An Analysis of the Impact of Multiple Environmental Goods on House Prices
Date of This Version
In order to be correctly specified, an hedonic model must include all the relevant housing, neighborhood, and environmental characteristics. If the characteristics are correlated with each other, then if they are not included in the regression the coefficients will not be correctly estimated. This paper uses a unique data set to examine how including multiple environmental indicates affects the price estimates of environmental goods. We find that when we include a single indicator as is done in the previous literature, our results are as expected. However, when we include multiple environmental indicators our results are unstable, with some coefficients having the incorrect sign. Our findings suggest that the environmental variables included might be correlated with unobserved variables other than house or neighborhood characteristics, or that the relationships between the included variables might be more complex than previously thought.
Working Paper Number
Kiel, Katherine and Bowen, Jennifer, "An Analysis of the Impact of Multiple Environmental Goods on House Prices" (2002). Economics Department Working Papers. Paper 117.