consistent individual factor influencing fear of crime; girls are thus found to be more fearful than boys (Cops, 2010; May and Dunaway, 2000; Wallace and May, 2005). With respect to age, SES and victimization, the evidence is mixed. May et al. (2002) point out a significant relationship between fear of crime, age and economic status, whereas Cops (2010) finds no evidence of such a relationship. Furthermore, the relationship between victimization and fear of crime is confirmed by De Groof (2008) and Wallace and May (2005), whereas other studies find only a weak (Cops, 2010, 2013) or non-existent (May et al., 2015) correlation. On the other hand, perceived neighbourhood disorganization seems to be closely linked to fear of crime, not only in adults but also in relation to adolescents. Moreover, adolescents spend a substantial part of the day at school. According to a number of studies, higher disorganization in the school environment substantially increases their fear of crime (Cops, 2010; May et al., 2002; May and Dunaway, 2000).

The role of parental supervision

Since children are, unlike adults, to a certain degree influenced by their parents in their decision-making and activities, the impact of parenting style is often taken into account when studying adolescent fear of crime (De Groof, 2008; May et al., 2002, 2015). According to De Vaus and Wise (1996), there is a link between parents’ fear and their children’s fear, suggesting that parental fear is likely to be partly transmitted to the children and that children seem to ‘internalize the perceptions of their parents about the safety of the world “out there”’ (De Vaus and Wise, 1996: 37).

An important aspect of parenting style that can substantially affect the activities of adolescents and that aims to ensure their safety and proper behaviour is parental supervision. A number of studies have confirmed a positive correlation between parental supervision and fear of crime, that is, adolescents who are closely monitored by their parents fear crime to a greater extent than those who declare only limited parental supervision (Cops, 2010, 2013; De Groof, 2008; May et al., 2002; May et al., 2015). Common interpretations suggest that high parental supervision reminds adolescents of the potential danger of victimization and that such a parenting style does not provide them with effective strategies for coping with risk, which results in a higher expressed fear of crime (for example, May et al., 2002). However, with respect to risk perception, the study by May et al. (2002) reveals that high parental supervision actually lowers the adolescent’s perceived victimization risk. The authors suggest that there is a ‘sheltering effect’: because close parental supervision often limits the child’s possibility to socialize with his/her peers and face dangerous situations, the child thus assesses his/her victimization risk to be rather low.

The current study

To our knowledge, studies that focus on adolescent fear of crime are limited and there is no study that would test Ferraro’s (1995) risk interpretation model of fear of crime among adolescents. Therefore, the first aim of the study is to examine whether this model can also be applied to the adolescent population. The core of the theory is articulated in the first three hypotheses:

Hypothesis 1: There is a strong relationship between risk perception and fear of crime. Those who perceive their victimization risk to be high fear crime to a greater extent than those who perceive it to be relatively low.

Hypothesis 2: Risk perception is associated with avoidance behaviour (in the sense that those who perceive their risk of victimization to be high tend to use different behavioural strategies to protect themselves from victimization), which in turn leads to higher fear of crime.

Hypothesis 3: Risk perception mediates the influence of other individual and social factors, that is, gender, family SES, victimization experience and perceived neighbourhood disorganization, on fear of crime.

In addition to testing the classic model, we argue that adolescents spend a substantial part of the day at school and that their perception of the school environment can significantly influence their risk perception and consequently their fear of crime (May and Dunaway, 2000). We thus estimate a second model where we examine the role of perceived school disorganization.

Hypothesis 4: Perceived school disorganization is positively correlated with risk perception and consequently fear of crime.

