Many citizens fear that they may fall victim to crime amid the increasing crime rates in the society. However, the greater perception of fear of crime among the elderly and women - who are not likely very much to fall victim to crime in routine life - adds the plausibility to the ‘paradox of the fear of crime'. Moreover, the growing interest in the quality of life with economic growth leads to the idea that any restrictions on the behavior which is imposed by the unreasonable fear are not appropriate.
The fear of crime has been the theme of important researches in criminology and other fields for a long time. However, the studies on the crime in this field have focused only on shedding light on the cause of fear and little attention has been paid to such the restrictions on the behavior which stems from the fear of crime. Thus, this study was intended to examine the cause of crime with a focus on the fear of crime, and look into the relationship among the behaviors which result from the fear.
In this study, 519 people, who were over 20 years old and residing in Seoul, were selected as subjects through judgment sampling and quota sampling. In relation to the geographic area of the investigation, three areas of high risk region and low risk region, respectively were selected based on the number of incidents that fell under the five categories of crime(murder, robbery, rape, burglary, violence) and the crime rates(crime incidents per population of 10 million) which were estimated by the Police in Seoul.
In addition, the subjects of this study were grouped by the region according to the age and gender, so that the geographic characteristics of each selected district can be reflected. The collected data was analyzed with SPSS 12.0 and AMOS 18.0 through the confirmatory factor analysis and structural equation modeling.
The research model designed by the analysis of the preceding studies used the independent variables such as the experience of being victimized by crime which is the cause of fear of crime, vulnerability to crime, disorder, crime control. Meanwhile, the perception of risk was based on the fear of crime and the parameter among those independent variables. And the reaction to fear was defined as the dependent variable, assuming that the fear played the role as a parameter.
Specifically, the experience of being victimized by crime was determined by the direct experience of having fallen victim to the crime, indirect experience of being victimized by the crime, and exposures to news report on the crime incidents that occurred in and around their neighborhood. Regarding the vulnerability, the vulnerability to crime that the subjects perceived relatively in comparison to other people, the attractiveness, the ability to resist the victimization of crime, and the ability to recover after being victimized by crime were selected as observational variables.
The disorder in the community was divided into the physical disorder and the social disorder, while the crime control was divided into official control over crimes and the non-official control over crimes.
Additionally, the perception of risk was divided into the perception of risk of crime in the house and the perception of risk of crime on the street or road. Meanwhile, the fear of crime was divided into the fear of the crime in house, the fear of the crime on the street/road, and the general fear of crime. The reactions to fear were divided into the behavior to avoid being victimized by crime and the behavioral and physical actions of defense.
The results of the above-mentioned model analysis suggest that this model is suitable for this study(=479.229 (df=131), p=.000, NFI=.865, CFI=.897, RMSEA=.072). By improving the model through the elimination of the route that was not significant, the revised model 1 was developed(=481.150 (df=136), p=.000, NFI=.865, CFI=.898, RMSEA=.070), and the final model was completed by refining the model again(=443.257 (df=134), p=.000, NFI=.875, CFI=.909, RMSEA=.067).

① The research hypothesis was validated for the relationship among the independent variables, the perception of risk, and the fear of crime. More experience of having fallen victim to crime, higher vulnerability to crime, and more serious disorder in the community were associated with the increased perception of risk(H1-1, H2-1, H3-1 adoption) and the increased fear of crime(H1-2, H2-2, H3-2 adoption).
In this process, it was found that the experience of having fallen victim to crime and the vulnerability to crime had an impact on the fear through the medium of the perception of risk, while the mediating effect of risk perception could not be substantiated in connection with the disorder in the community. Meanwhile, it was shown that crime control in the community did not have any significant effect on the perception of risk and the fear of crime(H 4-1, H4-2 dismissed).
② Based on the review of the reaction to fear of crime and the respective independent variables, it was revealed that more experience of having fallen victim to crime, higher vulnerability to crime, and more serious disorder in the community led to increased reactions to fear(H1-3, H2-3, H3-3 adoption). However, no significant relationship with the fear was substantiated in the case of the crime control in the community, and no mediating effect was found, either(H4-3 dismissed).
On the basis of the review about the mediating effect on the fear of crime and the perception of risk, the vulnerability to crime had the significant mediating effect both in terms of the fear and the fear-risk perception. However, the disorder in the community did not have any significant mediating effect if only the fear is mediated, but showed the significant result when it was mediated by the fear-risk perception.
③ Review on the relationship among the perception of risk, the fear of crime, and the reaction to fear showed that the higher perception of risk resulted in the greater fear of crime and the reaction to the fear(H5, H6 adoption), and the greater fear of crime increased the reaction to fear(H7 adoption). Along with that, the perception of risk was found to have an effect on the reaction to fear through the medium of fear.
This study used the variation of the logical structure inherent in the risk interpretation model proposed by Ferraro(1995), and suggested valuable results in some respects.
① This study was validated to have appropriate logical structure. In other words, the restriction of behaviors does not lead to the fear of crime in the process that the probability of being victimized by the crime is reminded. Rather, the perception of fear of the probability of falling victim to the crime or the crime itself is caused by various reasons, and as a result, behaviors to avoid being victimized by the crime or a defensive behavior are selected.
② The distinction between the perception of risk and the fear of crime became clear. Many studies do not distinguish between the perception of risk and the fear of crime. In this study, it was found that both played clearly different roles when the reaction to fear was considered.
③ It was found that the variable of crime control did not have any effect on the fear of crime and the perception of risk, and others. Many existing studies argue that the social solidarity, police activities at the level of community are important causes of fear of crime. However, they may not become important influential factors if the experience of being victimized by crime at the personal level or the vulnerability to crime, etc, are taken into account altogether.

The results of this study also make significant suggestions as listed below in the aspect of policy:
① Strategies are required which can reduce the experience of falling victim to the crime. It is necessary to consider the discretionary regulatory strategies of the media in relation to the news report of crime because the indirect experience of being victimized by the crime through the mass media plays a more important role compared to the direct experience of being victimized by the crime.
② The crime prevention should be promoted and the ability to defend against the crime should be developed at the personal level, considering that the individual vulnerability to crime has a significant effect on not only the fear but also the restriction of behavior.
③ The disorder in the community should be addressed. Though the elements related to the disorder in the community have decreased very much with the economic development and the improved awareness of citizens, streets overflowing with drunken people, homeless people, etc., still pose problems to the community. Accordingly, social welfare strategies should be considered at the same time, along with the strategies for strict control.
This study is an integrated research which examined both the cause of and reaction to fear of crime with a focus on the fear of crime and the perception of risk, and presents many theoretical and political implications. However, it has some limitations. In other words, the perception of risk and the fear of crime were not determined at the same level in the phase of determination, and in-depth explanation could not be derived on the basis of each observed variables. And there were some limitations in clarifying the causal relationship among respective variables because of the limit of cross-sectional study.
Notwithstanding these limitations, this study is deemed to be meaningful in the light that it attempted the integrated discussion through the expansion of the existing research models and drew more diverse discussions based on the explanation concerning the restrictions of behavior not as a cause but as a result of fear of crime.