The property values of urban single-family housing: a case study of Bloemfontein
Enslin, Albert George
MetadataShow full item record
The ultimate purpose of this study is the construction of a prognostic model of the value of single-family housing. This model can be applied by valuers to determine unbiased, mass valuations of single-family housing for property tax purposes. The property tax, which is the most important independent source of income available to local authorities in South Africa, has continuously been attacked by laymen and academics alike. The primary motivation for these attacks has been the apparent inequitability and inefficiency of the property tax. By improving the administration of the property tax through upgrading and improvement of the valuation procedure, the efficiency of the property tax is enhanced while the equitability of the valuations is ensured. This in turn maximises the financial-resource-generating possibilities of the property tax and makes the tax politically and economically more acceptable. The present erosion of political owing to the increased dependence oon the undershoring of their financial base through financial assistance extended by central government, would to a large extent also be reversed. Improvement of the valuation procedure through the application of the housing price model developed in this study results in spatially neutral valuations being obtained with vertical and horizontal valuation equity resulting. The relative varying contribution of each property within a certain price and land use category to the income of local authorities would now not be the result of the valuation process but the result of a political decision by the locally elected representatives who would, through the determination of the randage or differential randage, determine this relative tax contribution. Through the combined application of factor analysis, cluster analysis and multiple regression analysis, the model of the price of housing that has been constructed, produces market value valuations that show a high degree of correspondence with observed market prices. The predictive ability of the model has been further improved by basing the theoretical foundation of the model on the ecological paradigm. The identification, within the larger overall White single dwelling residential submarket of Bloemfontein and Langenhoven Park, of spatially specific submarkets of integrated economic and socio-geographic status into which the owner-occupiers of single-family housing differentiate and the subsequent construction of separate prediction equations for each submarket, lead to the determination of superior estimates of the market value of housing. Contrary to the conclusion arrived at by Schanre and Struyk in Boston and Ball and Kirwin in Bristol, the greater Bloemfontein property market operates as a series of submarkets. Primary property submarkets (based on the restriction of the property interest activities and segregation of the property activity space of various racial groups) and secondary property submarkets (based on the segregation and restriction of the property interests and property activity space of the agents of supply and allocation and consumers of properties of different land usages) have been distinguished in the greater Bloemfontein area. Within the White single dwelling residential submarket, owner-occupiers, of dwellings, of a similar social status, cluster in spatially specific subareas. This is the result of the propensity of households to live close to other households of similar social status. These spatially specific subareas effectively form submarkets at a third level of disaggregation. The weak cross-price elasticities between submarkets result in independent hedonic price equations of the attributes of housing developing for each submarket. The strength of the coefficients differs for the submarkets because of the importance of different sets of attributes accounting for housing prices in particular submarkets. The importance of specific attributes in the hedonic price equations thus seems to be a function of the peculiar composition of the submarket and of the variability in each submarket of the attributes of housing.