Finally, since there is evidence (De Vaus and Wise, 1996) that parents can transmit some of their fears to their children and that the level of control that parents exercise over their children can significantly influence their fear of crime and risk perception (Cops, 2013; De Groof, 2008; May et al., 2002, 2015), we propose an extension of Ferraro’s (1995) model by incorporating parental supervision as a factor relevant for the adolescent population (see Figure 1). In light of the study by May et al. (2002) which indicates a positive relationship between parental supervision and fear of crime and a negative relationship between parental supervision and risk perception, we believe that the influence of parental supervision on fear of crime is complex because it is the product of two opposite forces. On the one hand, parental supervision is, indeed, likely to decrease risk perception and consequently fear of crime. On the other hand, we suggest that parental supervision is linked with increased avoidance behaviour, which in turn increases fear of crime. Not only can avoidance behaviour be learned from and supported by parents, but parents frequently play an active part in it, for instance by driving adolescents to school or picking them up when they go out at night. Because avoidance behaviour has not been included in any previous studies analysing the impact of parental supervision on fear of crime, it is very possible that the previously identified strong direct effect (see, Cops, 2010, 2013; De Groof, 2008; May et al., 2002; May et al., 2015) might (partly) be accounted for by this omission.

Hypothesis 5: The impact of parental supervision on adolescent fear of crime is complex. On the one hand, higher parental supervision leads to more frequent avoidance behaviour and thus higher fear of crime. On the other hand, it lowers adolescent risk perception and consequently fear of crime.

Figure 1.

Model to be fitted.

Methods

Data

The subsequent analysis is based on data from the Urban Youth Victimization Survey (UYVS), which is a self-report school-based survey conducted in 2015 in the Czech Republic. It focuses on 9th grade students and it was conducted in the four largest municipalities, that is, cities with over 150,000 inhabitants. The design of the survey and several parts of the questionnaire are identical to the International Self-Report Delinquency Study 3 (ISRD3 Working Group, 2013; Marshall et al., 2015). Furthermore, some parts of the questionnaire are a follow-up to a victimization survey conducted in Pilsen in 1999 as a part of a comparative project of several European cities. Sampling was based on a random selection of classes and was stratified by school type and city, with the intention of collecting a comparable number of respondents in each city. The response rate was 64 percent at the school level and 78 percent at the individual level, which was primarily accounted for by student absences (refusal to participate in the survey expressed either by students or by their parents was minimal, approximately 2 percent).

The final UYVS sample comprises 1546 children from 85 classes located in 69 schools (54 middle schools and 15 grammar schools). The number of students per city varies from 346 to 410 students. Males and females are represented equally in the sample and the mean age is 15.0 years.

Measures

All latent constructs used in the subsequent analysis as well as the results of a confirmatory factor analysis (CFA) with factor loadings and fit statistics are presented in Table 1. The measurement model fits the data well, that is, fit statistics are acceptable.

Table 1.

Latent constructs and confirmatory factor analysis.

Question	Factor loadings
Fear of crime (4-point scale)
How often do you worry someone will…
…hit you violently or hurt you?	0.585
…threaten you or blackmail you?	0.587
…take your money or something else by force?	0.747
…steal your belongings or money?	0.650
Neighbourhood disorganization (4-point scale)
How much do you agree or disagree with the following statements about your neighbourhood?
There is a lot of crime in my neighbourhood	0.697
There is a lot of drug selling in my neighbourhood	0.812
There is a lot of fighting in my neighbourhood	0.878
There are a lot of empty and abandoned buildings in my neighbourhood	0.588
School disorganization (4-point scale)
How strongly do you agree or disagree with the following statements about your school?
There is a lot of stealing in my school	0.704
There is a lot of fighting in my school	0.716
Many things are broken or vandalized in my school	0.725
There is a lot of drug use in my school	0.460
Constrained behaviour (5-point scale)
Do you sometimes do the following things because you fear something can happen to you or in order to protect yourself better?
I avoid certain streets, places, parks	0.729
I try not to go out in the evening alone	0.726
I avoid coming into contact with certain people. I go, for example, to the other side of the street	0.731
I let others give me a lift or accompany me	0.522
Parental control (5-point scale)
How often do the following statements apply to you?
My parents know where I am when I go out	0.852
My parents know what I am doing when I go out	0.836
My parents know what friends I am with when I go out	0.697
Fit statistics
RMSEA	0.040
CFI	0.955
TLI	0.946

Notes: CFA (Mplus 6); UYVS; N = 1402; standardized coefficients (StdYX); MLR estimation. The variables are coded so that a higher number indicates more frequent worries, avoidance behaviour or parental supervision and higher perceived disorganization.

