Analyzing the impact of soil contamination on farmland values

Abstract

Hedonic studies estimating the effect of environmental disamenities such as soil contamination have focused solely on residential property values thus far. However, since contamination of agricultural land can also cause considerable risks to food safety and hence public health, it is expected to impact farmland values as well. This empirical application aims to fill this research gap by incorporating soil contaminants into a hedonic farmland model. The Campine region, an agricultural area in Belgium that has been historically contaminated with heavy metals – particularly cadmium (Cd) – was used as a case study. Soil Cd concentrations were predicted by means of spatial interpolation techniques and added to 599 farmland transactions that have occurred in the area between 2004 and 2011. In order to take into account some of the issues related to standard spatial econometric techniques, classic linear regression is complemented by quantile regression and a spatiotemporal framework is introduced that only incorporates sales preceding other sales by maximum one year. Goal was to particularly explain the spatial spillover effect caused by previous sales. All regression models found that soil Cd levels did not significantly impact farmland values in the area. Apparently, other factors such as redevelopment potential, current land use and zoning regulations are more important price determinants for agricultural land buyers than soil contamination. The spatiotemporal lag coefficient was found to be highly significant in both mean and median regression models.