The definition and identification of housing submarkets

Although it is widely accepted that urban housing markets are too complex to be described by unitary market, equilibrium models, the role of submarkets has not been embraced in applied research. In this paper it is argued that this is unsurprising and can be traced to the failure to establish a theoretical or empirical basis for submarket modeling. I note that, throughout the housing economics literature, the term 'submarket' is subject to a range of definitions; the means of identifying submarkets has varied; empirical analyses have employed differing tests; and case studies have focused on a range of different cities and different time periods. This inconsistency has prevented the development of a coherent analytical approach. By using data from the Glasgow housing market a range of alternative definition and identification schemes are compared. The evidence suggests that submarkets are important and that, rather than being based exclusively on the similarity of property characteristics or geographical contiguity, the dimensions of housing submarkets are determined by both spatial and structural factors simultaneously.