Fear of crime is measured using questions asking respondents how often they worry someone will hit them violently or hurt them, threaten or blackmail them, take money or something else from them by force, or steal some of their belongings or money; answer categories range from ‘never’ (1) to ‘often’ (4). The same acts are then used to measure risk perception, that is, we asked respondents how likely it was, in their opinion, that some of the mentioned acts would happen to them in the following 12 months. Unlike the original model of fear of crime (Ferraro, 1995), we did not enter risk perception as a latent variable in the analysis. Instead, we calculated an index by averaging values of all four indicators.¹

Both perceived neighbourhood and school disorganization consist of indicators focusing on problematic phenomena that may occur in these environments. In terms of neighbourhood disorganization, these include crime, drug dealing, fighting or the presence of empty and abandoned buildings. Stealing, fighting, vandalism and drug use are considered with respect to the school environment. Respondents were asked to agree or disagree with the occurrence of the phenomena in a given area, with answer categories ranging from ‘I fully disagree’, coded as 1, and ‘I fully agree’, coded as 4.

To measure avoidance behaviour, the respondents were asked if they sometimes behave in a certain way (avoid streets, places, parks, try not to go out in the evening alone, avoid coming into contact with certain people, let others give them a lift or accompany them) because they fear something could happen to them or in order to better protect themselves. Answer categories range from ‘never’ (1) to ‘very often’ (4).

Finally, the level of parental supervision is based on how often parents know where their child is, what he or she is doing and what friends he or she is with while being out. Answer categories range from ‘almost never’ (1) to ‘almost always’ (4).

Table 2 shows descriptive statistics for all manifest variables used in the subsequent analysis, including perceived victimization risk.² Gender is a binary variable where 1 = ‘boy’ and 0 = ‘girl’. In order to measure family SES, we asked respondents to evaluate the position of their family compared with others, with answer categories ranging from 1 = ‘much worse’ to 6 = ‘much better’. Direct victimization is calculated as an index measuring the incidence of various types of victimization (robbery, sexual assault, assault, theft, vandalism and hate crime).³ Indirect victimization represents an index measuring how often a respondent is exposed to violence at school, in the neighbourhood, or between parents. The orientation of the scale is similar in both cases: a higher index value means a higher number of victimization incidents or greater exposure to violent situations. As already mentioned above, risk perception is constructed as a mean of a set of variables measuring the subjectively assessed likelihood that someone would hit the respondent violently or hurt them, threaten or blackmail them, take money or something else from them by force, or steal their belongings or money.

Table 2.

Manifest variables – descriptive statistics.

Variable	Mean	SD	Min.	Max.
Gender	0.50	0.50	0	1
Family SES	4.49	0.99	1	6
Direct victimization	0.41	0.62	0	4.52
Indirect victimization	1.49	0.45	1	3
Risk perception	1.54	0.61	1	5

Notes: Gender is coded as 1 = ‘boy’ and 0 = ‘girl’. Family SES is coded as 1 = ‘low’ and 6 = ‘high’. Direct and indirect victimization and perceived risk of victimization are coded so that a higher number indicates a higher number of cases of victimization or greater exposure to violent situations and higher risk perception.

Analytic strategy

In order to find out whether Ferraro’s risk interpretation model of fear of crime also holds true among the adolescent population and to test our hypotheses, we used Structural Equation Modelling (SEM) in Mplus (version 6). Given the nested structure of our data, we controlled for within-cluster correlation.

First, we estimated the original model (Ferraro, 1995) with gender, family SES, victimization and perceived neighbourhood disorganization as individual and social factors (Model 1). Second, we replaced perceived neighbourhood disorganization with perceived school disorganization, while all other factors in the model remained unchanged (Model 2). Finally, Model 3 reflects both perceived neighbourhood and school disorganization as well as parental supervision. We thus also examined to what extent the level of control parents exercise over their children is associated with the children’s risk perception, perceived neighbourhood or school disorganization, avoidance behaviour and fear of crime.

