Environmental hazards : The micro-geography of land-use negative externalities Somerville, Tsur; Wetzel, Jake
The decisions on the siting of hazardous facilities and compensation for nearby landowners depends on an accurate estimation of the negative externalities these facilities place on proximate land uses, primarily residential properties. In this paper we highlight the sensitivity of these estimates to the treatment of distance from the hazard and to the presence of other nearby land uses identified at a highly granular geographic level. Recent opposition to the expansion of North American pipeline capacity has been intense, mixing concerns about climate change, environmental damage, and local opposition to the physical presence of pipelines in their neighbourhoods. This paper studies the disamenity effects associated with the last factor. In doing so we generate results that more broadly address the specification and left out variable bias challenges of measuring the capitalization of negative location-specific environmental externalities. The key contributions of this paper are first showing that disamenity effects can be highly localized and easily susceptible to errors with parametric specifications. Second, that the magnitude of the effect on house prices arising from proximity are sensitive to land uses that are not the hazard in question, but whose presence may be correlated with the hazard. And third, that negative news about a hazard increases the assessment of risk and lowers nearby house values, but that this effect is temporary. We find that the quantitative effects of proximity to oil pipelines are relatively small: prices are lower by 5.7 percent ($39.3k) for properties with a pipeline easement, 2.1 percent ($14.4k) lower for those properties adjacent to a property with an easement, and 1.4 percent ($9.6k) for those adjacent to the former, one property further away from the pipeline. Though this last result is sensitive to specification choice. The prices of all residential properties further away from the pipeline in our data are unaffected. When expressed in cardinal distance, only the prices for residential properties within 100 meters of the pipeline easement are affected. The findings here suggest that care and flexibility with functional forms, the perception of hazards, and attention to land use contexts is necessary for an analysis of the negative externalities for residential property associated with proximity to environmental disamenities and that simple parametric treatments are highly likely to result in biased estimates.
Item Citations and Data
Attribution-NonCommercial-NoDerivatives 4.0 International