House Value, Crime and Residential Location Choice

Households choose where to live by trading off wages, house prices and local amenities. In this paper, I estimate the effect of crime on household location choice using a two-stage residential sorting model which incorporates the effect of mobility cost. The choice set in this paper is defined at the level of the metropolitan areas. The results from the second stage show that people are willing to pay more to move to a location with lower violent crime occurrences and are willing to pay more to move to a place with higher property crime; however, the effect of violent crime is larger than property crime. When recovering the willingness to pay (WTP) for the two types of crime using elasticities, the results show that people are willing to pay $651 and $977 for a one hundred unit decrease in violent crime and $23 and $27 for a one hundred unit increase in property crime for 2005 and 2010 respectively. The difference in difference results for the sorting model show that people are willing to pay less to move to a location in which the police number increases, and pay more to move to a location where the crime rate decreases while police force increases. The results of the difference in difference analysis, shows that the elasticity of WTP for the increase in police number in the hedonic price model, is slightly lower than that from the sorting model.