Proximity to Natural Amenities: A Seemingly Unrelated Hedonic Regression Model with Spatial Durbin and Spatial Error Processes

Abstract The existing literature suggests that the presence of natural amenities, such as open spaces, can be highly valued and affect economic decisions about where people live and work. This article contributes to previous research by testing this hypothesis using a unique micro‐level data set and by examining spatial variations in income levels and housing prices in the presence of natural amenities in a case study of Arizona. Proximity effects are estimated based on a geographic information system road network in which each variable represents the road mile distance from house i to its closest natural amenity within each category. Using a seemingly unrelated regression approach, spatial hedonic regressions of housing prices and income levels indicate that the total effect of various natural amenities calculated for the sample average income household and average home value, ranges from $2,382 (National Forests) to $1,560 (Wilderness areas). The presence of compensating differentials has policy relevance in considering the regional value of natural amenities. It also implies that valuation approaches such as the travel cost method may not reflect the full price of recreation site access, and may lead to underestimates of such values.