A Spatial Model of Housing Attributes: Theory and Evidence

Abstract

The last decade has witnessed rapid growth of an empirical literature designed to estimate and interpret hedonic price functions. Although early studies (see Kain and Quigley 1970 and Grether and Mieszkowski 1974) lacked an explicit theoretical foundation, recent papers (see Linneman 1981 and Witte et al. 1979) reflect the impact of Rosen's (1974) 'implicitmarket' analysis of hedonic price determination. In spite of this empirical focus on the pricing of attributes, theoretical work in economics has relied almost exclusively on housing models, in which is unrealistically portrayed as a unidimensional commodity. In a recent attempt to reconcile the empirical and theoretical traditions, Brueckner (1980) developed a Rosen-type model of attributes which borrows heavily from the service approach. In the model, consumers value two attributes, floor space and yard space. Developers react to a consumer bid-rent function, which relates dwelling rent to the levels of the attributes, in choosing profitmaximizing characteristics of complexes. Among the theoretical questions addressed by the model, those concerning the spatial behavior of floor and yard space per dwelling are of particular interest: Will dwelling size (floor space) increase with distance to the central business district (CBD), as in the standard model? Will yards be larger farther from the CBD? The next section of the paper sketches the theory and its predictions, providing an-