Residential living structure as a basis for the spatial delineation of residential submarkets

Whilst it is generally accepted that residential submarkets exist, this is not the case for either a definition or delineation of residential submarkets. There is a need to incorporate a full range of property attributes (including structural, environmenta, and socioeconomic) describing all dwellings in the study area in order to provide an understanding of the whole residential structure in which submarkets reside. This paper proposes the derivation of a residential living structure (RLS) to achieve this. Principal Component Analysis (PCA) is used to derive the underlying RLS from a wide range of variables including age, size, building construction, house condition, amenity, accessibility to various services and the CBD, together with a range of socio-economic variables. Altogether 65 variables were collected for each of the approximate 440,000 residential properties in the study area. The results presented in this paper reveal 15 underlying components of the RLS in the metropolitan area of Adelaide, South Australia explaining 74% of the variation in the original data. This study extends the existing work in this area by including a wide range of property attributes for all properties in the study area and by establishing the resulting principal components as a set of comprehensive and independent surrogate property characteristics as a basis for further submarket analysis.