The effectiveness of McFaddens’s nested logit model in valuing amenity improvement

Abstract

The paper presents an application of the nested logit model to a large, Chicago housing data set, with the goal of valuing environmental amenities. Four alternative hierarchical nestings based on dwelling, neighborhood, and city attributes are considered, and a sampling rule satisfying McFadden’s uniform conditioning property is adopted. The benefit estimates of dwelling and city attributes are found to be less sensitive, while those of neighborhood attributes are found to be more sensitive to alternative nesting strategies. The effects of household demographics on parameter estimates generally conform to standard notions. A comparison of the model with the standard two-step hedonic model reveals that the benefit estimates are consistently lower in the case of the former.