Findings

Bivariate correlations

Before the SEM analysis was conducted, we checked for the bivariate corrections between all latent and manifest variables (see Table 3 in the Appendix). As expected, we found a relatively strong positive relationship between fear of crime and risk perception (r = 0.57) and fear of crime and avoidance behaviour (r = 0.36). Fear of crime is also somewhat influenced by gender and victimization experience, whereas the correlation between fear of crime and family SES, perceived neighbourhood and school disorganization, indirect victimization and parental supervision is very weak or non-existent.

Furthermore, moderately strong correlations were found between (1) gender and avoidance behaviour (r = −0.40) – revealing that girls have a higher tendency to apply different avoidance strategies than boys; (2) indirect victimization and both perceived neighbourhood and school disorganization (r > 0.30); and (3) direct and indirect victimization (r = 0.31).

SEM results

First, we estimated a model based on Ferraro’s (1995) original model of fear of crime. Its results, as depicted in Model 1 (Figure 2), are consistent with Hypotheses 1 and 2. There is a strong positive relationship between risk perception and fear of crime. In a similar vein, those who declare high risk perception constrain their behaviour more often, which in turn leads to higher fear of crime. Hypothesis 3 can be only partly corroborated since we found no relationship either between perceived neighbourhood disorganization and risk perception or between SES and risk perception. On the other hand, a moderate relationship was found between gender, victimization (both direct and indirect) and risk perception, which in turn influences fear of crime. Nevertheless, the effect of gender and direct victimization on fear of crime is not fully mediated through risk perception; there is still a weak direct effect on fear of crime, as well as a moderate direct effect of gender on avoidance behaviour. Surprisingly, and unlike models estimated for the adult population (Ferraro, 1995), a negative relationship between perceived neighbourhood disorganization and avoidance behaviour was revealed.

Figure 2.

Model 1: Fitted model with neighbourhood disorganization.

Second, we replaced perceived neighbourhood disorganization in Model 1 with perceived school disorganization (Model 2 in Figure 3). In accordance with Hypothesis 4, a relationship between perceived school disorganization and risk perception was found and risk perception was revealed to further mediate the influence of perceived school disorganization on fear of crime.

Figure 3.

Model 2. Fitted model with school disorganization.

Finally, Model 3 (Figure 4) takes into account parental supervision as well as both perceived neighbourhood and school disorganization. In line with Hypothesis 5, parental supervision is positively related to avoidance behaviour, both directly and indirectly, with a total effect of 0.24. On the other hand, the influence of parental supervision on risk perception is negligible (total effect = −0.001). Similarly, the effect of parental supervision on fear of crime – mediated through avoidance behaviour and risk perception – is small and nonsignificant (total effect = 0.07). More information about the total, direct and indirect effects of parental supervision can be found in Appendix (Table 4).

Figure 4.

Model 3: Fitted model with neighbourhood and school disorganization and parental supervision.

Interestingly, the correlation between parental supervision and both perceived neighbourhood and school disorganization is negative: adolescents who are more supervised by their parents thus evaluate their environment to be better than their less supervised peers. In addition, there is also a positive relationship between neighbourhood and school disorganization.

Discussion

As Ferraro (1995) already confirmed in the case of adults, fear of crime presents a complex phenomenon that is influenced by risk perception and constrained behaviour, as well as many individual and social factors. Since knowledge about adolescent fear of crime is still largely limited, the current study broadens the scope of existing research on fear of crime by examining whether Ferraro’s risk interpretation model of fear of crime can also be applied to the adolescent population. Children in mid-adolescence nevertheless differ from adults in many respects – primarily in that they are obliged to attend school and are still dependent on their parents, who are responsible for them. Therefore, we have proposed an extension of Ferraro’s classic model to better reflect the situation of this age group. First, we consider the role of the school environment, where adolescents spend a majority of the day and whose quality can affect their fear of crime (May and Dunaway, 2000). Second, parental supervision is incorporated in the model as a factor relevant for adolescent fear of crime (Cops, 2013; De Groof, 2008; May et al., 2002, 2015).

