Essays on housing space analysis Open Access
Downloadable ContentDownload PDF
This dissertation demonstrates that the standard urban model (SUM) has important, previously unknown, and rather counterintuitive predictions about the determinants of housing consumption in cities. For example, the SUM predicts that, as higher wages in the CBD prompt city growth, the housing space per household falls, i.e. rising income is associated with falling housing consumption. Empirical testing using a specially constructed panel dataset of U.S. cities, confirms this prediction. When city size, income, and housing price rise, housing space per household falls.Virtually all research on urban housing supply uses the number of housing units as a measure of housing output. Supply differences among cities are based on changes in the number of units as housing price changes. Microeconomic models of the housing market treat housing units as heterogeneous combinations of attributes producing housing services. Logically the supply of housing services should be measured by aggregating housing services rather than units. The dissertation focuses on measuring housing services as square feet of interior space and provides insights in understanding the spatial and temporal variations in housing space. It demonstrates the consequences of conducting housing market analysis based only on housing units instead of housing space. In addition, the theoretical literature has developed many propositions about the determinants of total housing units in cities but neglected propositions regarding the total supply of housing services. The theoretical component of this dissertation therefore fills the gap in the literature as well.