|
Article Excerpt
Introduction
Throughout the last decade in North America, Canadian and American crime rates have been declining, including high-profile offences such as homicide (Wallace 2004). However, surprisingly, opinion polls show that crime remains a consistent topic of concern, despite the fall in reported offences (Roberts 2001). This puzzle leads to our central question: Are people fearful of crime, regardless of whether crime rates increase or decline? If not, what is the connection between official crime rate changes and people's perception of their victimization risk?
Changes in fear levels regarding crime over time are important to policy makers. Higher fear levels within the general public indicate that greater resources are needed in areas of crime control and community development. Intuitively, one would expect that reported crime rates would directly influence fear. If this is not the case, more effort may be needed by government agencies to provide a realistic estimate of risk to the public (Roberts and Stalans 1997).
Our paper examines the changes in fear of crime over time, focusing on the medium-sized city of Winnipeg, Manitoba. We assess whether or not any increases or decreases in fear levels are related to changes in the official crime rates. Three years (1984, 1994, and 2004) of survey data from the Winnipeg Area Study and 26 years of official crime data from the Winnipeg Police Service are used to examine the changes in fear over a 20-year time span. We incorporate a multivariate model to control for possible demographic shifts that might influence changes in fear over time. In addition, we use fear survey data and official crime data to partially test two theories: the risk interpretation model and the indirect victimization model.
Theoretical models
The three most influential fear-of-crime theories are the risk interpretation model, the indirect victimization model, and the vulnerability model. Our study focuses on the first two theories, while we use the vulnerability model to identify suitable control variables for individual attributes. The risk interpretation model proposed by Kenneth Ferraro focuses on the role that perceived risk plays in fear of crime. Ferraro (1995) derived his model by integrating three sociological theories: symbolic interactionism, incivility theory, and criminal opportunity theory.
The first part of Ferraro's model involves symbolic interactionism, originally derived from Herbert Blumer, which examines how people gather and interpret information around them, changing perceptions over time. Ferraro (1995) uses symbolic interactionism to argue that people redefine risk in relation to local events or experiences. The incivility perspective claims that individuals see certain features of a physical environment as "signs of crime," which often serve as cues to actors that risk might be higher in those areas, e.g., groups of teens hanging out, abandoned/broken down buildings, use of drugs, and various forms of vandalism (Ferraro 1995; Skogan and Maxfield 1981). Finally, Ferraro's risk interpretation model incorporates criminal opportunity theory. This theory states that offenders make "rational" choices about various criminal opportunities in the environment. Potential offenders may take the advantage of information about living quarters, police protection, and neighbourhood surveillance in judging the risk of apprehension. In a similar fashion, Ferraro proposes that potential victims use this information to judge the risk of being victimized. For example, one is aware that assault is more likely while walking down an unlit back lane late at night, as opposed to walking down a busy street. Taken together, symbolic interactionism, incivility, and criminal opportunity models infer that social actors will redefine their circumstances as crime rates go up and report higher fear levels. This will be particularly true for lower-income areas where incivility is likely to coexist with high crime rates.
Taylor and Hale's (1986) indirect victimization model outlines a process that links crime rates with fear. This model is based on two premises: "shock waves" and vulnerability. Crimes are thought to generate shock waves that induce fear, because people who hear about crimes become afraid, thereby becoming indirect victims. Skogan and Maxfield (1981) find that most people do not learn about crime through their own recent and direct victimization experiences. Rather, people rely upon the media and stories from acquaintances to learn about crime. Personal experience and direct observation of victimization tends to be rare, particularly violent victimization. On the other hand, news of crime plays upon an individual's sense of vulnerability. Fear may become more intense when people learn of local crimes affecting their friends and neighbours, especially if they share some bond or sense of common fate. People feel sympathetic when consequences are tragic, but they are also reminded of their own vulnerability (Chiricos, Padgett, and Gertz 2000; Skogan and Maxfield 1981).
While local news is important, one cannot discount the importance of media generally in producing shock waves and promoting feelings of vulnerability. The media have been accused of engendering fear because, although crime information is spread widely, it does not parallel the distribution of actual victimization. Media coverage of crime is frequent and centered around violence, homicide in particular. The sheer volume of violence featured on television and in newspapers provides an exaggerated view of the actual frequency of such crime, and thus can engender higher levels of fear (Chricos et al. 2000; Heath and Gilbert 1996; Heath, Kavanagh, and Thompson 2001; Skogan and Maxfield 1981). The media seem to have an especially large influence upon fear in an urban setting such as Winnipeg, which city has experienced higher levels of violent crime than other Canadian cities (Sauve 2005).
If the amount of violent crime has increased in an area, one would expect that that media reports and neighbourhood conversations about crime would increase significantly as well (Skogan and Maxfield 1981). This may be particularly true in high-crime neighbourhoods, which might experience a "double whammy" of media coverage of serious overall crime in addition to local information. The indirect victimization model suggests that increases in violent crime will be consistent with increases in urban residents' fear of crime. Local increases in violent crime are thought likely to incur greater increases in fear.
The vulnerability model is the most common theoretical framework used in the literature. It postulates that groups that perceive themselves more likely to be victimized are more fearful. We do not intend to test this model, but do intend to use it to guide our introduction of control variables. Studies involving the relationship between fear and vulnerability most consistently focus on variables such as gender, age, race, income, and martial status, as these are found to reflect the underlying dimensions of physical and social vulnerability to crime (Ferraro 1995; Ferraro and La Grange 1988; Houts and Kassab 1997; Madriz 1997; Parker and Ray 1990; Rountree and Land 1996; Sacco 1994; Smith and Hill 1991; Weinrath 1999; Weinrath and Gartrell 1996).
Fear definitions
Most of the survey research done in the area of fear of crime relies on what Pantazis (2000) calls the "global measure of fear"; in other words, the "formless fear" question. Fear of crime is typically measured by asking, "Is there any area right around here, that is within a mile where you would be afraid to walk alone at night?" (Forde 1993; Haynie 1998; Roberts 2001; Warr 1995). This measure has been criticized as being too hypothetical, limited to night-time, not mentioning crime, and only crudely estimating intensity (Ferraro and LaGrange 1988; Warr 2000). It is also criticized for expecting respondents to define what "safe" means, and what constitutes a neighbourhood (Christian 2001).
A key conceptual issue frequently discussed in the literature is that of fear vs. perceived risk. Mesch (2000) argues that the distinction between fear of crime and perceived risk has been an important contribution to the field. Fear of crime is defined as "an emotional response of dread or anxiety to crime or symbols that a person associates with crime" (Ferraro 1995: 4). Ferraro defines perceived risk as "a recognition of a situation as possessing at least potential danger, real or imagined" (1995: 4). While the two concepts appear related, a number of studies have adduced empirical evidence that they are quite distinct and are affected by different factors (Mesch 2000). Warr (2000) has measured fear and risk estimation by looking at five offence-specific indicators: theft, fraud, assault, sexual assault, and robbery. For example, he measures fear of sexual predation by asking respondents, "How often do you worry about someone sexually assaulting you?". His measures give the respondent some context in which to answer the question (Warr 2000; Rountree 1998). We favour this approach and use similar questions in...
|