In accordance with many studies focusing on fear of crime (for example, Farrall et al., 2009; Ferraro, 1995; Gainey et al., 2011; Jackson, 2004; Rountree and Land, 1996), our results confirm a strong relationship between risk perception and fear of crime. Risk perception thus remains the most important predictor of fear of crime, not only among the adult population but also among adolescents. Moreover, risk perception is associated with more frequent avoidance behaviour, which in turn heightens fear of crime and mediates the relationship between fear of crime and various individual factors such as gender and victimization experience. Our study thus presents evidence that the core relationships designated in Ferraro’s (1995) original model of fear of crime are relevant not only for adults but also for the adolescent population.

Nevertheless, there are deviation from Ferraro’s (1995) model. Most importantly, the perception of neighbourhood disorganization seems to function differently among adolescents. Our data reveal no relationship between perceived neighbourhood disorganization and risk perception; moreover, there is a negative correlation between perceived neighbourhood disorganization and avoidance behaviour, suggesting that a higher perception of neighbourhood disorganization results in the less frequent use of behavioural strategies. One explanation for such a negative correlation could be that adolescents who live in disorganized neighbourhoods are accustomed to this fact and do not consider it necessary to change their behaviour in order to protect themselves from victimization (Franklin et al., 2008). However, there is another plausible explanation for the negative relationship between these variables: it is very possible that the direction of the relationship is actually inverse (or reciprocal) and that adolescents who more frequently employ behavioural strategies to avoid victimization have more limited contact with their neighbourhood and, as a result, perceive this setting to be less dangerous. Given the cross-sectional nature of our study, this hypothesis is, however, to be tested in future research using longitudinal data. In addition, with respect to the complex relationship between avoidance behaviour, neighbourhood disorganization and fear of crime, we should keep in mind that, although avoidance behaviour may arouse fear of crime, it may just as well be a consequence of it (for example, Liska et al., 1988; Yuan & McNeeley, 2017).

On the other hand, the perception of school disorganization fits well into Ferraro’s (1995) model, because its influence on fear of crime is fully mediated through risk perception (Models 2 and 3). Hence, it seems that adolescents, unlike adults, evaluate their victimization risk primarily on the basis of stimuli stemming from the school environment and not from the neighbourhood, although it should be acknowledged that there is a positive correlation between the perception of both environments. The importance of the nature of the school setting for the perception of victimization risk is not surprising, given that school is a place where adolescents have to spend a majority of the day in the presence of peers and under the (more or less effective) supervision of teachers. Conversely, the amount of time spent in the neighbourhood is highly variable among adolescents and is, to a certain degree, a matter of choice.

The significant role that parents play in the formation of fear of crime in adolescents is acknowledged by the incorporation of parental supervision in the model (Model 3). The results show that higher parental supervision leads to substantially higher avoidance behaviour, which in turn increases fear of crime. The influence of parental supervision on risk perception, and more importantly on fear of crime, is, however, negligible. This finding is not consistent with a number of previous studies that found a significant direct effect between parental supervision and fear of crime (see Cops, 2010, 2013; De Groof, 2008; May et al., 2002; May et al., 2015). Given that these studies failed to control for avoidance behaviour, it is likely that this purported direct effect arose from the omission of this significant factor.

Finally, our results reveal an unexpected negative relationship between parental supervision and both neighbourhood and school disorganization, signifying that adolescents who are more supervised by their parents perceive both settings more positively. This finding might be the result of the ‘sheltering effect’ of parental supervision, suggested already by May and his colleagues (2002). It is likely that adolescents who are under the intensive control of their parents do not have many opportunities to familiarize themselves with their neighbourhood (see also Cops, 2013) and are thus not as exposed to the negative phenomena occurring in this setting as are their less supervised peers. With respect to the school environment, parents can influence the decision of which school their child will attend. We thus presuppose that parents who are used to intensively supervising their children are also more concerned about their safety and will more likely strive to enrol them in a school with a better reputation than will parents with a more ‘liberal’ parenting style. Such school selectivity is, however, only possible in sufficiently large municipalities with a larger variety of schools; this is precisely the case with our data, which include only youth from large cities. It is thus possible that, if adolescents living in different types of communities were analysed, the correlation between parental supervision and perceived school disorganization might be weaker unless the municipality size is controlled for. Nevertheless, the flip side of the relationship between parental supervision and neighbourhood and school disorganization is also plausible, because families who live in more advantaged neighbourhoods or who place their children in better schools may have more resources to supervise them.

