The nature and impacts of environmental spillovers on housing prices: A spatial hedonic analysis

This paper investigates the spatial dimension of the environmental effects. We use recent advances in spatial econometrics to show that hedonic equations produce estimates to be differently interpreted as implicit prices according to spatial models. In particularly, the implicit price of housing attribute combines a feedback effect and a propagation effect and may be interpreted in terms of local or global spillovers. We drive an empirical study in the estuary of the Loire, a rural and urban area well occupied by various natural areas and more artificialized ones. We study various spatial interaction patterns to test the robustness of our estimates and we find that spatial dependencies based on inverse distance and small neighborhoods provide stable estimations. It is consistent too with realistic spatial interaction patterns for household behaviors: information on closer housings is more reliable and comparison areas are in fact limited by the research process. As expected, positive impacts are concentrated on traditional attributes like the proximity to the ocean frontage and quiet places. On the contrary, the presence of various natural wet amenities is negatively valued because of the impression of housing density associated to flood risk. If urban places are more valued by households, it's rather because rural location are less desired than because of urban intrinsic attributes.