The effect of crime on housing values in Rotterdam

Purpose: The purpose of this study is to test the effect of crime on housing values in Rotterdam and to consider how this effect varies for different types of houses in different neighborhoods. Methodology: A hedonic price model is applied in this study. Regarding the statistical tests, one null single level model and five multilevel models have been performed in STATA. To accomplish this, housing data of Rotterdam in 2012 is used. The statistical test is based on a sample of 265.650 houses, which are dispersed over 68 neighborhoods. Results: If the SafetyIndex decreases with 1 point, then the housing value decreases with €9.926, which confirms that crime indeed has a negative effect on housing values. Furthermore, it is found that crime not only has a larger negative effect on the value of big houses compared to small houses, but also on owner-occupied houses compared to rental houses. All of these findings are in line with the expectations created in advance. Limitations: There was no data available for house characteristics such as number of bathrooms and garage, which are likely to have a statistically significant effect on housing prices. In addition, there is a high probability that the dataset still contains exceptions, as it has a large number of observations. Furthermore, this study did not focus on spatial effects, which may affect to some extent the results.