As is the case in other empirical research, our study has several limitations. First, we are working with a relatively narrow range of indicators measuring adolescent fear and perceived risk, which reflect common violent and property offences. Future research could thus examine whether Ferraro’s model of fear of crime also holds true for adolescents when other offences typical for young people such as (cyber) bullying or hate crimes are considered. Second, our data contain only urban pupils of the same age group (that is, 9th graders). Caution should therefore be taken when generalizing our results onto other youth populations. Furthermore, the UYVS was a standard school-based survey and it is likely that our data do not encompass problematic students who often play truant. Last but not least, the cross-sectional nature of our data does not allow us to examine the direction of analysed relationships, and claims of causality are thus based solely on knowledge arising from fear of crime theories and previous research in this field.

Despite these limitations, we can conclude that our study has yielded important results with respect to the applicability of Ferraro’s (1995) risk interpretation model among adolescents. We have confirmed that the core of the model also holds true for this population; nevertheless, it is also necessary to take into account specific factors pertaining to this particular period of adolescence. First of all, it is clear that the quality of the school environment should not be ignored either in future research or by practitioners, because an adverse school setting is linked to an increased awareness of victimization risk, which indirectly increases fear of crime as well. Local authorities should definitely pay attention not only to the educational quality of their schools, but also to the quality of their environments in terms of disorder, bullying and other negative phenomena. It is vital to support schools struggling with such issues or suffering from a negative reputation – for instance through preventive efforts and the transfer of best practices – so that both actual and perceived differences among schools can be diminished. As long as the quality of schools within a single municipality remains grossly disparate, many parents who are concerned about their children’s education and well-being will keep striving to place their children in ‘better’ schools, which is a process that only contributes to the preservation of existing differences among schools.

Second, although the total influence of parental supervision on fear of crime is negligible in our model, it does have an impact on related phenomena, specifically avoidance behaviour. Although the intention of parents to protect their children from harm is generally a positive one, parents should also realize that adolescence is a transitory period of life, in which children should gradually learn to gain autonomy and assume responsibility. Parents should thus be careful not to overprotect their children too much and for too long. Without autonomously exploring the world, they will not learn the necessary strategies to cope with potential dangers (see, for example, Cops, 2013), or subsequently manage negative emotions, including fear of crime.

Footnotes

Appendix

Table 4.

Total, direct and indirect effects of parental supervision on risk perception, avoidance behaviour and fear of crime (Model 3).

	Risk perception	Avoidance behaviour	Fear of crime
Total effect	−0.001	0.242***	0.067
Direct effect	0.016	0.223***	0.003
Indirect effect (total)	−0.017	0.019	0.063*
Specific indirect effects (IE)
IE through avoidance behaviour			0.069***
IE through risk perception		0.003	0.009
IE through neighbourhood disorganization	0.004	0.023**	−0.007
IE through school disorganization	−0.021*	−0.004	−0.004

Note: SEM (Mplus 6), UYVS; N = 1355; standardized coefficients (StdYX).

p < .05, **p < .01, ***p < .001.

Funding

This study was supported by the Czech Science Foundation grant ‘Youth victimization: Prevalence, forms, and social context’ (GP14-08021P).

Notes

References

Agnew

(1985) Neutralizing the impact of crime. Criminal Justice and Behavior 12(2): 221–239.

Akers

La Greca

Sellers

, et al. (1987) Fear of crime and victimization among the elderly in different types of communities. Criminology 25(3): 487–505.

Andreescu

(2010) Victimization and fear of crime in Romania and Hungary: A comparative analysis. Revista Romana de Sociologie 21(3–4): 163–183.

