Parental Fear of Crime: A Discursive Analysis
Tulloch, Marian I., Journal of Sociology
The construct of 'altruistic fear of crime' presents a challenge to the traditional fear of crime research paradigm in which individuals have been questioned about their personal fear of criminal victimization. Self-reported avoidant and protective behaviours have been presented as motivated by a desire for self protection. More recently research into fear of crime has extended to examine fear for the safety of family members. A study by Warr and Ellison (2000) analysed 1996 data from a large-scale, randomized-dialling phone survey of Texas residents, which included questions on respondents' worry about the victimization of significant others, in particular partners and children. They found that respondents' fear for their children, particularly their daughters, was as great or greater than fear for themselves, and although mothers were somewhat more fearful than fathers, the difference in levels of worry was quite small. While fear for sons fluctuated in childhood and early adolescence before declining, fear for daughters remained high into adulthood. Moreover, levels of concern for other family members predicted many anti-crime protective behaviours, which Warr and Ellison argue points to the primacy of social relationships in understanding fear of crime and its influence on behaviour.
Yet, although the study by Wart and Ellison (2000) presents a new perspective on fear of crime with extensive data on worry about victimization of family members, their methodology conforms to the traditional standardized social survey techniques common to much quantitative fear of crime research. Developments in fear of crime research paradigms pose two challenges to the conclusions drawn from generalized fear of crime questions of the type used by Warr and Ellison. Qualitative methodologies have opened up more complex and situated understandings of the meaning of fear of crime in people's lives (e.g. Taylor, 1995; Pain, 1997; Hollway and Jefferson, 2000). Many qualitative studies of fear of crime have taken a narrative or interpretative stance, answering quite different questions than quantitative methodology. Other researchers, however, have argued for a triangulated approach, in which the contextualized and respondent-generated meanings emerging from in-depth interviews provide a more accurate basis for understanding fear of crime than closed survey questions, which may misrepresent the concerns of respondents. Researchers who have combined quantitative and qualitative methods with the same group of respondents have found significant discrepancies in their findings using the two methods. On this basis it is claimed (Farrall et al., 1997: 672) that in-depth qualitative interviews indicate that quantitative measures derived from closed survey questions 'substantially' and 'consistently overestimate levels of worry'. However, they deny privileging of one method over another, instead pointing to the need for avoiding generalities in framing questions, making careful linguistic distinctions (e.g. between 'worrying about' and 'thinking about' a crime) and providing greater specificity, encompassing time, space and social context. More precise, situated questions can elicit more accurate ratings of fear of crime, reducing the probability of measurement error (Farrall et al., 1997: 675). Underlying these prescriptions for accuracy is the positivist premise that fear of crime is operationally definable and measurable. Moreover, when appropriately questioned an individual can produce an 'accurate self assessment of their own crime fear level' by means of introspective self-reporting (Farrall et al., 1997: 662). Hale (1996: 94), in a major review of the fear of crime literature, has a similar view of the way qualitative research can support quantitative methods in reducing 'theoretical casualness and empirical chaos'. The place of qualitative research within this paradigm is in assisting the identification of multiple measurable aspects of a concept. …