Measuring the effects of perceptions of crime on neighborhood quality and housing markets

This dissertation examines the effect of crime on neighborhood quality through an examination of house prices. Instead of a straight examination of the effect of measured levels of crime on housing prices, this research explores the effect of how crime is perceived. Studies using crime statistics may be misleading due to issues due to issues with mismeasurement of crime and homebuyers' awareness of crime statistics and riskaverse behavior of homebuyers. This is achieved using two approaches. The first uses survey data to compare how the effect of the satisfaction with safety from crime on house prices differs from the effect of measured crime rates on house prices. The second examines whether there are asymmetries in the perceptions of increases and decreases in crime rates and also examines how long the effects last as measured through effects on house prices.; When hedonic models of house prices using the perception of crime are compared to models using measured crimes, neither measure is a good indicator of house prices. The results show that measured crime rates and perception of crime do not robustly predict house prices and in some cases indicate higher house prices associated with higher crime rates and dissatisfaction with safety in the neighborhood. The results from the analysis of hedonic price models using increases and decreases in crime do not show evidence that changes in crime should not be measured as a continuous linear variable. The results suffer from a lack of significant regression coefficients, leaving little to say about whether increases in crime are different from decreases in crime.