Cops

(2010) Socializing into fear: The impact of socializing institutions on adolescents’ fear of crime. Young 18(4): 385–402.

Cops

(2013) The role of autonomous mobility in public space on fear of crime among adolescents. Journal of Youth Studies 16(8): 1105–1122.

De Groof

(2008) And my mama said: The (relative) parental influence on fear of crime among adolescent girls and boys. Youth & Society 39(3): 267–293.

De Vaus

Wise

(1996) Parents’ concerns for the safety of their children. Family Matters 43: 34–38.

Doran

Burgess

(2012) Putting Fear of Crime on the Map. New York: Springer.

Farrall

Jackson

Gray

(2009) Social Order and the Fear of Crime in Contemporary Times. New York: Oxford University Press.

10.

Ferraro

(1995) Fear of Crime: Interpreting Victimization Risk. Albany: State University of New York Press.

11.

Ferraro

LaGrange

(1987) The measurement of fear of crime. Sociological Inquiry 57(1): 70–97.

12.

Finkelhor

Turner

Ormrod

, et al. (2009) Violence, abuse, and crime exposure in a national sample of children and youth. Pediatrics 124(5): 1411–1423.

13.

Franklin

Fearn

(2008) A multilevel analysis of the vulnerability, disorder, and social integration models of fear of crime. Social Justice Research 21: 204–227.

14.

Furstenberg

(1971) Public reaction to crime in the streets. The American Scholar 40(4): 601–610.

15.

Gainey

Alper

Chappell

(2011) Fear of crime revisited: Examining the direct and indirect effects of disorder, risk perception, and social capital. American Journal of Criminal Justice 36(2): 120–137.

16.

Garofalo

(1979) Victimization and the fear of crime. Journal of Research in Crime and Delinquency 16(1): 80–97.

17.

Garofalo

(1981) The fear of crime: Causes and consequences. Journal of Criminal Law and Criminology 72(2): 839–857.

18.

Gerbner

Gross

(1976) Living with television: The violence profile. Journal of Communication 26(2): 172–194.

19.

Gilchrist

Bannister

Ditton

, et al. (1998) Women and the ‘fear of crime’: Challenging the accepted stereotype. British Journal of Criminology 38(2): 283–298.

20.

Goodey

(1994) Fear of crime: What can children tell us? International Review of Victimology 3(3): 195–210.

21.

Hale

(1996) Fear of crime: A review of the literature. International Review of Victimology 4(2): 79–150.

22.

Hanslmaier

(2013) Crime, fear and subjective well-being: How victimization and street crime affect fear and life satisfaction. European Journal of Criminology 10(5): 515–533.

23.

Hashima

Finkelhor

(1999) Violent victimization of youth versus adults in the National Crime Victimization Survey. Journal of Interpersonal Violence 14(8): 799–820.

24.

ISRD3 Working Group (2013) Questionnaire ISRD3: Standard Student Questionnaire. ISRD3 Technical Report Series #2. URL (accessed 8 July 2017): http://www.northeastern.edu/isrd/wpcontent/uploads/2016/01/ISRD3_TechRep_2.pdf.

25.

Jackson

(2004) Experience and expression: Social and cultural significance in the fear of crime. British Journal of Criminology 44(6): 946–966.

26.

Jackson

Stafford

(2009) Public health and fear of crime: A prospective cohort study. British Journal of Criminology 49(6): 832–847.

27.

Killias

(1990) Vulnerability: Towards a better understanding of a key variable in the genesis of fear of crime. Violence and Victims 5(2): 97–108.

28.

LaGrange

Ferraro

(1989) Assessing age and gender differences in perceived risk and fear of crime. Criminology 27(4): 697–719.

29.

LaGrange

Ferraro

Supancic

(1992) Perceived risk and fear of crime: Role of social and physical incivilities. Journal of Research in Crime and Delinquency 29(3): 311–334.

30.

Lewis

Salem

(1986) Fear of Crime: Incivility and the Production of a Social Problem. New Brunswick, NJ: Transaction Publishers.

31.

