Résumé
This thesis attempts to answer the questions of whether or not Winnipeg residentshave become more or less fearful of crime over time and whether changes in fear levelsare related to changes in official crime rates. It also attempts to answer the question of whether certain individuals (women, minority and elderly) experience heightened levelsof fear. Four fear of crime theories; the risk interpretation model, the indirectvictimization model, the vulnerability model and the multiple jeopardy hypothesis aretested using Winnipeg Police official crime data, and 1984, 1994, and2004 WinnipegArea Study survey results. Relationships between fear of crime and official crime ratesare explored using a multiple comparison technique while multiple regression techniqueswere used to estimate the effects of demographic variables on fear of crime.Contrary to the risk interpretation and indirect victimization models, resultsindicate that generally, mean fear levels over the twenty year time span are low, and thereare no consistent associations between fear levels and official crime over the twenty yeartime span. Fear of crime levels increased from 1984 to 1994 then decreased from 1994 to2004. 'Women, visible minorities, the less educated arid married individuals expressed higher levels of fear, while the elderly expressed lower levels of fear. Higher levels offear were expressed by females, particularly when other indicators of vulnerability were added to the regression models. These findings lend partial support to the vulnerability model and multiple jeopardy hypothesis.The results of this thesis suggest that most people in Winnipeg are not that fearful.Fear rates may fluctuate but tend to revert to relatively low levels, even when crime ratesincrease. From a policy perspective, crime is not as important an issue to the public andmay not merit a general increase of resources for crime suppression. A more prudentallocation of resources would see them directed to those most vulnerable: low-incomeminority women.