Liska

Sanchirico

Reed

(1988) Fear of crime and constrained behavior specifying and estimating a reciprocal effects model. Social Forces 66(3): 827–837.

32.

Lotz

(1979) Public anxiety about crime. Pacific Sociological Review 22(2): 241–254.

33.

Marshall

Enzmann

Hough

Killias

Kivivuori

Steketee

(2015) Youth Victimization and Reporting to Police. First Results from the Third Round of the International Self-Report Delinquency Study (ISRD3). ISRD3 Technical Report Series #3. URL (accessed 8 July 2017): http://www.northeastern.edu/isrd/wpcontent/uploads/2016/01/ISRD3_TechRep_03.pdf.

34.

May

Dunaway

(2000) Predictors of fear of criminal victimization at school among adolescents. Sociological Spectrum 20(2): 149–168.

35.

May

Vartanian

Virgo

(2002) The impact of parental attachment and supervision on fear of crime among adolescent males. Adolescence 37(146): 267–287.

36.

May

Keith

Rader

, et al. (2015) Predicting adolescent fear of crime through the lens of general strain theory. Sociological Focus 48(2): 172–189.

37.

President’s Commission on Law Enforcement and Administration of Justice: Task Force Report (1967) Washington, DC: US Government Printing Office.

38.

Reiner

(2002) Media made criminality: The representation of crime in the mass media. In: Maguire

Morgan

Reiner

(eds) The Oxford Handbook of Criminology. Oxford: Oxford University Press, 302–340.

39.

Rountree

Land

(1996) Perceived risk versus fear of crime: Empirical evidence of conceptually distinct reactions in survey data. Social Forces 74(4): 1353–1376.

40.

Russo

Roccato

Vieno

(2013) Criminal victimization and crime risk perception: A multilevel longitudinal study. Social Indicators Research 112: 535–548.

41.

Sacco

(1995) Media constructions of crime. Annals of the American Academy of Political and Social Science 539: 141–154.

42.

Skogan

Maxfield

(1981) Coping with Crime: Individual and Neighborhood Reactions. London: Sage.

43.

Sutton

Farrall

(2005) Gender, socially desirable responding and the fear of crime: Are women really more anxious about crime? British Journal of Criminology 45(2): 212–224.

44.

Taylor

Hale

(1986) Testing alternative models of fear of crime. Journal of Criminal Law and Criminology 77(1): 151–189.

45.

Tseloni

Zarafonitou

(2008) Fear of crime and victimization: A multivariate multilevel analysis of competing measurements. European Journal of Criminology 5(4): 387–409.

46.

Wallace

May

(2005) The impact of parental attachment and feelings of isolation on adolescent fear of crime at school. Adolescence 40(159): 457–474.

47.

Warr

(1984) Fear of victimization: Why are women and the elderly more afraid. Social Science Quarterly 65(3): 681–702.

48.

Warr

(1985) Fear of rape among urban women. Social Problems 32(3): 238–250.

49.

Warr

(2000) Fear of crime in the United States: Avenues for research and policy. Criminal Justice 4(4): 451–489.

50.

Warr

Stafford

(1983) Fear of victimization: A look at the proximate causes. Social Forces 61(4): 1033–1043.

51.

Weinrath

Gartrell

(1996) Victimization and fear of crime. Violence and Victims 11(3): 187–197.

52.

Williams

Singh

(1994) Urban youth, fear of crime, and resulting defensive actions. Adolescence 29(114): 323–330.

53.

Wyant

(2008) Multilevel impacts of perceived incivilities and perceptions of crime risk on fear of crime: Isolating endogenous impacts. Journal of Research in Crime and Delinquency 45(1): 39–64.

54.

Yuan

McNeeley

(2017) Fear of crime and behavioral adaptations: Testing the effects of fear of violence on unstructured socializing with peers. Deviant Behavior. URL: https://doi.org/10.1080/01639625.2017.1410625.

55.

Zani

Cicognani

Albanesi

(2001) Adolescents’ sense of community and feeling of unsafety in the urban environment. Journal of Community & Applied Social Psychology 11(6): 475–489.